[linux-2.6-block.git] / arch / x86 / kernel / dumpstack_64.c

/*
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *  Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
 */
#include <linux/kallsyms.h>
#include <linux/kprobes.h>
#include <linux/uaccess.h>
#include <linux/hardirq.h>
#include <linux/kdebug.h>
#include <linux/module.h>
#include <linux/ptrace.h>
#include <linux/kexec.h>
#include <linux/sysfs.h>
#include <linux/bug.h>
#include <linux/nmi.h>

#include <asm/stacktrace.h>

#include "dumpstack.h"

#define N_EXCEPTION_STACKS_END \
		(N_EXCEPTION_STACKS + DEBUG_STKSZ/EXCEPTION_STKSZ - 2)

static char x86_stack_ids[][8] = {
		[ DEBUG_STACK-1			]	= "#DB",
		[ NMI_STACK-1			]	= "NMI",
		[ DOUBLEFAULT_STACK-1		]	= "#DF",
		[ STACKFAULT_STACK-1		]	= "#SS",
		[ MCE_STACK-1			]	= "#MC",
#if DEBUG_STKSZ > EXCEPTION_STKSZ
		[ N_EXCEPTION_STACKS ...
		  N_EXCEPTION_STACKS_END	]	= "#DB[?]"
#endif
};

static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
					 unsigned *usedp, char **idp)
{
	unsigned k;

	/*
	 * Iterate over all exception stacks, and figure out whether
	 * 'stack' is in one of them:
	 */
	for (k = 0; k < N_EXCEPTION_STACKS; k++) {
		unsigned long end = per_cpu(orig_ist, cpu).ist[k];
		/*
		 * Is 'stack' above this exception frame's end?
		 * If yes then skip to the next frame.
		 */
		if (stack >= end)
			continue;
		/*
		 * Is 'stack' above this exception frame's start address?
		 * If yes then we found the right frame.
		 */
		if (stack >= end - EXCEPTION_STKSZ) {
			/*
			 * Make sure we only iterate through an exception
			 * stack once. If it comes up for the second time
			 * then there's something wrong going on - just
			 * break out and return NULL:
			 */
			if (*usedp & (1U << k))
				break;
			*usedp |= 1U << k;
			*idp = x86_stack_ids[k];
			return (unsigned long *)end;
		}
		/*
		 * If this is a debug stack, and if it has a larger size than
		 * the usual exception stacks, then 'stack' might still
		 * be within the lower portion of the debug stack:
		 */
#if DEBUG_STKSZ > EXCEPTION_STKSZ
		if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
			unsigned j = N_EXCEPTION_STACKS - 1;

			/*
			 * Black magic. A large debug stack is composed of
			 * multiple exception stack entries, which we
			 * iterate through now. Dont look:
			 */
			do {
				++j;
				end -= EXCEPTION_STKSZ;
				x86_stack_ids[j][4] = '1' +
						(j - N_EXCEPTION_STACKS);
			} while (stack < end - EXCEPTION_STKSZ);
			if (*usedp & (1U << j))
				break;
			*usedp |= 1U << j;
			*idp = x86_stack_ids[j];
			return (unsigned long *)end;
		}
#endif
	}
	return NULL;
}

static inline int
in_irq_stack(unsigned long *stack, unsigned long *irq_stack,
	     unsigned long *irq_stack_end)
{
	return (stack >= irq_stack && stack < irq_stack_end);
}

/*
 * We are returning from the irq stack and go to the previous one.
 * If the previous stack is also in the irq stack, then bp in the first
 * frame of the irq stack points to the previous, interrupted one.
 * Otherwise we have another level of indirection: We first save
 * the bp of the previous stack, then we switch the stack to the irq one
 * and save a new bp that links to the previous one.
 * (See save_args())
 */
static inline unsigned long
fixup_bp_irq_link(unsigned long bp, unsigned long *stack,
		  unsigned long *irq_stack, unsigned long *irq_stack_end)
{
#ifdef CONFIG_FRAME_POINTER
	struct stack_frame *frame = (struct stack_frame *)bp;

	if (!in_irq_stack(stack, irq_stack, irq_stack_end))
		return (unsigned long)frame->next_frame;
#endif
	return bp;
}

/*
 * x86-64 can have up to three kernel stacks:
 * process stack
 * interrupt stack
 * severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
 */

void dump_trace(struct task_struct *task, struct pt_regs *regs,
		unsigned long *stack, unsigned long bp,
		const struct stacktrace_ops *ops, void *data)
{
	const unsigned cpu = get_cpu();
	unsigned long *irq_stack_end =
		(unsigned long *)per_cpu(irq_stack_ptr, cpu);
	unsigned used = 0;
	struct thread_info *tinfo;
	int graph = 0;

	if (!task)
		task = current;

	if (!stack) {
		unsigned long dummy;
		stack = &dummy;
		if (task && task != current)
			stack = (unsigned long *)task->thread.sp;
	}

#ifdef CONFIG_FRAME_POINTER
	if (!bp) {
		if (task == current) {
			/* Grab bp right from our regs */
			get_bp(bp);
		} else {
			/* bp is the last reg pushed by switch_to */
			bp = *(unsigned long *) task->thread.sp;
		}
	}
#endif

	/*
	 * Print function call entries in all stacks, starting at the
	 * current stack address. If the stacks consist of nested
	 * exceptions
	 */
	tinfo = task_thread_info(task);
	for (;;) {
		char *id;
		unsigned long *estack_end;
		estack_end = in_exception_stack(cpu, (unsigned long)stack,
						&used, &id);

		if (estack_end) {
			if (ops->stack(data, id) < 0)
				break;

			bp = ops->walk_stack(tinfo, stack, bp, ops,
					     data, estack_end, &graph);
			ops->stack(data, "<EOE>");
			/*
			 * We link to the next stack via the
			 * second-to-last pointer (index -2 to end) in the
			 * exception stack:
			 */
			stack = (unsigned long *) estack_end[-2];
			continue;
		}
		if (irq_stack_end) {
			unsigned long *irq_stack;
			irq_stack = irq_stack_end -
				(IRQ_STACK_SIZE - 64) / sizeof(*irq_stack);

			if (in_irq_stack(stack, irq_stack, irq_stack_end)) {
				if (ops->stack(data, "IRQ") < 0)
					break;
				bp = print_context_stack(tinfo, stack, bp,
					ops, data, irq_stack_end, &graph);
				/*
				 * We link to the next stack (which would be
				 * the process stack normally) the last
				 * pointer (index -1 to end) in the IRQ stack:
				 */
				stack = (unsigned long *) (irq_stack_end[-1]);
				bp = fixup_bp_irq_link(bp, stack, irq_stack,
						       irq_stack_end);
				irq_stack_end = NULL;
				ops->stack(data, "EOI");
				continue;
			}
		}
		break;
	}

	/*
	 * This handles the process stack:
	 */
	bp = print_context_stack(tinfo, stack, bp, ops, data, NULL, &graph);
	put_cpu();
}
EXPORT_SYMBOL(dump_trace);

void
show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs,
		   unsigned long *sp, unsigned long bp, char *log_lvl)
{
	unsigned long *irq_stack_end;
	unsigned long *irq_stack;
	unsigned long *stack;
	int cpu;
	int i;

	preempt_disable();
	cpu = smp_processor_id();

	irq_stack_end	= (unsigned long *)(per_cpu(irq_stack_ptr, cpu));
	irq_stack	= (unsigned long *)(per_cpu(irq_stack_ptr, cpu) - IRQ_STACK_SIZE);

	/*
	 * Debugging aid: "show_stack(NULL, NULL);" prints the
	 * back trace for this cpu:
	 */
	if (sp == NULL) {
		if (task)
			sp = (unsigned long *)task->thread.sp;
		else
			sp = (unsigned long *)&sp;
	}

	stack = sp;
	for (i = 0; i < kstack_depth_to_print; i++) {
		if (stack >= irq_stack && stack <= irq_stack_end) {
			if (stack == irq_stack_end) {
				stack = (unsigned long *) (irq_stack_end[-1]);
				printk(" <EOI> ");
			}
		} else {
		if (((long) stack & (THREAD_SIZE-1)) == 0)
			break;
		}
		if (i && ((i % STACKSLOTS_PER_LINE) == 0))
			printk("\n%s", log_lvl);
		printk(" %016lx", *stack++);
		touch_nmi_watchdog();
	}
	preempt_enable();

	printk("\n");
	show_trace_log_lvl(task, regs, sp, bp, log_lvl);
}

void show_registers(struct pt_regs *regs)
{
	int i;
	unsigned long sp;
	const int cpu = smp_processor_id();
	struct task_struct *cur = current;

	sp = regs->sp;
	printk("CPU %d ", cpu);
	print_modules();
	__show_regs(regs, 1);
	printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
		cur->comm, cur->pid, task_thread_info(cur), cur);

	/*
	 * When in-kernel, we also print out the stack and code at the
	 * time of the fault..
	 */
	if (!user_mode(regs)) {
		unsigned int code_prologue = code_bytes * 43 / 64;
		unsigned int code_len = code_bytes;
		unsigned char c;
		u8 *ip;

		printk(KERN_EMERG "Stack:\n");
		show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
				regs->bp, KERN_EMERG);

		printk(KERN_EMERG "Code: ");

		ip = (u8 *)regs->ip - code_prologue;
		if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) {
			/* try starting at IP */
			ip = (u8 *)regs->ip;
			code_len = code_len - code_prologue + 1;
		}
		for (i = 0; i < code_len; i++, ip++) {
			if (ip < (u8 *)PAGE_OFFSET ||
					probe_kernel_address(ip, c)) {
				printk(" Bad RIP value.");
				break;
			}
			if (ip == (u8 *)regs->ip)
				printk("<%02x> ", c);
			else
				printk("%02x ", c);
		}
	}
	printk("\n");
}

int is_valid_bugaddr(unsigned long ip)
{
	unsigned short ud2;

	if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
		return 0;

	return ud2 == 0x0b0f;
}
Commit	Line	Data
6fcbede3 AH	1	/*
	2	* Copyright (C) 1991, 1992 Linus Torvalds
	3	* Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
	4	*/
	5	#include <linux/kallsyms.h>
	6	#include <linux/kprobes.h>
	7	#include <linux/uaccess.h>
6fcbede3 AH	8	#include <linux/hardirq.h>
	9	#include <linux/kdebug.h>
	10	#include <linux/module.h>
	11	#include <linux/ptrace.h>
	12	#include <linux/kexec.h>
b8030906	13	#include <linux/sysfs.h>
6fcbede3 AH	14	#include <linux/bug.h>
	15	#include <linux/nmi.h>
	16
	17	#include <asm/stacktrace.h>
	18
878719e8	19	#include "dumpstack.h"
6fcbede3	20
b8030906 IM	21	#define N_EXCEPTION_STACKS_END \
b8030906 IM	22	(N_EXCEPTION_STACKS + DEBUG_STKSZ/EXCEPTION_STKSZ - 2)
0406ca6d FW	23
0406ca6d FW	24	static char x86_stack_ids[][8] = {
b8030906 IM	25	[ DEBUG_STACK-1 ] = "#DB",
	26	[ NMI_STACK-1 ] = "NMI",
	27	[ DOUBLEFAULT_STACK-1 ] = "#DF",
	28	[ STACKFAULT_STACK-1 ] = "#SS",
	29	[ MCE_STACK-1 ] = "#MC",
6fcbede3	30	#if DEBUG_STKSZ > EXCEPTION_STKSZ
b8030906 IM	31	[ N_EXCEPTION_STACKS ...
b8030906 IM	32	N_EXCEPTION_STACKS_END ] = "#DB[?]"
6fcbede3	33	#endif
b8030906	34	};
0406ca6d	35
0406ca6d	36	static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
b8030906	37	unsigned usedp, char *idp)
0406ca6d	38	{
6fcbede3 AH	39	unsigned k;
	40
	41	/*
	42	* Iterate over all exception stacks, and figure out whether
	43	* 'stack' is in one of them:
	44	*/
	45	for (k = 0; k < N_EXCEPTION_STACKS; k++) {
	46	unsigned long end = per_cpu(orig_ist, cpu).ist[k];
	47	/*
	48	* Is 'stack' above this exception frame's end?
	49	* If yes then skip to the next frame.
	50	*/
	51	if (stack >= end)
	52	continue;
	53	/*
	54	* Is 'stack' above this exception frame's start address?
	55	* If yes then we found the right frame.
	56	*/
	57	if (stack >= end - EXCEPTION_STKSZ) {
	58	/*
	59	* Make sure we only iterate through an exception
	60	* stack once. If it comes up for the second time
	61	* then there's something wrong going on - just
	62	* break out and return NULL:
	63	*/
	64	if (*usedp & (1U << k))
	65	break;
	66	*usedp \|= 1U << k;
0406ca6d	67	*idp = x86_stack_ids[k];
6fcbede3 AH	68	return (unsigned long *)end;
	69	}
	70	/*
	71	* If this is a debug stack, and if it has a larger size than
	72	* the usual exception stacks, then 'stack' might still
	73	* be within the lower portion of the debug stack:
	74	*/
	75	#if DEBUG_STKSZ > EXCEPTION_STKSZ
	76	if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
	77	unsigned j = N_EXCEPTION_STACKS - 1;
	78
	79	/*
	80	* Black magic. A large debug stack is composed of
	81	* multiple exception stack entries, which we
	82	* iterate through now. Dont look:
	83	*/
	84	do {
	85	++j;
	86	end -= EXCEPTION_STKSZ;
0406ca6d FW	87	x86_stack_ids[j][4] = '1' +
0406ca6d FW	88	(j - N_EXCEPTION_STACKS);
6fcbede3 AH	89	} while (stack < end - EXCEPTION_STKSZ);
	90	if (*usedp & (1U << j))
	91	break;
	92	*usedp \|= 1U << j;
0406ca6d	93	*idp = x86_stack_ids[j];
6fcbede3 AH	94	return (unsigned long *)end;
	95	}
	96	#endif
	97	}
	98	return NULL;
	99	}
	100
af2d8289 FW	101	static inline int
	102	in_irq_stack(unsigned long stack, unsigned long irq_stack,
	103	unsigned long *irq_stack_end)
	104	{
	105	return (stack >= irq_stack && stack < irq_stack_end);
	106	}
	107
	108	/*
	109	* We are returning from the irq stack and go to the previous one.
	110	* If the previous stack is also in the irq stack, then bp in the first
	111	* frame of the irq stack points to the previous, interrupted one.
	112	* Otherwise we have another level of indirection: We first save
	113	* the bp of the previous stack, then we switch the stack to the irq one
	114	* and save a new bp that links to the previous one.
	115	* (See save_args())
	116	*/
	117	static inline unsigned long
	118	fixup_bp_irq_link(unsigned long bp, unsigned long *stack,
	119	unsigned long irq_stack, unsigned long irq_stack_end)
	120	{
	121	#ifdef CONFIG_FRAME_POINTER
	122	struct stack_frame frame = (struct stack_frame )bp;
	123
	124	if (!in_irq_stack(stack, irq_stack, irq_stack_end))
	125	return (unsigned long)frame->next_frame;
	126	#endif
	127	return bp;
	128	}
	129
6fcbede3 AH	130	/*
	131	* x86-64 can have up to three kernel stacks:
	132	* process stack
	133	* interrupt stack
	134	* severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
	135	*/
	136
6fcbede3 AH	137	void dump_trace(struct task_struct task, struct pt_regs regs,
	138	unsigned long *stack, unsigned long bp,
	139	const struct stacktrace_ops ops, void data)
	140	{
	141	const unsigned cpu = get_cpu();
26f80bd6 BG	142	unsigned long *irq_stack_end =
26f80bd6 BG	143	(unsigned long *)per_cpu(irq_stack_ptr, cpu);
6fcbede3 AH	144	unsigned used = 0;
6fcbede3 AH	145	struct thread_info *tinfo;
7ee991fb	146	int graph = 0;
6fcbede3 AH	147
	148	if (!task)
	149	task = current;
	150
	151	if (!stack) {
	152	unsigned long dummy;
	153	stack = &dummy;
	154	if (task && task != current)
	155	stack = (unsigned long *)task->thread.sp;
	156	}
	157
	158	#ifdef CONFIG_FRAME_POINTER
	159	if (!bp) {
	160	if (task == current) {
	161	/* Grab bp right from our regs */
8a541665	162	get_bp(bp);
6fcbede3 AH	163	} else {
	164	/* bp is the last reg pushed by switch_to */
	165	bp = (unsigned long ) task->thread.sp;
	166	}
	167	}
	168	#endif
	169
	170	/*
	171	* Print function call entries in all stacks, starting at the
	172	* current stack address. If the stacks consist of nested
	173	* exceptions
	174	*/
	175	tinfo = task_thread_info(task);
	176	for (;;) {
	177	char *id;
	178	unsigned long *estack_end;
	179	estack_end = in_exception_stack(cpu, (unsigned long)stack,
	180	&used, &id);
	181
	182	if (estack_end) {
	183	if (ops->stack(data, id) < 0)
	184	break;
	185
61c1917f FW	186	bp = ops->walk_stack(tinfo, stack, bp, ops,
61c1917f FW	187	data, estack_end, &graph);
6fcbede3 AH	188	ops->stack(data, "<EOE>");
	189	/*
	190	* We link to the next stack via the
	191	* second-to-last pointer (index -2 to end) in the
	192	* exception stack:
	193	*/
	194	stack = (unsigned long *) estack_end[-2];
	195	continue;
	196	}
26f80bd6 BG	197	if (irq_stack_end) {
	198	unsigned long *irq_stack;
	199	irq_stack = irq_stack_end -
	200	(IRQ_STACK_SIZE - 64) / sizeof(*irq_stack);
6fcbede3	201
af2d8289	202	if (in_irq_stack(stack, irq_stack, irq_stack_end)) {
6fcbede3 AH	203	if (ops->stack(data, "IRQ") < 0)
	204	break;
	205	bp = print_context_stack(tinfo, stack, bp,
26f80bd6	206	ops, data, irq_stack_end, &graph);
6fcbede3 AH	207	/*
	208	* We link to the next stack (which would be
	209	* the process stack normally) the last
	210	* pointer (index -1 to end) in the IRQ stack:
	211	*/
26f80bd6	212	stack = (unsigned long *) (irq_stack_end[-1]);
af2d8289 FW	213	bp = fixup_bp_irq_link(bp, stack, irq_stack,
af2d8289 FW	214	irq_stack_end);
26f80bd6	215	irq_stack_end = NULL;
6fcbede3 AH	216	ops->stack(data, "EOI");
	217	continue;
	218	}
	219	}
	220	break;
	221	}
	222
	223	/*
	224	* This handles the process stack:
	225	*/
7ee991fb	226	bp = print_context_stack(tinfo, stack, bp, ops, data, NULL, &graph);
6fcbede3 AH	227	put_cpu();
	228	}
	229	EXPORT_SYMBOL(dump_trace);
	230
878719e8	231	void
6fcbede3	232	show_stack_log_lvl(struct task_struct task, struct pt_regs regs,
b8030906	233	unsigned long sp, unsigned long bp, char log_lvl)
6fcbede3	234	{
67f2de0b IM	235	unsigned long *irq_stack_end;
67f2de0b IM	236	unsigned long *irq_stack;
6fcbede3	237	unsigned long *stack;
67f2de0b	238	int cpu;
6fcbede3	239	int i;
67f2de0b IM	240
	241	preempt_disable();
	242	cpu = smp_processor_id();
	243
	244	irq_stack_end = (unsigned long *)(per_cpu(irq_stack_ptr, cpu));
	245	irq_stack = (unsigned long *)(per_cpu(irq_stack_ptr, cpu) - IRQ_STACK_SIZE);
6fcbede3 AH	246
6fcbede3 AH	247	/*
67f2de0b IM	248	* Debugging aid: "show_stack(NULL, NULL);" prints the
67f2de0b IM	249	* back trace for this cpu:
6fcbede3	250	*/
6fcbede3 AH	251	if (sp == NULL) {
	252	if (task)
	253	sp = (unsigned long *)task->thread.sp;
	254	else
	255	sp = (unsigned long *)&sp;
	256	}
	257
	258	stack = sp;
	259	for (i = 0; i < kstack_depth_to_print; i++) {
26f80bd6 BG	260	if (stack >= irq_stack && stack <= irq_stack_end) {
	261	if (stack == irq_stack_end) {
	262	stack = (unsigned long *) (irq_stack_end[-1]);
6fcbede3 AH	263	printk(" <EOI> ");
	264	}
	265	} else {
	266	if (((long) stack & (THREAD_SIZE-1)) == 0)
	267	break;
	268	}
8a541665	269	if (i && ((i % STACKSLOTS_PER_LINE) == 0))
ca0a8164	270	printk("\n%s", log_lvl);
6fcbede3 AH	271	printk(" %016lx", *stack++);
	272	touch_nmi_watchdog();
	273	}
67f2de0b IM	274	preempt_enable();
67f2de0b IM	275
6fcbede3 AH	276	printk("\n");
	277	show_trace_log_lvl(task, regs, sp, bp, log_lvl);
	278	}
	279
6fcbede3 AH	280	void show_registers(struct pt_regs *regs)
	281	{
	282	int i;
	283	unsigned long sp;
	284	const int cpu = smp_processor_id();
c6f5e0ac	285	struct task_struct *cur = current;
6fcbede3 AH	286
	287	sp = regs->sp;
	288	printk("CPU %d ", cpu);
f266d7f5	289	print_modules();
6fcbede3 AH	290	__show_regs(regs, 1);
	291	printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
	292	cur->comm, cur->pid, task_thread_info(cur), cur);
	293
	294	/*
	295	* When in-kernel, we also print out the stack and code at the
	296	* time of the fault..
	297	*/
	298	if (!user_mode(regs)) {
	299	unsigned int code_prologue = code_bytes * 43 / 64;
	300	unsigned int code_len = code_bytes;
	301	unsigned char c;
	302	u8 *ip;
	303
ca0a8164	304	printk(KERN_EMERG "Stack:\n");
6fcbede3	305	show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
ca0a8164	306	regs->bp, KERN_EMERG);
6fcbede3 AH	307
	308	printk(KERN_EMERG "Code: ");
	309
	310	ip = (u8 *)regs->ip - code_prologue;
	311	if (ip < (u8 *)PAGE_OFFSET \|\| probe_kernel_address(ip, c)) {
8a541665	312	/* try starting at IP */
6fcbede3 AH	313	ip = (u8 *)regs->ip;
	314	code_len = code_len - code_prologue + 1;
	315	}
	316	for (i = 0; i < code_len; i++, ip++) {
	317	if (ip < (u8 *)PAGE_OFFSET \|\|
	318	probe_kernel_address(ip, c)) {
	319	printk(" Bad RIP value.");
	320	break;
	321	}
	322	if (ip == (u8 *)regs->ip)
	323	printk("<%02x> ", c);
	324	else
	325	printk("%02x ", c);
	326	}
	327	}
	328	printk("\n");
	329	}
	330
	331	int is_valid_bugaddr(unsigned long ip)
	332	{
	333	unsigned short ud2;
	334
	335	if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
	336	return 0;
	337
	338	return ud2 == 0x0b0f;
	339	}