[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;
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.
 *  'A' - ACPI table overridden.
 *  'W' - Taint on warning.
 *
 *	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%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' : ' ',
			tainted & TAINT_OVERRIDDEN_ACPI_TABLE ? 'A' : ' ',
			tainted & TAINT_WARN ? 'W' : ' ');
	}
	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);

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_on_slowpath(const char *file, int line)
{
	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);
	print_modules();
	dump_stack();
	print_oops_end_marker();
	add_taint(TAINT_WARN);
}
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 in: %p\n",
		__builtin_return_address(0));
}
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>
79b4cc5e	23	#include <linux/kallsyms.h>
1da177e4	24
1da177e4 LT	25	int panic_on_oops;
1da177e4 LT	26	int tainted;
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
	37	static int __init panic_setup(char *str)
	38	{
	39	panic_timeout = simple_strtoul(str, NULL, 0);
	40	return 1;
	41	}
	42	__setup("panic=", panic_setup);
	43
	44	static long no_blink(long time)
	45	{
	46	return 0;
	47	}
	48
	49	/* Returns how long it waited in ms */
	50	long (*panic_blink)(long time);
	51	EXPORT_SYMBOL(panic_blink);
	52
	53	/**
	54	* panic - halt the system
	55	* @fmt: The text string to print
	56	*
	57	* Display a message, then perform cleanups.
	58	*
	59	* This function never returns.
	60	*/
c277e63f	61
1da177e4 LT	62	NORET_TYPE void panic(const char * fmt, ...)
	63	{
	64	long i;
	65	static char buf[1024];
	66	va_list args;
347a8dc3	67	#if defined(CONFIG_S390)
c277e63f	68	unsigned long caller = (unsigned long) __builtin_return_address(0);
1da177e4 LT	69	#endif
1da177e4 LT	70
dc009d92 EB	71	/*
	72	* It's possible to come here directly from a panic-assertion and not
	73	* have preempt disabled. Some functions called from here want
	74	* preempt to be disabled. No point enabling it later though...
	75	*/
	76	preempt_disable();
	77
1da177e4 LT	78	bust_spinlocks(1);
	79	va_start(args, fmt);
	80	vsnprintf(buf, sizeof(buf), fmt, args);
	81	va_end(args);
	82	printk(KERN_EMERG "Kernel panic - not syncing: %s\n",buf);
	83	bust_spinlocks(0);
	84
dc009d92 EB	85	/*
	86	* If we have crashed and we have a crash kernel loaded let it handle
	87	* everything else.
	88	* Do we want to call this before we try to display a message?
	89	*/
6e274d14	90	crash_kexec(NULL);
dc009d92	91
1da177e4	92	#ifdef CONFIG_SMP
dc009d92 EB	93	/*
	94	* Note smp_send_stop is the usual smp shutdown function, which
	95	* unfortunately means it may not be hardened to work in a panic
	96	* situation.
	97	*/
1da177e4 LT	98	smp_send_stop();
	99	#endif
	100
e041c683	101	atomic_notifier_call_chain(&panic_notifier_list, 0, buf);
1da177e4 LT	102
	103	if (!panic_blink)
	104	panic_blink = no_blink;
	105
dc009d92	106	if (panic_timeout > 0) {
1da177e4 LT	107	/*
	108	* Delay timeout seconds before rebooting the machine.
	109	* We can't use the "normal" timers since we just panicked..
	110	*/
	111	printk(KERN_EMERG "Rebooting in %d seconds..",panic_timeout);
	112	for (i = 0; i < panic_timeout*1000; ) {
	113	touch_nmi_watchdog();
	114	i += panic_blink(i);
	115	mdelay(1);
	116	i++;
	117	}
2f048ea8 EB	118	/* This will not be a clean reboot, with everything
	119	* shutting down. But if there is a chance of
	120	* rebooting the system it will be rebooted.
1da177e4	121	*/
2f048ea8	122	emergency_restart();
1da177e4 LT	123	}
	124	#ifdef __sparc__
	125	{
	126	extern int stop_a_enabled;
a271c241	127	/* Make sure the user can actually press Stop-A (L1-A) */
1da177e4	128	stop_a_enabled = 1;
a271c241	129	printk(KERN_EMERG "Press Stop-A (L1-A) to return to the boot prom\n");
1da177e4 LT	130	}
1da177e4 LT	131	#endif
347a8dc3	132	#if defined(CONFIG_S390)
c277e63f	133	disabled_wait(caller);
1da177e4 LT	134	#endif
	135	local_irq_enable();
	136	for (i = 0;;) {
c22db941	137	touch_softlockup_watchdog();
1da177e4 LT	138	i += panic_blink(i);
	139	mdelay(1);
	140	i++;
	141	}
	142	}
	143
	144	EXPORT_SYMBOL(panic);
	145
	146	/**
	147	* print_tainted - return a string to represent the kernel taint state.
	148	*
	149	* 'P' - Proprietary module has been loaded.
	150	* 'F' - Module has been forcibly loaded.
	151	* 'S' - SMP with CPUs not designed for SMP.
	152	* 'R' - User forced a module unload.
9aa5e993	153	* 'M' - System experienced a machine check exception.
1da177e4	154	* 'B' - System has hit bad_page.
34f5a398	155	* 'U' - Userspace-defined naughtiness.
95b570c9 NH	156	* 'A' - ACPI table overridden.
95b570c9 NH	157	* 'W' - Taint on warning.
1da177e4 LT	158	*
	159	* The string is overwritten by the next call to print_taint().
	160	*/
c277e63f	161
1da177e4 LT	162	const char *print_tainted(void)
	163	{
	164	static char buf[20];
	165	if (tainted) {
95b570c9	166	snprintf(buf, sizeof(buf), "Tainted: %c%c%c%c%c%c%c%c%c%c",
1da177e4 LT	167	tainted & TAINT_PROPRIETARY_MODULE ? 'P' : 'G',
	168	tainted & TAINT_FORCED_MODULE ? 'F' : ' ',
	169	tainted & TAINT_UNSAFE_SMP ? 'S' : ' ',
	170	tainted & TAINT_FORCED_RMMOD ? 'R' : ' ',
c277e63f	171	tainted & TAINT_MACHINE_CHECK ? 'M' : ' ',
34f5a398	172	tainted & TAINT_BAD_PAGE ? 'B' : ' ',
bcdcd8e7	173	tainted & TAINT_USER ? 'U' : ' ',
6ed31e92	174	tainted & TAINT_DIE ? 'D' : ' ',
95b570c9 NH	175	tainted & TAINT_OVERRIDDEN_ACPI_TABLE ? 'A' : ' ',
95b570c9 NH	176	tainted & TAINT_WARN ? 'W' : ' ');
1da177e4 LT	177	}
	178	else
	179	snprintf(buf, sizeof(buf), "Not tainted");
	180	return(buf);
	181	}
	182
	183	void add_taint(unsigned flag)
	184	{
068c4579	185	debug_locks = 0; /* can't trust the integrity of the kernel anymore */
1da177e4 LT	186	tainted \|= flag;
	187	}
	188	EXPORT_SYMBOL(add_taint);
dd287796 AM	189
	190	static int __init pause_on_oops_setup(char *str)
	191	{
	192	pause_on_oops = simple_strtoul(str, NULL, 0);
	193	return 1;
	194	}
	195	__setup("pause_on_oops=", pause_on_oops_setup);
	196
	197	static void spin_msec(int msecs)
	198	{
	199	int i;
	200
	201	for (i = 0; i < msecs; i++) {
	202	touch_nmi_watchdog();
	203	mdelay(1);
	204	}
	205	}
	206
	207	/*
	208	* It just happens that oops_enter() and oops_exit() are identically
	209	* implemented...
	210	*/
	211	static void do_oops_enter_exit(void)
	212	{
	213	unsigned long flags;
	214	static int spin_counter;
	215
	216	if (!pause_on_oops)
	217	return;
	218
	219	spin_lock_irqsave(&pause_on_oops_lock, flags);
	220	if (pause_on_oops_flag == 0) {
	221	/* This CPU may now print the oops message */
	222	pause_on_oops_flag = 1;
	223	} else {
	224	/* We need to stall this CPU */
	225	if (!spin_counter) {
	226	/* This CPU gets to do the counting */
	227	spin_counter = pause_on_oops;
	228	do {
	229	spin_unlock(&pause_on_oops_lock);
	230	spin_msec(MSEC_PER_SEC);
	231	spin_lock(&pause_on_oops_lock);
	232	} while (--spin_counter);
	233	pause_on_oops_flag = 0;
	234	} else {
	235	/* This CPU waits for a different one */
	236	while (spin_counter) {
	237	spin_unlock(&pause_on_oops_lock);
	238	spin_msec(1);
	239	spin_lock(&pause_on_oops_lock);
	240	}
	241	}
	242	}
	243	spin_unlock_irqrestore(&pause_on_oops_lock, flags);
	244	}
	245
	246	/*
	247	* Return true if the calling CPU is allowed to print oops-related info. This
	248	* is a bit racy..
	249	*/
	250	int oops_may_print(void)
	251	{
	252	return pause_on_oops_flag == 0;
253	}
254
255	/*
256	* Called when the architecture enters its oops handler, before it prints
257	* anything. If this is the first CPU to oops, and it's oopsing the first time
258	* then let it proceed.
259	*
260	* This is all enabled by the pause_on_oops kernel boot option. We do all this
261	* to ensure that oopses don't scroll off the screen. It has the side-effect
262	* of preventing later-oopsing CPUs from mucking up the display, too.
263	*
264	* It turns out that the CPU which is allowed to print ends up pausing for the
265	* right duration, whereas all the other CPUs pause for twice as long: once in
266	* oops_enter(), once in oops_exit().
267	*/
268	void oops_enter(void)
269	{
2c16e9c8	270	debug_locks_off(); /* can't trust the integrity of the kernel anymore */
dd287796 AM	271	do_oops_enter_exit();
	272	}
	273
2c3b20e9 AV	274	/*
	275	* 64-bit random ID for oopses:
	276	*/
	277	static u64 oops_id;
	278
	279	static int init_oops_id(void)
	280	{
	281	if (!oops_id)
	282	get_random_bytes(&oops_id, sizeof(oops_id));
	283
	284	return 0;
	285	}
	286	late_initcall(init_oops_id);
	287
71c33911 AV	288	static void print_oops_end_marker(void)
	289	{
	290	init_oops_id();
	291	printk(KERN_WARNING "---[ end trace %016llx ]---\n",
	292	(unsigned long long)oops_id);
	293	}
	294
dd287796 AM	295	/*
	296	* Called when the architecture exits its oops handler, after printing
	297	* everything.
	298	*/
	299	void oops_exit(void)
	300	{
	301	do_oops_enter_exit();
71c33911	302	print_oops_end_marker();
dd287796	303	}
3162f751	304
79b4cc5e AV	305	#ifdef WANT_WARN_ON_SLOWPATH
	306	void warn_on_slowpath(const char *file, int line)
	307	{
	308	char function[KSYM_SYMBOL_LEN];
	309	unsigned long caller = (unsigned long) __builtin_return_address(0);
79b4cc5e	310	sprint_symbol(function, caller);
71c33911 AV	311
71c33911 AV	312	printk(KERN_WARNING "------------[ cut here ]------------\n");
79b4cc5e AV	313	printk(KERN_WARNING "WARNING: at %s:%d %s()\n", file,
79b4cc5e AV	314	line, function);
71c33911	315	print_modules();
79b4cc5e	316	dump_stack();
71c33911	317	print_oops_end_marker();
95b570c9	318	add_taint(TAINT_WARN);
79b4cc5e AV	319	}
	320	EXPORT_SYMBOL(warn_on_slowpath);
	321	#endif
	322
3162f751 AV	323	#ifdef CONFIG_CC_STACKPROTECTOR
	324	/*
	325	* Called when gcc's -fstack-protector feature is used, and
	326	* gcc detects corruption of the on-stack canary value
	327	*/
	328	void __stack_chk_fail(void)
	329	{
517a92c4 IM	330	panic("stack-protector: Kernel stack is corrupted in: %p\n",
517a92c4 IM	331	__builtin_return_address(0));
3162f751 AV	332	}
	333	EXPORT_SYMBOL(__stack_chk_fail);
	334	#endif