[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>

int panic_on_oops;
int tainted;
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 int __init panic_setup(char *str)
{
	panic_timeout = simple_strtoul(str, NULL, 0);
	return 1;
}
__setup("panic=", panic_setup);

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);

/**
 *	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.
 *
 *	The string is overwritten by the next call to print_taint().
 */

const char *print_tainted(void)
{
	static char buf[20];
	if (tainted) {
		snprintf(buf, sizeof(buf), "Tainted: %c%c%c%c%c%c%c%c",
			tainted & TAINT_PROPRIETARY_MODULE ? 'P' : 'G',
			tainted & TAINT_FORCED_MODULE ? 'F' : ' ',
			tainted & TAINT_UNSAFE_SMP ? 'S' : ' ',
			tainted & TAINT_FORCED_RMMOD ? 'R' : ' ',
			tainted & TAINT_MACHINE_CHECK ? 'M' : ' ',
			tainted & TAINT_BAD_PAGE ? 'B' : ' ',
			tainted & TAINT_USER ? 'U' : ' ',
			tainted & TAINT_DIE ? 'D' : ' ');
	}
	else
		snprintf(buf, sizeof(buf), "Not tainted");
	return(buf);
}

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

static int __init pause_on_oops_setup(char *str)
{
	pause_on_oops = simple_strtoul(str, NULL, 0);
	return 1;
}
__setup("pause_on_oops=", pause_on_oops_setup);

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);

/*
 * Called when the architecture exits its oops handler, after printing
 * everything.
 */
void oops_exit(void)
{
	do_oops_enter_exit();
	init_oops_id();
	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
		(unsigned long long)oops_id);
}

#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
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>
1da177e4	23
1da177e4 LT	24	int panic_on_oops;
1da177e4 LT	25	int tainted;
dd287796 AM	26	static int pause_on_oops;
	27	static int pause_on_oops_flag;
	28	static DEFINE_SPINLOCK(pause_on_oops_lock);
1da177e4	29
dd287796	30	int panic_timeout;
1da177e4	31
e041c683	32	ATOMIC_NOTIFIER_HEAD(panic_notifier_list);
1da177e4 LT	33
	34	EXPORT_SYMBOL(panic_notifier_list);
	35
	36	static int __init panic_setup(char *str)
	37	{
	38	panic_timeout = simple_strtoul(str, NULL, 0);
	39	return 1;
	40	}
	41	__setup("panic=", panic_setup);
	42
	43	static long no_blink(long time)
	44	{
	45	return 0;
	46	}
	47
	48	/* Returns how long it waited in ms */
	49	long (*panic_blink)(long time);
	50	EXPORT_SYMBOL(panic_blink);
	51
	52	/**
	53	* panic - halt the system
	54	* @fmt: The text string to print
	55	*
	56	* Display a message, then perform cleanups.
	57	*
	58	* This function never returns.
	59	*/
c277e63f	60
1da177e4 LT	61	NORET_TYPE void panic(const char * fmt, ...)
	62	{
	63	long i;
	64	static char buf[1024];
	65	va_list args;
347a8dc3	66	#if defined(CONFIG_S390)
c277e63f	67	unsigned long caller = (unsigned long) __builtin_return_address(0);
1da177e4 LT	68	#endif
1da177e4 LT	69
dc009d92 EB	70	/*
	71	* It's possible to come here directly from a panic-assertion and not
	72	* have preempt disabled. Some functions called from here want
	73	* preempt to be disabled. No point enabling it later though...
	74	*/
	75	preempt_disable();
	76
1da177e4 LT	77	bust_spinlocks(1);
	78	va_start(args, fmt);
	79	vsnprintf(buf, sizeof(buf), fmt, args);
	80	va_end(args);
	81	printk(KERN_EMERG "Kernel panic - not syncing: %s\n",buf);
	82	bust_spinlocks(0);
	83
dc009d92 EB	84	/*
	85	* If we have crashed and we have a crash kernel loaded let it handle
	86	* everything else.
	87	* Do we want to call this before we try to display a message?
	88	*/
6e274d14	89	crash_kexec(NULL);
dc009d92	90
1da177e4	91	#ifdef CONFIG_SMP
dc009d92 EB	92	/*
	93	* Note smp_send_stop is the usual smp shutdown function, which
	94	* unfortunately means it may not be hardened to work in a panic
	95	* situation.
	96	*/
1da177e4 LT	97	smp_send_stop();
	98	#endif
	99
e041c683	100	atomic_notifier_call_chain(&panic_notifier_list, 0, buf);
1da177e4 LT	101
	102	if (!panic_blink)
	103	panic_blink = no_blink;
	104
dc009d92	105	if (panic_timeout > 0) {
1da177e4 LT	106	/*
	107	* Delay timeout seconds before rebooting the machine.
	108	* We can't use the "normal" timers since we just panicked..
	109	*/
	110	printk(KERN_EMERG "Rebooting in %d seconds..",panic_timeout);
	111	for (i = 0; i < panic_timeout*1000; ) {
	112	touch_nmi_watchdog();
	113	i += panic_blink(i);
	114	mdelay(1);
	115	i++;
	116	}
2f048ea8 EB	117	/* This will not be a clean reboot, with everything
	118	* shutting down. But if there is a chance of
	119	* rebooting the system it will be rebooted.
1da177e4	120	*/
2f048ea8	121	emergency_restart();
1da177e4 LT	122	}
	123	#ifdef __sparc__
	124	{
	125	extern int stop_a_enabled;
a271c241	126	/* Make sure the user can actually press Stop-A (L1-A) */
1da177e4	127	stop_a_enabled = 1;
a271c241	128	printk(KERN_EMERG "Press Stop-A (L1-A) to return to the boot prom\n");
1da177e4 LT	129	}
1da177e4 LT	130	#endif
347a8dc3	131	#if defined(CONFIG_S390)
c277e63f	132	disabled_wait(caller);
1da177e4 LT	133	#endif
	134	local_irq_enable();
	135	for (i = 0;;) {
c22db941	136	touch_softlockup_watchdog();
1da177e4 LT	137	i += panic_blink(i);
	138	mdelay(1);
	139	i++;
	140	}
	141	}
	142
	143	EXPORT_SYMBOL(panic);
	144
	145	/**
	146	* print_tainted - return a string to represent the kernel taint state.
	147	*
	148	* 'P' - Proprietary module has been loaded.
	149	* 'F' - Module has been forcibly loaded.
	150	* 'S' - SMP with CPUs not designed for SMP.
	151	* 'R' - User forced a module unload.
9aa5e993	152	* 'M' - System experienced a machine check exception.
1da177e4	153	* 'B' - System has hit bad_page.
34f5a398	154	* 'U' - Userspace-defined naughtiness.
1da177e4 LT	155	*
	156	* The string is overwritten by the next call to print_taint().
	157	*/
c277e63f	158
1da177e4 LT	159	const char *print_tainted(void)
	160	{
	161	static char buf[20];
	162	if (tainted) {
bcdcd8e7	163	snprintf(buf, sizeof(buf), "Tainted: %c%c%c%c%c%c%c%c",
1da177e4 LT	164	tainted & TAINT_PROPRIETARY_MODULE ? 'P' : 'G',
	165	tainted & TAINT_FORCED_MODULE ? 'F' : ' ',
	166	tainted & TAINT_UNSAFE_SMP ? 'S' : ' ',
	167	tainted & TAINT_FORCED_RMMOD ? 'R' : ' ',
c277e63f	168	tainted & TAINT_MACHINE_CHECK ? 'M' : ' ',
34f5a398	169	tainted & TAINT_BAD_PAGE ? 'B' : ' ',
bcdcd8e7 PE	170	tainted & TAINT_USER ? 'U' : ' ',
bcdcd8e7 PE	171	tainted & TAINT_DIE ? 'D' : ' ');
1da177e4 LT	172	}
	173	else
	174	snprintf(buf, sizeof(buf), "Not tainted");
	175	return(buf);
	176	}
	177
	178	void add_taint(unsigned flag)
	179	{
068c4579	180	debug_locks = 0; /* can't trust the integrity of the kernel anymore */
1da177e4 LT	181	tainted \|= flag;
	182	}
	183	EXPORT_SYMBOL(add_taint);
dd287796 AM	184
	185	static int __init pause_on_oops_setup(char *str)
	186	{
	187	pause_on_oops = simple_strtoul(str, NULL, 0);
	188	return 1;
	189	}
	190	__setup("pause_on_oops=", pause_on_oops_setup);
	191
	192	static void spin_msec(int msecs)
	193	{
	194	int i;
	195
	196	for (i = 0; i < msecs; i++) {
	197	touch_nmi_watchdog();
	198	mdelay(1);
	199	}
	200	}
	201
	202	/*
	203	* It just happens that oops_enter() and oops_exit() are identically
	204	* implemented...
	205	*/
	206	static void do_oops_enter_exit(void)
	207	{
	208	unsigned long flags;
	209	static int spin_counter;
	210
	211	if (!pause_on_oops)
	212	return;
	213
	214	spin_lock_irqsave(&pause_on_oops_lock, flags);
	215	if (pause_on_oops_flag == 0) {
	216	/* This CPU may now print the oops message */
	217	pause_on_oops_flag = 1;
	218	} else {
	219	/* We need to stall this CPU */
	220	if (!spin_counter) {
	221	/* This CPU gets to do the counting */
	222	spin_counter = pause_on_oops;
	223	do {
	224	spin_unlock(&pause_on_oops_lock);
	225	spin_msec(MSEC_PER_SEC);
	226	spin_lock(&pause_on_oops_lock);
	227	} while (--spin_counter);
	228	pause_on_oops_flag = 0;
	229	} else {
	230	/* This CPU waits for a different one */
	231	while (spin_counter) {
	232	spin_unlock(&pause_on_oops_lock);
	233	spin_msec(1);
	234	spin_lock(&pause_on_oops_lock);
	235	}
	236	}
	237	}
	238	spin_unlock_irqrestore(&pause_on_oops_lock, flags);
	239	}
	240
	241	/*
	242	* Return true if the calling CPU is allowed to print oops-related info. This
	243	* is a bit racy..
	244	*/
	245	int oops_may_print(void)
	246	{
	247	return pause_on_oops_flag == 0;
248	}
249
250	/*
251	* Called when the architecture enters its oops handler, before it prints
252	* anything. If this is the first CPU to oops, and it's oopsing the first time
253	* then let it proceed.
254	*
255	* This is all enabled by the pause_on_oops kernel boot option. We do all this
256	* to ensure that oopses don't scroll off the screen. It has the side-effect
257	* of preventing later-oopsing CPUs from mucking up the display, too.
258	*
259	* It turns out that the CPU which is allowed to print ends up pausing for the
260	* right duration, whereas all the other CPUs pause for twice as long: once in
261	* oops_enter(), once in oops_exit().
262	*/
263	void oops_enter(void)
264	{
2c16e9c8	265	debug_locks_off(); /* can't trust the integrity of the kernel anymore */
dd287796 AM	266	do_oops_enter_exit();
	267	}
	268
2c3b20e9 AV	269	/*
	270	* 64-bit random ID for oopses:
	271	*/
	272	static u64 oops_id;
	273
	274	static int init_oops_id(void)
	275	{
	276	if (!oops_id)
	277	get_random_bytes(&oops_id, sizeof(oops_id));
	278
	279	return 0;
	280	}
	281	late_initcall(init_oops_id);
	282
dd287796 AM	283	/*
	284	* Called when the architecture exits its oops handler, after printing
	285	* everything.
	286	*/
	287	void oops_exit(void)
	288	{
	289	do_oops_enter_exit();
2c3b20e9 AV	290	init_oops_id();
	291	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
	292	(unsigned long long)oops_id);
dd287796	293	}
3162f751 AV	294
	295	#ifdef CONFIG_CC_STACKPROTECTOR
	296	/*
	297	* Called when gcc's -fstack-protector feature is used, and
	298	* gcc detects corruption of the on-stack canary value
	299	*/
	300	void __stack_chk_fail(void)
	301	{
	302	panic("stack-protector: Kernel stack is corrupted");
	303	}
	304	EXPORT_SYMBOL(__stack_chk_fail);
	305	#endif