[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/utsname.h>
#include <linux/hardirq.h>
#include <linux/kdebug.h>
#include <linux/module.h>
#include <linux/ptrace.h>
#include <linux/kexec.h>
#include <linux/bug.h>
#include <linux/nmi.h>
#include <linux/sysfs.h>

#include <asm/stacktrace.h>

#include "dumpstack.h"

static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
					unsigned *usedp, char **idp)
{
	static char 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 + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
#endif
	};
	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 = 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;
				ids[j][4] = '1' + (j - N_EXCEPTION_STACKS);
			} while (stack < end - EXCEPTION_STKSZ);
			if (*usedp & (1U << j))
				break;
			*usedp |= 1U << j;
			*idp = ids[j];
			return (unsigned long *)end;
		}
#endif
	}
	return NULL;
}

/*
 * 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 *irqstack_end = (unsigned long *)cpu_pda(cpu)->irqstackptr;
	unsigned used = 0;
	struct thread_info *tinfo;

	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 = print_context_stack(tinfo, stack, bp, ops,
							data, estack_end);
			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 (irqstack_end) {
			unsigned long *irqstack;
			irqstack = irqstack_end -
				(IRQSTACKSIZE - 64) / sizeof(*irqstack);

			if (stack >= irqstack && stack < irqstack_end) {
				if (ops->stack(data, "IRQ") < 0)
					break;
				bp = print_context_stack(tinfo, stack, bp,
						ops, data, irqstack_end);
				/*
				 * 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 *) (irqstack_end[-1]);
				irqstack_end = NULL;
				ops->stack(data, "EOI");
				continue;
			}
		}
		break;
	}

	/*
	 * This handles the process stack:
	 */
	bp = print_context_stack(tinfo, stack, bp, ops, data, NULL);
	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 *stack;
	int i;
	const int cpu = smp_processor_id();
	unsigned long *irqstack_end =
		(unsigned long *) (cpu_pda(cpu)->irqstackptr);
	unsigned long *irqstack =
		(unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);

	/*
	 * 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 >= irqstack && stack <= irqstack_end) {
			if (stack == irqstack_end) {
				stack = (unsigned long *) (irqstack_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();
	}
	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 = cpu_pda(cpu)->pcurrent;

	sp = regs->sp;
	printk("CPU %d ", cpu);
	__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>
	8	#include <linux/utsname.h>
	9	#include <linux/hardirq.h>
	10	#include <linux/kdebug.h>
	11	#include <linux/module.h>
	12	#include <linux/ptrace.h>
	13	#include <linux/kexec.h>
	14	#include <linux/bug.h>
	15	#include <linux/nmi.h>
ae87221d	16	#include <linux/sysfs.h>
6fcbede3 AH	17
	18	#include <asm/stacktrace.h>
	19
878719e8	20	#include "dumpstack.h"
6fcbede3 AH	21
	22	static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack,
	23	unsigned usedp, char *idp)
	24	{
	25	static char ids[][8] = {
	26	[DEBUG_STACK - 1] = "#DB",
	27	[NMI_STACK - 1] = "NMI",
	28	[DOUBLEFAULT_STACK - 1] = "#DF",
	29	[STACKFAULT_STACK - 1] = "#SS",
	30	[MCE_STACK - 1] = "#MC",
	31	#if DEBUG_STKSZ > EXCEPTION_STKSZ
	32	[N_EXCEPTION_STACKS ...
	33	N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]"
	34	#endif
	35	};
	36	unsigned k;
	37
	38	/*
	39	* Iterate over all exception stacks, and figure out whether
	40	* 'stack' is in one of them:
	41	*/
	42	for (k = 0; k < N_EXCEPTION_STACKS; k++) {
	43	unsigned long end = per_cpu(orig_ist, cpu).ist[k];
	44	/*
	45	* Is 'stack' above this exception frame's end?
	46	* If yes then skip to the next frame.
	47	*/
	48	if (stack >= end)
	49	continue;
	50	/*
	51	* Is 'stack' above this exception frame's start address?
	52	* If yes then we found the right frame.
	53	*/
	54	if (stack >= end - EXCEPTION_STKSZ) {
	55	/*
	56	* Make sure we only iterate through an exception
	57	* stack once. If it comes up for the second time
	58	* then there's something wrong going on - just
	59	* break out and return NULL:
	60	*/
	61	if (*usedp & (1U << k))
	62	break;
	63	*usedp \|= 1U << k;
	64	*idp = ids[k];
	65	return (unsigned long *)end;
	66	}
	67	/*
	68	* If this is a debug stack, and if it has a larger size than
	69	* the usual exception stacks, then 'stack' might still
	70	* be within the lower portion of the debug stack:
	71	*/
	72	#if DEBUG_STKSZ > EXCEPTION_STKSZ
	73	if (k == DEBUG_STACK - 1 && stack >= end - DEBUG_STKSZ) {
	74	unsigned j = N_EXCEPTION_STACKS - 1;
	75
	76	/*
	77	* Black magic. A large debug stack is composed of
	78	* multiple exception stack entries, which we
	79	* iterate through now. Dont look:
	80	*/
	81	do {
	82	++j;
	83	end -= EXCEPTION_STKSZ;
	84	ids[j][4] = '1' + (j - N_EXCEPTION_STACKS);
85	} while (stack < end - EXCEPTION_STKSZ);
86	if (*usedp & (1U << j))
87	break;
88	*usedp \|= 1U << j;
89	*idp = ids[j];
90	return (unsigned long *)end;
91	}
92	#endif
93	}
94	return NULL;
95	}
96
97	/*
98	* x86-64 can have up to three kernel stacks:
99	* process stack
100	* interrupt stack
101	* severe exception (double fault, nmi, stack fault, debug, mce) hardware stack
102	*/
103
6fcbede3 AH	104	void dump_trace(struct task_struct task, struct pt_regs regs,
	105	unsigned long *stack, unsigned long bp,
	106	const struct stacktrace_ops ops, void data)
	107	{
	108	const unsigned cpu = get_cpu();
	109	unsigned long irqstack_end = (unsigned long )cpu_pda(cpu)->irqstackptr;
	110	unsigned used = 0;
	111	struct thread_info *tinfo;
	112
	113	if (!task)
	114	task = current;
	115
	116	if (!stack) {
	117	unsigned long dummy;
	118	stack = &dummy;
	119	if (task && task != current)
	120	stack = (unsigned long *)task->thread.sp;
	121	}
	122
	123	#ifdef CONFIG_FRAME_POINTER
	124	if (!bp) {
	125	if (task == current) {
	126	/* Grab bp right from our regs */
8a541665	127	get_bp(bp);
6fcbede3 AH	128	} else {
	129	/* bp is the last reg pushed by switch_to */
	130	bp = (unsigned long ) task->thread.sp;
	131	}
	132	}
	133	#endif
	134
	135	/*
	136	* Print function call entries in all stacks, starting at the
	137	* current stack address. If the stacks consist of nested
	138	* exceptions
	139	*/
	140	tinfo = task_thread_info(task);
	141	for (;;) {
	142	char *id;
	143	unsigned long *estack_end;
	144	estack_end = in_exception_stack(cpu, (unsigned long)stack,
	145	&used, &id);
	146
	147	if (estack_end) {
	148	if (ops->stack(data, id) < 0)
	149	break;
	150
	151	bp = print_context_stack(tinfo, stack, bp, ops,
	152	data, estack_end);
	153	ops->stack(data, "<EOE>");
	154	/*
	155	* We link to the next stack via the
	156	* second-to-last pointer (index -2 to end) in the
	157	* exception stack:
	158	*/
	159	stack = (unsigned long *) estack_end[-2];
	160	continue;
	161	}
	162	if (irqstack_end) {
	163	unsigned long *irqstack;
	164	irqstack = irqstack_end -
	165	(IRQSTACKSIZE - 64) / sizeof(*irqstack);
	166
	167	if (stack >= irqstack && stack < irqstack_end) {
	168	if (ops->stack(data, "IRQ") < 0)
	169	break;
	170	bp = print_context_stack(tinfo, stack, bp,
	171	ops, data, irqstack_end);
	172	/*
	173	* We link to the next stack (which would be
	174	* the process stack normally) the last
	175	* pointer (index -1 to end) in the IRQ stack:
	176	*/
	177	stack = (unsigned long *) (irqstack_end[-1]);
	178	irqstack_end = NULL;
	179	ops->stack(data, "EOI");
	180	continue;
	181	}
	182	}
	183	break;
	184	}
	185
	186	/*
	187	* This handles the process stack:
	188	*/
	189	bp = print_context_stack(tinfo, stack, bp, ops, data, NULL);
	190	put_cpu();
	191	}
192	EXPORT_SYMBOL(dump_trace);
193
878719e8	194	void
6fcbede3 AH	195	show_stack_log_lvl(struct task_struct task, struct pt_regs regs,
	196	unsigned long sp, unsigned long bp, char log_lvl)
	197	{
	198	unsigned long *stack;
	199	int i;
	200	const int cpu = smp_processor_id();
	201	unsigned long *irqstack_end =
	202	(unsigned long *) (cpu_pda(cpu)->irqstackptr);
	203	unsigned long *irqstack =
	204	(unsigned long *) (cpu_pda(cpu)->irqstackptr - IRQSTACKSIZE);
	205
	206	/*
	207	* debugging aid: "show_stack(NULL, NULL);" prints the
	208	* back trace for this cpu.
	209	*/
	210
	211	if (sp == NULL) {
	212	if (task)
	213	sp = (unsigned long *)task->thread.sp;
	214	else
	215	sp = (unsigned long *)&sp;
	216	}
	217
	218	stack = sp;
	219	for (i = 0; i < kstack_depth_to_print; i++) {
	220	if (stack >= irqstack && stack <= irqstack_end) {
	221	if (stack == irqstack_end) {
	222	stack = (unsigned long *) (irqstack_end[-1]);
	223	printk(" <EOI> ");
	224	}
	225	} else {
	226	if (((long) stack & (THREAD_SIZE-1)) == 0)
	227	break;
	228	}
8a541665	229	if (i && ((i % STACKSLOTS_PER_LINE) == 0))
ca0a8164	230	printk("\n%s", log_lvl);
6fcbede3 AH	231	printk(" %016lx", *stack++);
	232	touch_nmi_watchdog();
	233	}
	234	printk("\n");
	235	show_trace_log_lvl(task, regs, sp, bp, log_lvl);
	236	}
	237
6fcbede3 AH	238	void show_registers(struct pt_regs *regs)
	239	{
	240	int i;
	241	unsigned long sp;
	242	const int cpu = smp_processor_id();
	243	struct task_struct *cur = cpu_pda(cpu)->pcurrent;
	244
	245	sp = regs->sp;
	246	printk("CPU %d ", cpu);
	247	__show_regs(regs, 1);
	248	printk("Process %s (pid: %d, threadinfo %p, task %p)\n",
	249	cur->comm, cur->pid, task_thread_info(cur), cur);
	250
	251	/*
	252	* When in-kernel, we also print out the stack and code at the
	253	* time of the fault..
	254	*/
	255	if (!user_mode(regs)) {
	256	unsigned int code_prologue = code_bytes * 43 / 64;
	257	unsigned int code_len = code_bytes;
	258	unsigned char c;
	259	u8 *ip;
	260
ca0a8164	261	printk(KERN_EMERG "Stack:\n");
6fcbede3	262	show_stack_log_lvl(NULL, regs, (unsigned long *)sp,
ca0a8164	263	regs->bp, KERN_EMERG);
6fcbede3 AH	264
	265	printk(KERN_EMERG "Code: ");
	266
	267	ip = (u8 *)regs->ip - code_prologue;
	268	if (ip < (u8 *)PAGE_OFFSET \|\| probe_kernel_address(ip, c)) {
8a541665	269	/* try starting at IP */
6fcbede3 AH	270	ip = (u8 *)regs->ip;
	271	code_len = code_len - code_prologue + 1;
	272	}
	273	for (i = 0; i < code_len; i++, ip++) {
	274	if (ip < (u8 *)PAGE_OFFSET \|\|
	275	probe_kernel_address(ip, c)) {
	276	printk(" Bad RIP value.");
	277	break;
	278	}
	279	if (ip == (u8 *)regs->ip)
	280	printk("<%02x> ", c);
	281	else
	282	printk("%02x ", c);
	283	}
	284	}
	285	printk("\n");
	286	}
	287
	288	int is_valid_bugaddr(unsigned long ip)
	289	{
	290	unsigned short ud2;
	291
	292	if (__copy_from_user(&ud2, (const void __user *) ip, sizeof(ud2)))
	293	return 0;
	294
	295	return ud2 == 0x0b0f;
	296	}
	297