[net-next-2.6.git] / kernel / panic.c

/*
 *  linux/kernel/panic.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

/*
 * This function is used through-out the kernel (including mm and fs)
 * to indicate a major problem.
 */
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/delay.h>
#include <linux/reboot.h>
#include <linux/notifier.h>
#include <linux/init.h>
#include <linux/sysrq.h>
#include <linux/interrupt.h>
#include <linux/nmi.h>
#include <linux/kexec.h>
#include <linux/debug_locks.h>
#include <linux/random.h>
#include <linux/kallsyms.h>

int panic_on_oops;
static unsigned long tainted_mask;
static int pause_on_oops;
static int pause_on_oops_flag;
static DEFINE_SPINLOCK(pause_on_oops_lock);

int panic_timeout;

ATOMIC_NOTIFIER_HEAD(panic_notifier_list);

EXPORT_SYMBOL(panic_notifier_list);

static long no_blink(long time)
{
	return 0;
}

/* Returns how long it waited in ms */
long (*panic_blink)(long time);
EXPORT_SYMBOL(panic_blink);

/**
 *	panic - halt the system
 *	@fmt: The text string to print
 *
 *	Display a message, then perform cleanups.
 *
 *	This function never returns.
 */

NORET_TYPE void panic(const char * fmt, ...)
{
	long i;
	static char buf[1024];
	va_list args;
#if defined(CONFIG_S390)
	unsigned long caller = (unsigned long) __builtin_return_address(0);
#endif

	/*
	 * It's possible to come here directly from a panic-assertion and not
	 * have preempt disabled. Some functions called from here want
	 * preempt to be disabled. No point enabling it later though...
	 */
	preempt_disable();

	bust_spinlocks(1);
	va_start(args, fmt);
	vsnprintf(buf, sizeof(buf), fmt, args);
	va_end(args);
	printk(KERN_EMERG "Kernel panic - not syncing: %s\n",buf);
	bust_spinlocks(0);

	/*
	 * If we have crashed and we have a crash kernel loaded let it handle
	 * everything else.
	 * Do we want to call this before we try to display a message?
	 */
	crash_kexec(NULL);

#ifdef CONFIG_SMP
	/*
	 * Note smp_send_stop is the usual smp shutdown function, which
	 * unfortunately means it may not be hardened to work in a panic
	 * situation.
	 */
	smp_send_stop();
#endif

	atomic_notifier_call_chain(&panic_notifier_list, 0, buf);

	if (!panic_blink)
		panic_blink = no_blink;

	if (panic_timeout > 0) {
		/*
	 	 * Delay timeout seconds before rebooting the machine. 
		 * We can't use the "normal" timers since we just panicked..
	 	 */
		printk(KERN_EMERG "Rebooting in %d seconds..",panic_timeout);
		for (i = 0; i < panic_timeout*1000; ) {
			touch_nmi_watchdog();
			i += panic_blink(i);
			mdelay(1);
			i++;
		}
		/*	This will not be a clean reboot, with everything
		 *	shutting down.  But if there is a chance of
		 *	rebooting the system it will be rebooted.
		 */
		emergency_restart();
	}
#ifdef __sparc__
	{
		extern int stop_a_enabled;
		/* Make sure the user can actually press Stop-A (L1-A) */
		stop_a_enabled = 1;
		printk(KERN_EMERG "Press Stop-A (L1-A) to return to the boot prom\n");
	}
#endif
#if defined(CONFIG_S390)
	disabled_wait(caller);
#endif
	local_irq_enable();
	for (i = 0;;) {
		touch_softlockup_watchdog();
		i += panic_blink(i);
		mdelay(1);
		i++;
	}
}

EXPORT_SYMBOL(panic);


struct tnt {
	u8 bit;
	char true;
	char false;
};

static const struct tnt tnts[] = {
	{ TAINT_PROPRIETARY_MODULE, 'P', 'G' },
	{ TAINT_FORCED_MODULE, 'F', ' ' },
	{ TAINT_UNSAFE_SMP, 'S', ' ' },
	{ TAINT_FORCED_RMMOD, 'R', ' ' },
	{ TAINT_MACHINE_CHECK, 'M', ' ' },
	{ TAINT_BAD_PAGE, 'B', ' ' },
	{ TAINT_USER, 'U', ' ' },
	{ TAINT_DIE, 'D', ' ' },
	{ TAINT_OVERRIDDEN_ACPI_TABLE, 'A', ' ' },
	{ TAINT_WARN, 'W', ' ' },
	{ TAINT_CRAP, 'C', ' ' },
};

/**
 *	print_tainted - return a string to represent the kernel taint state.
 *
 *  'P' - Proprietary module has been loaded.
 *  'F' - Module has been forcibly loaded.
 *  'S' - SMP with CPUs not designed for SMP.
 *  'R' - User forced a module unload.
 *  'M' - System experienced a machine check exception.
 *  'B' - System has hit bad_page.
 *  'U' - Userspace-defined naughtiness.
 *  'A' - ACPI table overridden.
 *  'W' - Taint on warning.
 *  'C' - modules from drivers/staging are loaded.
 *
 *	The string is overwritten by the next call to print_taint().
 */
const char *print_tainted(void)
{
	static char buf[ARRAY_SIZE(tnts) + sizeof("Tainted: ") + 1];

	if (tainted_mask) {
		char *s;
		int i;

		s = buf + sprintf(buf, "Tainted: ");
		for (i = 0; i < ARRAY_SIZE(tnts); i++) {
			const struct tnt *t = &tnts[i];
			*s++ = test_bit(t->bit, &tainted_mask) ?
					t->true : t->false;
		}
		*s = 0;
	} else
		snprintf(buf, sizeof(buf), "Not tainted");
	return(buf);
}

int test_taint(unsigned flag)
{
	return test_bit(flag, &tainted_mask);
}
EXPORT_SYMBOL(test_taint);

unsigned long get_taint(void)
{
	return tainted_mask;
}

void add_taint(unsigned flag)
{
	debug_locks = 0; /* can't trust the integrity of the kernel anymore */
	set_bit(flag, &tainted_mask);
}
EXPORT_SYMBOL(add_taint);

static void spin_msec(int msecs)
{
	int i;

	for (i = 0; i < msecs; i++) {
		touch_nmi_watchdog();
		mdelay(1);
	}
}

/*
 * It just happens that oops_enter() and oops_exit() are identically
 * implemented...
 */
static void do_oops_enter_exit(void)
{
	unsigned long flags;
	static int spin_counter;

	if (!pause_on_oops)
		return;

	spin_lock_irqsave(&pause_on_oops_lock, flags);
	if (pause_on_oops_flag == 0) {
		/* This CPU may now print the oops message */
		pause_on_oops_flag = 1;
	} else {
		/* We need to stall this CPU */
		if (!spin_counter) {
			/* This CPU gets to do the counting */
			spin_counter = pause_on_oops;
			do {
				spin_unlock(&pause_on_oops_lock);
				spin_msec(MSEC_PER_SEC);
				spin_lock(&pause_on_oops_lock);
			} while (--spin_counter);
			pause_on_oops_flag = 0;
		} else {
			/* This CPU waits for a different one */
			while (spin_counter) {
				spin_unlock(&pause_on_oops_lock);
				spin_msec(1);
				spin_lock(&pause_on_oops_lock);
			}
		}
	}
	spin_unlock_irqrestore(&pause_on_oops_lock, flags);
}

/*
 * Return true if the calling CPU is allowed to print oops-related info.  This
 * is a bit racy..
 */
int oops_may_print(void)
{
	return pause_on_oops_flag == 0;
}

/*
 * Called when the architecture enters its oops handler, before it prints
 * anything.  If this is the first CPU to oops, and it's oopsing the first time
 * then let it proceed.
 *
 * This is all enabled by the pause_on_oops kernel boot option.  We do all this
 * to ensure that oopses don't scroll off the screen.  It has the side-effect
 * of preventing later-oopsing CPUs from mucking up the display, too.
 *
 * It turns out that the CPU which is allowed to print ends up pausing for the
 * right duration, whereas all the other CPUs pause for twice as long: once in
 * oops_enter(), once in oops_exit().
 */
void oops_enter(void)
{
	debug_locks_off(); /* can't trust the integrity of the kernel anymore */
	do_oops_enter_exit();
}

/*
 * 64-bit random ID for oopses:
 */
static u64 oops_id;

static int init_oops_id(void)
{
	if (!oops_id)
		get_random_bytes(&oops_id, sizeof(oops_id));

	return 0;
}
late_initcall(init_oops_id);

static void print_oops_end_marker(void)
{
	init_oops_id();
	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
		(unsigned long long)oops_id);
}

/*
 * Called when the architecture exits its oops handler, after printing
 * everything.
 */
void oops_exit(void)
{
	do_oops_enter_exit();
	print_oops_end_marker();
}

#ifdef WANT_WARN_ON_SLOWPATH
void warn_slowpath(const char *file, int line, const char *fmt, ...)
{
	va_list args;
	char function[KSYM_SYMBOL_LEN];
	unsigned long caller = (unsigned long)__builtin_return_address(0);
	sprint_symbol(function, caller);

	printk(KERN_WARNING "------------[ cut here ]------------\n");
	printk(KERN_WARNING "WARNING: at %s:%d %s()\n", file,
		line, function);

	if (fmt) {
		va_start(args, fmt);
		vprintk(fmt, args);
		va_end(args);
	}

	print_modules();
	dump_stack();
	print_oops_end_marker();
	add_taint(TAINT_WARN);
}
EXPORT_SYMBOL(warn_slowpath);

void warn_on_slowpath(const char *file, int line)
{
	warn_slowpath(file, line, NULL);
}
EXPORT_SYMBOL(warn_on_slowpath);
#endif

#ifdef CONFIG_CC_STACKPROTECTOR
/*
 * Called when gcc's -fstack-protector feature is used, and
 * gcc detects corruption of the on-stack canary value
 */
void __stack_chk_fail(void)
{
	panic("stack-protector: Kernel stack is corrupted");
}
EXPORT_SYMBOL(__stack_chk_fail);
#endif

core_param(panic, panic_timeout, int, 0644);
core_param(pause_on_oops, pause_on_oops, int, 0644);
Commit	Line	Data
1da177e4 LT	1	/*
	2	* linux/kernel/panic.c
	3	*
	4	* Copyright (C) 1991, 1992 Linus Torvalds
	5	*/
	6
	7	/*
	8	* This function is used through-out the kernel (including mm and fs)
	9	* to indicate a major problem.
	10	*/
1da177e4 LT	11	#include <linux/module.h>
	12	#include <linux/sched.h>
	13	#include <linux/delay.h>
	14	#include <linux/reboot.h>
	15	#include <linux/notifier.h>
	16	#include <linux/init.h>
	17	#include <linux/sysrq.h>
	18	#include <linux/interrupt.h>
	19	#include <linux/nmi.h>
dc009d92	20	#include <linux/kexec.h>
657b3010	21	#include <linux/debug_locks.h>
2c3b20e9	22	#include <linux/random.h>
79b4cc5e	23	#include <linux/kallsyms.h>
1da177e4	24
1da177e4	25	int panic_on_oops;
25ddbb18	26	static unsigned long tainted_mask;
dd287796 AM	27	static int pause_on_oops;
	28	static int pause_on_oops_flag;
	29	static DEFINE_SPINLOCK(pause_on_oops_lock);
1da177e4	30
dd287796	31	int panic_timeout;
1da177e4	32
e041c683	33	ATOMIC_NOTIFIER_HEAD(panic_notifier_list);
1da177e4 LT	34
	35	EXPORT_SYMBOL(panic_notifier_list);
	36
1da177e4 LT	37	static long no_blink(long time)
	38	{
	39	return 0;
	40	}
	41
	42	/* Returns how long it waited in ms */
	43	long (*panic_blink)(long time);
	44	EXPORT_SYMBOL(panic_blink);
	45
	46	/**
	47	* panic - halt the system
	48	* @fmt: The text string to print
	49	*
	50	* Display a message, then perform cleanups.
	51	*
	52	* This function never returns.
	53	*/
c277e63f	54
1da177e4 LT	55	NORET_TYPE void panic(const char * fmt, ...)
	56	{
	57	long i;
	58	static char buf[1024];
	59	va_list args;
347a8dc3	60	#if defined(CONFIG_S390)
c277e63f	61	unsigned long caller = (unsigned long) __builtin_return_address(0);
1da177e4 LT	62	#endif
1da177e4 LT	63
dc009d92 EB	64	/*
	65	* It's possible to come here directly from a panic-assertion and not
	66	* have preempt disabled. Some functions called from here want
	67	* preempt to be disabled. No point enabling it later though...
	68	*/
	69	preempt_disable();
	70
1da177e4 LT	71	bust_spinlocks(1);
	72	va_start(args, fmt);
	73	vsnprintf(buf, sizeof(buf), fmt, args);
	74	va_end(args);
	75	printk(KERN_EMERG "Kernel panic - not syncing: %s\n",buf);
	76	bust_spinlocks(0);
	77
dc009d92 EB	78	/*
	79	* If we have crashed and we have a crash kernel loaded let it handle
	80	* everything else.
	81	* Do we want to call this before we try to display a message?
	82	*/
6e274d14	83	crash_kexec(NULL);
dc009d92	84
1da177e4	85	#ifdef CONFIG_SMP
dc009d92 EB	86	/*
	87	* Note smp_send_stop is the usual smp shutdown function, which
	88	* unfortunately means it may not be hardened to work in a panic
	89	* situation.
	90	*/
1da177e4 LT	91	smp_send_stop();
	92	#endif
	93
e041c683	94	atomic_notifier_call_chain(&panic_notifier_list, 0, buf);
1da177e4 LT	95
	96	if (!panic_blink)
	97	panic_blink = no_blink;
	98
dc009d92	99	if (panic_timeout > 0) {
1da177e4 LT	100	/*
	101	* Delay timeout seconds before rebooting the machine.
	102	* We can't use the "normal" timers since we just panicked..
	103	*/
	104	printk(KERN_EMERG "Rebooting in %d seconds..",panic_timeout);
	105	for (i = 0; i < panic_timeout*1000; ) {
	106	touch_nmi_watchdog();
	107	i += panic_blink(i);
	108	mdelay(1);
	109	i++;
	110	}
2f048ea8 EB	111	/* This will not be a clean reboot, with everything
	112	* shutting down. But if there is a chance of
	113	* rebooting the system it will be rebooted.
1da177e4	114	*/
2f048ea8	115	emergency_restart();
1da177e4 LT	116	}
	117	#ifdef __sparc__
	118	{
	119	extern int stop_a_enabled;
a271c241	120	/* Make sure the user can actually press Stop-A (L1-A) */
1da177e4	121	stop_a_enabled = 1;
a271c241	122	printk(KERN_EMERG "Press Stop-A (L1-A) to return to the boot prom\n");
1da177e4 LT	123	}
1da177e4 LT	124	#endif
347a8dc3	125	#if defined(CONFIG_S390)
c277e63f	126	disabled_wait(caller);
1da177e4 LT	127	#endif
	128	local_irq_enable();
	129	for (i = 0;;) {
c22db941	130	touch_softlockup_watchdog();
1da177e4 LT	131	i += panic_blink(i);
	132	mdelay(1);
	133	i++;
	134	}
	135	}
	136
	137	EXPORT_SYMBOL(panic);
	138
c277e63f	139
25ddbb18 AK	140	struct tnt {
	141	u8 bit;
	142	char true;
	143	char false;
	144	};
	145
	146	static const struct tnt tnts[] = {
	147	{ TAINT_PROPRIETARY_MODULE, 'P', 'G' },
	148	{ TAINT_FORCED_MODULE, 'F', ' ' },
	149	{ TAINT_UNSAFE_SMP, 'S', ' ' },
	150	{ TAINT_FORCED_RMMOD, 'R', ' ' },
	151	{ TAINT_MACHINE_CHECK, 'M', ' ' },
	152	{ TAINT_BAD_PAGE, 'B', ' ' },
	153	{ TAINT_USER, 'U', ' ' },
	154	{ TAINT_DIE, 'D', ' ' },
	155	{ TAINT_OVERRIDDEN_ACPI_TABLE, 'A', ' ' },
	156	{ TAINT_WARN, 'W', ' ' },
26e9a397	157	{ TAINT_CRAP, 'C', ' ' },
25ddbb18 AK	158	};
25ddbb18 AK	159
6d5cd6ef RD	160	/**
	161	* print_tainted - return a string to represent the kernel taint state.
	162	*
	163	* 'P' - Proprietary module has been loaded.
	164	* 'F' - Module has been forcibly loaded.
	165	* 'S' - SMP with CPUs not designed for SMP.
	166	* 'R' - User forced a module unload.
	167	* 'M' - System experienced a machine check exception.
	168	* 'B' - System has hit bad_page.
	169	* 'U' - Userspace-defined naughtiness.
	170	* 'A' - ACPI table overridden.
	171	* 'W' - Taint on warning.
061b1bd3	172	* 'C' - modules from drivers/staging are loaded.
6d5cd6ef RD	173	*
	174	* The string is overwritten by the next call to print_taint().
	175	*/
1da177e4 LT	176	const char *print_tainted(void)
1da177e4 LT	177	{
25ddbb18 AK	178	static char buf[ARRAY_SIZE(tnts) + sizeof("Tainted: ") + 1];
	179
	180	if (tainted_mask) {
	181	char *s;
	182	int i;
	183
	184	s = buf + sprintf(buf, "Tainted: ");
	185	for (i = 0; i < ARRAY_SIZE(tnts); i++) {
	186	const struct tnt *t = &tnts[i];
	187	*s++ = test_bit(t->bit, &tainted_mask) ?
	188	t->true : t->false;
	189	}
	190	*s = 0;
	191	} else
1da177e4 LT	192	snprintf(buf, sizeof(buf), "Not tainted");
	193	return(buf);
	194	}
	195
25ddbb18 AK	196	int test_taint(unsigned flag)
	197	{
	198	return test_bit(flag, &tainted_mask);
	199	}
	200	EXPORT_SYMBOL(test_taint);
	201
	202	unsigned long get_taint(void)
	203	{
	204	return tainted_mask;
	205	}
	206
1da177e4 LT	207	void add_taint(unsigned flag)
1da177e4 LT	208	{
068c4579	209	debug_locks = 0; /* can't trust the integrity of the kernel anymore */
25ddbb18	210	set_bit(flag, &tainted_mask);
1da177e4 LT	211	}
1da177e4 LT	212	EXPORT_SYMBOL(add_taint);
dd287796	213
dd287796 AM	214	static void spin_msec(int msecs)
	215	{
	216	int i;
	217
	218	for (i = 0; i < msecs; i++) {
	219	touch_nmi_watchdog();
	220	mdelay(1);
	221	}
	222	}
	223
	224	/*
	225	* It just happens that oops_enter() and oops_exit() are identically
	226	* implemented...
	227	*/
	228	static void do_oops_enter_exit(void)
	229	{
	230	unsigned long flags;
	231	static int spin_counter;
	232
	233	if (!pause_on_oops)
	234	return;
	235
	236	spin_lock_irqsave(&pause_on_oops_lock, flags);
	237	if (pause_on_oops_flag == 0) {
	238	/* This CPU may now print the oops message */
	239	pause_on_oops_flag = 1;
	240	} else {
	241	/* We need to stall this CPU */
	242	if (!spin_counter) {
	243	/* This CPU gets to do the counting */
	244	spin_counter = pause_on_oops;
	245	do {
	246	spin_unlock(&pause_on_oops_lock);
	247	spin_msec(MSEC_PER_SEC);
	248	spin_lock(&pause_on_oops_lock);
	249	} while (--spin_counter);
	250	pause_on_oops_flag = 0;
	251	} else {
	252	/* This CPU waits for a different one */
	253	while (spin_counter) {
	254	spin_unlock(&pause_on_oops_lock);
	255	spin_msec(1);
	256	spin_lock(&pause_on_oops_lock);
	257	}
	258	}
	259	}
	260	spin_unlock_irqrestore(&pause_on_oops_lock, flags);
	261	}
	262
	263	/*
	264	* Return true if the calling CPU is allowed to print oops-related info. This
	265	* is a bit racy..
	266	*/
	267	int oops_may_print(void)
	268	{
	269	return pause_on_oops_flag == 0;
	270	}
	271
	272	/*
	273	* Called when the architecture enters its oops handler, before it prints
	274	* anything. If this is the first CPU to oops, and it's oopsing the first time
	275	* then let it proceed.
	276	*
	277	* This is all enabled by the pause_on_oops kernel boot option. We do all this
278	* to ensure that oopses don't scroll off the screen. It has the side-effect
279	* of preventing later-oopsing CPUs from mucking up the display, too.
280	*
281	* It turns out that the CPU which is allowed to print ends up pausing for the
282	* right duration, whereas all the other CPUs pause for twice as long: once in
283	* oops_enter(), once in oops_exit().
284	*/
285	void oops_enter(void)
286	{
2c16e9c8	287	debug_locks_off(); /* can't trust the integrity of the kernel anymore */
dd287796 AM	288	do_oops_enter_exit();
	289	}
	290
2c3b20e9 AV	291	/*
	292	* 64-bit random ID for oopses:
	293	*/
	294	static u64 oops_id;
	295
	296	static int init_oops_id(void)
	297	{
	298	if (!oops_id)
	299	get_random_bytes(&oops_id, sizeof(oops_id));
	300
	301	return 0;
	302	}
	303	late_initcall(init_oops_id);
	304
71c33911 AV	305	static void print_oops_end_marker(void)
	306	{
	307	init_oops_id();
	308	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
	309	(unsigned long long)oops_id);
	310	}
	311
dd287796 AM	312	/*
	313	* Called when the architecture exits its oops handler, after printing
	314	* everything.
	315	*/
	316	void oops_exit(void)
	317	{
	318	do_oops_enter_exit();
71c33911	319	print_oops_end_marker();
dd287796	320	}
3162f751	321
79b4cc5e	322	#ifdef WANT_WARN_ON_SLOWPATH
a8f18b90 AV	323	void warn_slowpath(const char file, int line, const char fmt, ...)
	324	{
	325	va_list args;
	326	char function[KSYM_SYMBOL_LEN];
	327	unsigned long caller = (unsigned long)__builtin_return_address(0);
	328	sprint_symbol(function, caller);
	329
	330	printk(KERN_WARNING "------------[ cut here ]------------\n");
	331	printk(KERN_WARNING "WARNING: at %s:%d %s()\n", file,
	332	line, function);
74853dba AV	333
	334	if (fmt) {
	335	va_start(args, fmt);
	336	vprintk(fmt, args);
	337	va_end(args);
	338	}
a8f18b90 AV	339
	340	print_modules();
	341	dump_stack();
	342	print_oops_end_marker();
	343	add_taint(TAINT_WARN);
	344	}
	345	EXPORT_SYMBOL(warn_slowpath);
74853dba AV	346
	347	void warn_on_slowpath(const char *file, int line)
	348	{
	349	warn_slowpath(file, line, NULL);
	350	}
	351	EXPORT_SYMBOL(warn_on_slowpath);
79b4cc5e AV	352	#endif
79b4cc5e AV	353
3162f751 AV	354	#ifdef CONFIG_CC_STACKPROTECTOR
	355	/*
	356	* Called when gcc's -fstack-protector feature is used, and
	357	* gcc detects corruption of the on-stack canary value
	358	*/
	359	void __stack_chk_fail(void)
	360	{
	361	panic("stack-protector: Kernel stack is corrupted");
	362	}
	363	EXPORT_SYMBOL(__stack_chk_fail);
	364	#endif
f44dd164 RR	365
	366	core_param(panic, panic_timeout, int, 0644);
	367	core_param(pause_on_oops, pause_on_oops, int, 0644);