[deliverable/linux.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>
#include <linux/dmi.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);
#ifdef CONFIG_DEBUG_BUGVERBOSE
	dump_stack();
#endif

	/*
	 * 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);

	/*
	 * 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();

	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++;
	}
	bust_spinlocks(0);
}

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.
 *  'D' - Kernel has oopsed before
 *  '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));
	else
		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);
	const char *board;

	sprint_symbol(function, caller);

	printk(KERN_WARNING "------------[ cut here ]------------\n");
	printk(KERN_WARNING "WARNING: at %s:%d %s()\n", file,
		line, function);
	board = dmi_get_system_info(DMI_PRODUCT_NAME);
	if (board)
		printk(KERN_WARNING "Hardware name: %s\n", board);

	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);
#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 in: %p\n",
		__builtin_return_address(0));
}
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>
bd89bb29	24	#include <linux/dmi.h>
1da177e4	25
1da177e4	26	int panic_on_oops;
25ddbb18	27	static unsigned long tainted_mask;
dd287796 AM	28	static int pause_on_oops;
	29	static int pause_on_oops_flag;
	30	static DEFINE_SPINLOCK(pause_on_oops_lock);
1da177e4	31
dd287796	32	int panic_timeout;
1da177e4	33
e041c683	34	ATOMIC_NOTIFIER_HEAD(panic_notifier_list);
1da177e4 LT	35
	36	EXPORT_SYMBOL(panic_notifier_list);
	37
1da177e4 LT	38	static long no_blink(long time)
	39	{
	40	return 0;
	41	}
	42
	43	/* Returns how long it waited in ms */
	44	long (*panic_blink)(long time);
	45	EXPORT_SYMBOL(panic_blink);
	46
	47	/**
	48	* panic - halt the system
	49	* @fmt: The text string to print
	50	*
	51	* Display a message, then perform cleanups.
	52	*
	53	* This function never returns.
	54	*/
c277e63f	55
1da177e4 LT	56	NORET_TYPE void panic(const char * fmt, ...)
	57	{
	58	long i;
	59	static char buf[1024];
	60	va_list args;
347a8dc3	61	#if defined(CONFIG_S390)
c277e63f	62	unsigned long caller = (unsigned long) __builtin_return_address(0);
1da177e4 LT	63	#endif
1da177e4 LT	64
dc009d92 EB	65	/*
	66	* It's possible to come here directly from a panic-assertion and not
	67	* have preempt disabled. Some functions called from here want
	68	* preempt to be disabled. No point enabling it later though...
	69	*/
	70	preempt_disable();
	71
1da177e4 LT	72	bust_spinlocks(1);
	73	va_start(args, fmt);
	74	vsnprintf(buf, sizeof(buf), fmt, args);
	75	va_end(args);
	76	printk(KERN_EMERG "Kernel panic - not syncing: %s\n",buf);
5cb27301 IM	77	#ifdef CONFIG_DEBUG_BUGVERBOSE
	78	dump_stack();
	79	#endif
1da177e4	80
dc009d92 EB	81	/*
	82	* If we have crashed and we have a crash kernel loaded let it handle
	83	* everything else.
	84	* Do we want to call this before we try to display a message?
	85	*/
6e274d14	86	crash_kexec(NULL);
dc009d92	87
dc009d92 EB	88	/*
	89	* Note smp_send_stop is the usual smp shutdown function, which
	90	* unfortunately means it may not be hardened to work in a panic
	91	* situation.
	92	*/
1da177e4	93	smp_send_stop();
1da177e4	94
e041c683	95	atomic_notifier_call_chain(&panic_notifier_list, 0, buf);
1da177e4 LT	96
	97	if (!panic_blink)
	98	panic_blink = no_blink;
	99
dc009d92	100	if (panic_timeout > 0) {
1da177e4 LT	101	/*
	102	* Delay timeout seconds before rebooting the machine.
	103	* We can't use the "normal" timers since we just panicked..
	104	*/
	105	printk(KERN_EMERG "Rebooting in %d seconds..",panic_timeout);
	106	for (i = 0; i < panic_timeout*1000; ) {
	107	touch_nmi_watchdog();
	108	i += panic_blink(i);
	109	mdelay(1);
	110	i++;
	111	}
2f048ea8 EB	112	/* This will not be a clean reboot, with everything
	113	* shutting down. But if there is a chance of
	114	* rebooting the system it will be rebooted.
1da177e4	115	*/
2f048ea8	116	emergency_restart();
1da177e4 LT	117	}
	118	#ifdef __sparc__
	119	{
	120	extern int stop_a_enabled;
a271c241	121	/* Make sure the user can actually press Stop-A (L1-A) */
1da177e4	122	stop_a_enabled = 1;
a271c241	123	printk(KERN_EMERG "Press Stop-A (L1-A) to return to the boot prom\n");
1da177e4 LT	124	}
1da177e4 LT	125	#endif
347a8dc3	126	#if defined(CONFIG_S390)
c277e63f	127	disabled_wait(caller);
1da177e4 LT	128	#endif
	129	local_irq_enable();
	130	for (i = 0;;) {
c22db941	131	touch_softlockup_watchdog();
1da177e4 LT	132	i += panic_blink(i);
	133	mdelay(1);
	134	i++;
	135	}
ffd71da4	136	bust_spinlocks(0);
1da177e4 LT	137	}
	138
	139	EXPORT_SYMBOL(panic);
	140
c277e63f	141
25ddbb18 AK	142	struct tnt {
	143	u8 bit;
	144	char true;
	145	char false;
	146	};
	147
	148	static const struct tnt tnts[] = {
	149	{ TAINT_PROPRIETARY_MODULE, 'P', 'G' },
	150	{ TAINT_FORCED_MODULE, 'F', ' ' },
	151	{ TAINT_UNSAFE_SMP, 'S', ' ' },
	152	{ TAINT_FORCED_RMMOD, 'R', ' ' },
	153	{ TAINT_MACHINE_CHECK, 'M', ' ' },
	154	{ TAINT_BAD_PAGE, 'B', ' ' },
	155	{ TAINT_USER, 'U', ' ' },
	156	{ TAINT_DIE, 'D', ' ' },
	157	{ TAINT_OVERRIDDEN_ACPI_TABLE, 'A', ' ' },
	158	{ TAINT_WARN, 'W', ' ' },
26e9a397	159	{ TAINT_CRAP, 'C', ' ' },
25ddbb18 AK	160	};
25ddbb18 AK	161
1da177e4 LT	162	/**
	163	* print_tainted - return a string to represent the kernel taint state.
	164	*
	165	* 'P' - Proprietary module has been loaded.
	166	* 'F' - Module has been forcibly loaded.
	167	* 'S' - SMP with CPUs not designed for SMP.
	168	* 'R' - User forced a module unload.
9aa5e993	169	* 'M' - System experienced a machine check exception.
1da177e4	170	* 'B' - System has hit bad_page.
34f5a398	171	* 'U' - Userspace-defined naughtiness.
a8005992	172	* 'D' - Kernel has oopsed before
95b570c9 NH	173	* 'A' - ACPI table overridden.
95b570c9 NH	174	* 'W' - Taint on warning.
061b1bd3	175	* 'C' - modules from drivers/staging are loaded.
1da177e4 LT	176	*
	177	* The string is overwritten by the next call to print_taint().
	178	*/
1da177e4 LT	179	const char *print_tainted(void)
1da177e4 LT	180	{
25ddbb18 AK	181	static char buf[ARRAY_SIZE(tnts) + sizeof("Tainted: ") + 1];
	182
	183	if (tainted_mask) {
	184	char *s;
	185	int i;
	186
	187	s = buf + sprintf(buf, "Tainted: ");
	188	for (i = 0; i < ARRAY_SIZE(tnts); i++) {
	189	const struct tnt *t = &tnts[i];
	190	*s++ = test_bit(t->bit, &tainted_mask) ?
	191	t->true : t->false;
	192	}
	193	*s = 0;
	194	} else
1da177e4 LT	195	snprintf(buf, sizeof(buf), "Not tainted");
	196	return(buf);
	197	}
	198
25ddbb18	199	int test_taint(unsigned flag)
1da177e4	200	{
25ddbb18 AK	201	return test_bit(flag, &tainted_mask);
	202	}
	203	EXPORT_SYMBOL(test_taint);
	204
	205	unsigned long get_taint(void)
	206	{
	207	return tainted_mask;
1da177e4	208	}
dd287796	209
1da177e4	210	void add_taint(unsigned flag)
dd287796	211	{
068c4579	212	debug_locks = 0; /* can't trust the integrity of the kernel anymore */
25ddbb18	213	set_bit(flag, &tainted_mask);
dd287796	214	}
1da177e4	215	EXPORT_SYMBOL(add_taint);
dd287796 AM	216
	217	static void spin_msec(int msecs)
	218	{
	219	int i;
	220
	221	for (i = 0; i < msecs; i++) {
	222	touch_nmi_watchdog();
	223	mdelay(1);
	224	}
	225	}
	226
	227	/*
	228	* It just happens that oops_enter() and oops_exit() are identically
	229	* implemented...
	230	*/
	231	static void do_oops_enter_exit(void)
	232	{
	233	unsigned long flags;
	234	static int spin_counter;
	235
	236	if (!pause_on_oops)
	237	return;
	238
	239	spin_lock_irqsave(&pause_on_oops_lock, flags);
	240	if (pause_on_oops_flag == 0) {
	241	/* This CPU may now print the oops message */
	242	pause_on_oops_flag = 1;
	243	} else {
	244	/* We need to stall this CPU */
	245	if (!spin_counter) {
	246	/* This CPU gets to do the counting */
	247	spin_counter = pause_on_oops;
	248	do {
	249	spin_unlock(&pause_on_oops_lock);
	250	spin_msec(MSEC_PER_SEC);
	251	spin_lock(&pause_on_oops_lock);
	252	} while (--spin_counter);
	253	pause_on_oops_flag = 0;
	254	} else {
	255	/* This CPU waits for a different one */
	256	while (spin_counter) {
	257	spin_unlock(&pause_on_oops_lock);
	258	spin_msec(1);
	259	spin_lock(&pause_on_oops_lock);
	260	}
	261	}
	262	}
	263	spin_unlock_irqrestore(&pause_on_oops_lock, flags);
	264	}
	265
	266	/*
	267	* Return true if the calling CPU is allowed to print oops-related info. This
	268	* is a bit racy..
	269	*/
	270	int oops_may_print(void)
	271	{
	272	return pause_on_oops_flag == 0;
	273	}
	274
	275	/*
	276	* Called when the architecture enters its oops handler, before it prints
	277	* anything. If this is the first CPU to oops, and it's oopsing the first time
	278	* then let it proceed.
	279	*
280	* This is all enabled by the pause_on_oops kernel boot option. We do all this
281	* to ensure that oopses don't scroll off the screen. It has the side-effect
282	* of preventing later-oopsing CPUs from mucking up the display, too.
283	*
284	* It turns out that the CPU which is allowed to print ends up pausing for the
285	* right duration, whereas all the other CPUs pause for twice as long: once in
286	* oops_enter(), once in oops_exit().
287	*/
288	void oops_enter(void)
289	{
2c16e9c8	290	debug_locks_off(); /* can't trust the integrity of the kernel anymore */
dd287796 AM	291	do_oops_enter_exit();
	292	}
	293
2c3b20e9 AV	294	/*
	295	* 64-bit random ID for oopses:
	296	*/
	297	static u64 oops_id;
	298
	299	static int init_oops_id(void)
	300	{
	301	if (!oops_id)
	302	get_random_bytes(&oops_id, sizeof(oops_id));
d6624f99 AV	303	else
d6624f99 AV	304	oops_id++;
2c3b20e9 AV	305
	306	return 0;
	307	}
	308	late_initcall(init_oops_id);
	309
71c33911 AV	310	static void print_oops_end_marker(void)
	311	{
	312	init_oops_id();
	313	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
	314	(unsigned long long)oops_id);
	315	}
	316
dd287796 AM	317	/*
	318	* Called when the architecture exits its oops handler, after printing
	319	* everything.
	320	*/
	321	void oops_exit(void)
	322	{
	323	do_oops_enter_exit();
71c33911	324	print_oops_end_marker();
dd287796	325	}
3162f751	326
79b4cc5e	327	#ifdef WANT_WARN_ON_SLOWPATH
a8f18b90 AV	328	void warn_slowpath(const char file, int line, const char fmt, ...)
	329	{
	330	va_list args;
	331	char function[KSYM_SYMBOL_LEN];
	332	unsigned long caller = (unsigned long)__builtin_return_address(0);
bd89bb29 AV	333	const char *board;
bd89bb29 AV	334
a8f18b90 AV	335	sprint_symbol(function, caller);
	336
	337	printk(KERN_WARNING "------------[ cut here ]------------\n");
	338	printk(KERN_WARNING "WARNING: at %s:%d %s()\n", file,
	339	line, function);
bd89bb29 AV	340	board = dmi_get_system_info(DMI_PRODUCT_NAME);
	341	if (board)
	342	printk(KERN_WARNING "Hardware name: %s\n", board);
74853dba AV	343
	344	if (fmt) {
	345	va_start(args, fmt);
	346	vprintk(fmt, args);
	347	va_end(args);
	348	}
a8f18b90 AV	349
	350	print_modules();
	351	dump_stack();
	352	print_oops_end_marker();
	353	add_taint(TAINT_WARN);
	354	}
	355	EXPORT_SYMBOL(warn_slowpath);
79b4cc5e AV	356	#endif
79b4cc5e AV	357
3162f751	358	#ifdef CONFIG_CC_STACKPROTECTOR
54371a43	359
3162f751 AV	360	/*
	361	* Called when gcc's -fstack-protector feature is used, and
	362	* gcc detects corruption of the on-stack canary value
	363	*/
	364	void __stack_chk_fail(void)
	365	{
517a92c4 IM	366	panic("stack-protector: Kernel stack is corrupted in: %p\n",
517a92c4 IM	367	__builtin_return_address(0));
3162f751 AV	368	}
3162f751 AV	369	EXPORT_SYMBOL(__stack_chk_fail);
54371a43	370
3162f751	371	#endif
f44dd164 RR	372
	373	core_param(panic, panic_timeout, int, 0644);
	374	core_param(pause_on_oops, pause_on_oops, int, 0644);