[deliverable/linux.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>


#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;
	unsigned long next;

	if (!in_irq_stack(stack, irq_stack, irq_stack_end)) {
		if (!probe_kernel_address(&frame->next_frame, next))
			return next;
		else
			WARN_ONCE(1, "Perf: bad frame pointer = %p in "
				  "callchain\n", &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,
		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;
	unsigned long bp;

	if (!task)
		task = current;

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

	bp = stack_frame(task, regs);
	/*
	 * 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 = ops->walk_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 = ops->walk_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, 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(KERN_CONT " <EOI> ");
			}
		} else {
		if (((long) stack & (THREAD_SIZE-1)) == 0)
			break;
		}
		if (i && ((i % STACKSLOTS_PER_LINE) == 0))
			printk(KERN_CONT "\n");
		printk(KERN_CONT " %016lx", *stack++);
		touch_nmi_watchdog();
	}
	preempt_enable();

	printk(KERN_CONT "\n");
	show_trace_log_lvl(task, regs, sp, 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,
				   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
6fcbede3	19
b8030906 IM	20	#define N_EXCEPTION_STACKS_END \
b8030906 IM	21	(N_EXCEPTION_STACKS + DEBUG_STKSZ/EXCEPTION_STKSZ - 2)
0406ca6d FW	22
0406ca6d FW	23	static char x86_stack_ids[][8] = {
b8030906 IM	24	[ DEBUG_STACK-1 ] = "#DB",
	25	[ NMI_STACK-1 ] = "NMI",
	26	[ DOUBLEFAULT_STACK-1 ] = "#DF",
	27	[ STACKFAULT_STACK-1 ] = "#SS",
	28	[ MCE_STACK-1 ] = "#MC",
6fcbede3	29	#if DEBUG_STKSZ > EXCEPTION_STKSZ
b8030906 IM	30	[ N_EXCEPTION_STACKS ...
b8030906 IM	31	N_EXCEPTION_STACKS_END ] = "#DB[?]"
6fcbede3	32	#endif
b8030906	33	};
0406ca6d	34
0406ca6d	35	static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
b8030906	36	unsigned usedp, char *idp)
0406ca6d	37	{
6fcbede3 AH	38	unsigned k;
	39
	40	/*
	41	* Iterate over all exception stacks, and figure out whether
	42	* 'stack' is in one of them:
	43	*/
	44	for (k = 0; k < N_EXCEPTION_STACKS; k++) {
	45	unsigned long end = per_cpu(orig_ist, cpu).ist[k];
	46	/*
	47	* Is 'stack' above this exception frame's end?
	48	* If yes then skip to the next frame.
	49	*/
	50	if (stack >= end)
	51	continue;
	52	/*
	53	* Is 'stack' above this exception frame's start address?
	54	* If yes then we found the right frame.
	55	*/
	56	if (stack >= end - EXCEPTION_STKSZ) {
	57	/*
	58	* Make sure we only iterate through an exception
	59	* stack once. If it comes up for the second time
	60	* then there's something wrong going on - just
	61	* break out and return NULL:
	62	*/
	63	if (*usedp & (1U << k))
	64	break;
	65	*usedp \|= 1U << k;
0406ca6d	66	*idp = x86_stack_ids[k];
6fcbede3 AH	67	return (unsigned long *)end;
	68	}
	69	/*
	70	* If this is a debug stack, and if it has a larger size than
	71	* the usual exception stacks, then 'stack' might still
	72	* be within the lower portion of the debug stack:
	73	*/
	74	#if DEBUG_STKSZ > EXCEPTION_STKSZ
	75	if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
	76	unsigned j = N_EXCEPTION_STACKS - 1;
	77
	78	/*
	79	* Black magic. A large debug stack is composed of
	80	* multiple exception stack entries, which we
	81	* iterate through now. Dont look:
	82	*/
	83	do {
	84	++j;
	85	end -= EXCEPTION_STKSZ;
0406ca6d FW	86	x86_stack_ids[j][4] = '1' +
0406ca6d FW	87	(j - N_EXCEPTION_STACKS);
6fcbede3 AH	88	} while (stack < end - EXCEPTION_STKSZ);
	89	if (*usedp & (1U << j))
	90	break;
	91	*usedp \|= 1U << j;
0406ca6d	92	*idp = x86_stack_ids[j];
6fcbede3 AH	93	return (unsigned long *)end;
	94	}
	95	#endif
	96	}
	97	return NULL;
	98	}
	99
af2d8289 FW	100	static inline int
	101	in_irq_stack(unsigned long stack, unsigned long irq_stack,
	102	unsigned long *irq_stack_end)
	103	{
	104	return (stack >= irq_stack && stack < irq_stack_end);
	105	}
	106
	107	/*
	108	* We are returning from the irq stack and go to the previous one.
	109	* If the previous stack is also in the irq stack, then bp in the first
	110	* frame of the irq stack points to the previous, interrupted one.
	111	* Otherwise we have another level of indirection: We first save
	112	* the bp of the previous stack, then we switch the stack to the irq one
	113	* and save a new bp that links to the previous one.
	114	* (See save_args())
	115	*/
	116	static inline unsigned long
	117	fixup_bp_irq_link(unsigned long bp, unsigned long *stack,
	118	unsigned long irq_stack, unsigned long irq_stack_end)
	119	{
	120	#ifdef CONFIG_FRAME_POINTER
	121	struct stack_frame frame = (struct stack_frame )bp;
29044ad1	122	unsigned long next;
af2d8289	123
29044ad1 FW	124	if (!in_irq_stack(stack, irq_stack, irq_stack_end)) {
	125	if (!probe_kernel_address(&frame->next_frame, next))
	126	return next;
	127	else
	128	WARN_ONCE(1, "Perf: bad frame pointer = %p in "
	129	"callchain\n", &frame->next_frame);
	130	}
af2d8289 FW	131	#endif
	132	return bp;
	133	}
	134
6fcbede3 AH	135	/*
	136	* x86-64 can have up to three kernel stacks:
	137	* process stack
	138	* interrupt stack
	139	* severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
	140	*/
	141
9c0729dc SSP	142	void dump_trace(struct task_struct *task,
9c0729dc SSP	143	struct pt_regs regs, unsigned long stack,
6fcbede3 AH	144	const struct stacktrace_ops ops, void data)
	145	{
	146	const unsigned cpu = get_cpu();
26f80bd6 BG	147	unsigned long *irq_stack_end =
26f80bd6 BG	148	(unsigned long *)per_cpu(irq_stack_ptr, cpu);
6fcbede3 AH	149	unsigned used = 0;
6fcbede3 AH	150	struct thread_info *tinfo;
7ee991fb	151	int graph = 0;
9c0729dc	152	unsigned long bp;
6fcbede3 AH	153
	154	if (!task)
	155	task = current;
	156
	157	if (!stack) {
	158	unsigned long dummy;
	159	stack = &dummy;
	160	if (task && task != current)
	161	stack = (unsigned long *)task->thread.sp;
	162	}
	163
9c0729dc	164	bp = stack_frame(task, regs);
6fcbede3 AH	165	/*
	166	* Print function call entries in all stacks, starting at the
	167	* current stack address. If the stacks consist of nested
	168	* exceptions
	169	*/
	170	tinfo = task_thread_info(task);
	171	for (;;) {
	172	char *id;
	173	unsigned long *estack_end;
	174	estack_end = in_exception_stack(cpu, (unsigned long)stack,
	175	&used, &id);
	176
	177	if (estack_end) {
	178	if (ops->stack(data, id) < 0)
	179	break;
	180
61c1917f FW	181	bp = ops->walk_stack(tinfo, stack, bp, ops,
61c1917f FW	182	data, estack_end, &graph);
6fcbede3 AH	183	ops->stack(data, "<EOE>");
	184	/*
	185	* We link to the next stack via the
	186	* second-to-last pointer (index -2 to end) in the
	187	* exception stack:
	188	*/
	189	stack = (unsigned long *) estack_end[-2];
	190	continue;
	191	}
26f80bd6 BG	192	if (irq_stack_end) {
	193	unsigned long *irq_stack;
	194	irq_stack = irq_stack_end -
	195	(IRQ_STACK_SIZE - 64) / sizeof(*irq_stack);
6fcbede3	196
af2d8289	197	if (in_irq_stack(stack, irq_stack, irq_stack_end)) {
6fcbede3 AH	198	if (ops->stack(data, "IRQ") < 0)
6fcbede3 AH	199	break;
61e67fb9	200	bp = ops->walk_stack(tinfo, stack, bp,
26f80bd6	201	ops, data, irq_stack_end, &graph);
6fcbede3 AH	202	/*
	203	* We link to the next stack (which would be
	204	* the process stack normally) the last
	205	* pointer (index -1 to end) in the IRQ stack:
	206	*/
26f80bd6	207	stack = (unsigned long *) (irq_stack_end[-1]);
af2d8289 FW	208	bp = fixup_bp_irq_link(bp, stack, irq_stack,
af2d8289 FW	209	irq_stack_end);
26f80bd6	210	irq_stack_end = NULL;
6fcbede3 AH	211	ops->stack(data, "EOI");
	212	continue;
	213	}
	214	}
	215	break;
	216	}
	217
	218	/*
	219	* This handles the process stack:
	220	*/
61e67fb9	221	bp = ops->walk_stack(tinfo, stack, bp, ops, data, NULL, &graph);
6fcbede3 AH	222	put_cpu();
	223	}
	224	EXPORT_SYMBOL(dump_trace);
	225
878719e8	226	void
6fcbede3	227	show_stack_log_lvl(struct task_struct task, struct pt_regs regs,
9c0729dc	228	unsigned long sp, char log_lvl)
6fcbede3	229	{
67f2de0b IM	230	unsigned long *irq_stack_end;
67f2de0b IM	231	unsigned long *irq_stack;
6fcbede3	232	unsigned long *stack;
67f2de0b	233	int cpu;
6fcbede3	234	int i;
67f2de0b IM	235
	236	preempt_disable();
	237	cpu = smp_processor_id();
	238
	239	irq_stack_end = (unsigned long *)(per_cpu(irq_stack_ptr, cpu));
	240	irq_stack = (unsigned long *)(per_cpu(irq_stack_ptr, cpu) - IRQ_STACK_SIZE);
6fcbede3 AH	241
6fcbede3 AH	242	/*
67f2de0b IM	243	* Debugging aid: "show_stack(NULL, NULL);" prints the
67f2de0b IM	244	* back trace for this cpu:
6fcbede3	245	*/
6fcbede3 AH	246	if (sp == NULL) {
	247	if (task)
	248	sp = (unsigned long *)task->thread.sp;
	249	else
	250	sp = (unsigned long *)&sp;
	251	}
	252
	253	stack = sp;
	254	for (i = 0; i < kstack_depth_to_print; i++) {
26f80bd6 BG	255	if (stack >= irq_stack && stack <= irq_stack_end) {
	256	if (stack == irq_stack_end) {
	257	stack = (unsigned long *) (irq_stack_end[-1]);
e4072a9a	258	printk(KERN_CONT " <EOI> ");
6fcbede3 AH	259	}
	260	} else {
	261	if (((long) stack & (THREAD_SIZE-1)) == 0)
	262	break;
	263	}
8a541665	264	if (i && ((i % STACKSLOTS_PER_LINE) == 0))
e4072a9a JS	265	printk(KERN_CONT "\n");
e4072a9a JS	266	printk(KERN_CONT " %016lx", *stack++);
6fcbede3 AH	267	touch_nmi_watchdog();
6fcbede3 AH	268	}
67f2de0b IM	269	preempt_enable();
67f2de0b IM	270
e4072a9a	271	printk(KERN_CONT "\n");
9c0729dc	272	show_trace_log_lvl(task, regs, sp, log_lvl);
6fcbede3 AH	273	}
6fcbede3 AH	274
6fcbede3 AH	275	void show_registers(struct pt_regs *regs)
	276	{
	277	int i;
	278	unsigned long sp;
	279	const int cpu = smp_processor_id();
c6f5e0ac	280	struct task_struct *cur = current;
6fcbede3 AH	281
	282	sp = regs->sp;
	283	printk("CPU %d ", cpu);
f266d7f5	284	print_modules();
6fcbede3 AH	285	__show_regs(regs, 1);
	286	printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
	287	cur->comm, cur->pid, task_thread_info(cur), cur);
	288
	289	/*
	290	* When in-kernel, we also print out the stack and code at the
	291	* time of the fault..
	292	*/
	293	if (!user_mode(regs)) {
	294	unsigned int code_prologue = code_bytes * 43 / 64;
	295	unsigned int code_len = code_bytes;
	296	unsigned char c;
	297	u8 *ip;
	298
ca0a8164	299	printk(KERN_EMERG "Stack:\n");
6fcbede3	300	show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
9c0729dc	301	KERN_EMERG);
6fcbede3 AH	302
	303	printk(KERN_EMERG "Code: ");
	304
	305	ip = (u8 *)regs->ip - code_prologue;
	306	if (ip < (u8 *)PAGE_OFFSET \|\| probe_kernel_address(ip, c)) {
8a541665	307	/* try starting at IP */
6fcbede3 AH	308	ip = (u8 *)regs->ip;
	309	code_len = code_len - code_prologue + 1;
	310	}
	311	for (i = 0; i < code_len; i++, ip++) {
	312	if (ip < (u8 *)PAGE_OFFSET \|\|
	313	probe_kernel_address(ip, c)) {
	314	printk(" Bad RIP value.");
	315	break;
	316	}
	317	if (ip == (u8 *)regs->ip)
	318	printk("<%02x> ", c);
	319	else
	320	printk("%02x ", c);
	321	}
	322	}
	323	printk("\n");
	324	}
	325
	326	int is_valid_bugaddr(unsigned long ip)
	327	{
	328	unsigned short ud2;
	329
	330	if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
	331	return 0;
	332
	333	return ud2 == 0x0b0f;
	334	}