[GitHub/mt8127/android_kernel_alcatel_ttab.git] / kernel / trace / trace_syscalls.c

#include <trace/syscall.h>
#include <trace/events/syscalls.h>
#include <linux/slab.h>
#include <linux/kernel.h>
#include <linux/ftrace.h>
#include <linux/perf_event.h>
#include <asm/syscall.h>

#include "trace_output.h"
#include "trace.h"

static DEFINE_MUTEX(syscall_trace_lock);
static int sys_refcount_enter;
static int sys_refcount_exit;
static DECLARE_BITMAP(enabled_enter_syscalls, NR_syscalls);
static DECLARE_BITMAP(enabled_exit_syscalls, NR_syscalls);

static int syscall_enter_register(struct ftrace_event_call *event,
				 enum trace_reg type);
static int syscall_exit_register(struct ftrace_event_call *event,
				 enum trace_reg type);

struct ftrace_event_class event_class_syscall_enter = {
	.system			= "syscalls",
	.reg			= syscall_enter_register
};

struct ftrace_event_class event_class_syscall_exit = {
	.system			= "syscalls",
	.reg			= syscall_exit_register
};

extern unsigned long __start_syscalls_metadata[];
extern unsigned long __stop_syscalls_metadata[];

static struct syscall_metadata **syscalls_metadata;

static struct syscall_metadata *find_syscall_meta(unsigned long syscall)
{
	struct syscall_metadata *start;
	struct syscall_metadata *stop;
	char str[KSYM_SYMBOL_LEN];


	start = (struct syscall_metadata *)__start_syscalls_metadata;
	stop = (struct syscall_metadata *)__stop_syscalls_metadata;
	kallsyms_lookup(syscall, NULL, NULL, NULL, str);

	for ( ; start < stop; start++) {
		/*
		 * Only compare after the "sys" prefix. Archs that use
		 * syscall wrappers may have syscalls symbols aliases prefixed
		 * with "SyS" instead of "sys", leading to an unwanted
		 * mismatch.
		 */
		if (start->name && !strcmp(start->name + 3, str + 3))
			return start;
	}
	return NULL;
}

static struct syscall_metadata *syscall_nr_to_meta(int nr)
{
	if (!syscalls_metadata || nr >= NR_syscalls || nr < 0)
		return NULL;

	return syscalls_metadata[nr];
}

enum print_line_t
print_syscall_enter(struct trace_iterator *iter, int flags)
{
	struct trace_seq *s = &iter->seq;
	struct trace_entry *ent = iter->ent;
	struct syscall_trace_enter *trace;
	struct syscall_metadata *entry;
	int i, ret, syscall;

	trace = (typeof(trace))ent;
	syscall = trace->nr;
	entry = syscall_nr_to_meta(syscall);

	if (!entry)
		goto end;

	if (entry->enter_event->id != ent->type) {
		WARN_ON_ONCE(1);
		goto end;
	}

	ret = trace_seq_printf(s, "%s(", entry->name);
	if (!ret)
		return TRACE_TYPE_PARTIAL_LINE;

	for (i = 0; i < entry->nb_args; i++) {
		/* parameter types */
		if (trace_flags & TRACE_ITER_VERBOSE) {
			ret = trace_seq_printf(s, "%s ", entry->types[i]);
			if (!ret)
				return TRACE_TYPE_PARTIAL_LINE;
		}
		/* parameter values */
		ret = trace_seq_printf(s, "%s: %lx%s", entry->args[i],
				       trace->args[i],
				       i == entry->nb_args - 1 ? "" : ", ");
		if (!ret)
			return TRACE_TYPE_PARTIAL_LINE;
	}

	ret = trace_seq_putc(s, ')');
	if (!ret)
		return TRACE_TYPE_PARTIAL_LINE;

end:
	ret =  trace_seq_putc(s, '\n');
	if (!ret)
		return TRACE_TYPE_PARTIAL_LINE;

	return TRACE_TYPE_HANDLED;
}

enum print_line_t
print_syscall_exit(struct trace_iterator *iter, int flags)
{
	struct trace_seq *s = &iter->seq;
	struct trace_entry *ent = iter->ent;
	struct syscall_trace_exit *trace;
	int syscall;
	struct syscall_metadata *entry;
	int ret;

	trace = (typeof(trace))ent;
	syscall = trace->nr;
	entry = syscall_nr_to_meta(syscall);

	if (!entry) {
		trace_seq_printf(s, "\n");
		return TRACE_TYPE_HANDLED;
	}

	if (entry->exit_event->id != ent->type) {
		WARN_ON_ONCE(1);
		return TRACE_TYPE_UNHANDLED;
	}

	ret = trace_seq_printf(s, "%s -> 0x%lx\n", entry->name,
				trace->ret);
	if (!ret)
		return TRACE_TYPE_PARTIAL_LINE;

	return TRACE_TYPE_HANDLED;
}

extern char *__bad_type_size(void);

#define SYSCALL_FIELD(type, name)					\
	sizeof(type) != sizeof(trace.name) ?				\
		__bad_type_size() :					\
		#type, #name, offsetof(typeof(trace), name),		\
		sizeof(trace.name), is_signed_type(type)

static
int  __set_enter_print_fmt(struct syscall_metadata *entry, char *buf, int len)
{
	int i;
	int pos = 0;

	/* When len=0, we just calculate the needed length */
#define LEN_OR_ZERO (len ? len - pos : 0)

	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"");
	for (i = 0; i < entry->nb_args; i++) {
		pos += snprintf(buf + pos, LEN_OR_ZERO, "%s: 0x%%0%zulx%s",
				entry->args[i], sizeof(unsigned long),
				i == entry->nb_args - 1 ? "" : ", ");
	}
	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"");

	for (i = 0; i < entry->nb_args; i++) {
		pos += snprintf(buf + pos, LEN_OR_ZERO,
				", ((unsigned long)(REC->%s))", entry->args[i]);
	}

#undef LEN_OR_ZERO

	/* return the length of print_fmt */
	return pos;
}

static int set_syscall_print_fmt(struct ftrace_event_call *call)
{
	char *print_fmt;
	int len;
	struct syscall_metadata *entry = call->data;

	if (entry->enter_event != call) {
		call->print_fmt = "\"0x%lx\", REC->ret";
		return 0;
	}

	/* First: called with 0 length to calculate the needed length */
	len = __set_enter_print_fmt(entry, NULL, 0);

	print_fmt = kmalloc(len + 1, GFP_KERNEL);
	if (!print_fmt)
		return -ENOMEM;

	/* Second: actually write the @print_fmt */
	__set_enter_print_fmt(entry, print_fmt, len + 1);
	call->print_fmt = print_fmt;

	return 0;
}

static void free_syscall_print_fmt(struct ftrace_event_call *call)
{
	struct syscall_metadata *entry = call->data;

	if (entry->enter_event == call)
		kfree(call->print_fmt);
}

int syscall_enter_define_fields(struct ftrace_event_call *call)
{
	struct syscall_trace_enter trace;
	struct syscall_metadata *meta = call->data;
	int ret;
	int i;
	int offset = offsetof(typeof(trace), args);

	ret = trace_define_field(call, SYSCALL_FIELD(int, nr), FILTER_OTHER);
	if (ret)
		return ret;

	for (i = 0; i < meta->nb_args; i++) {
		ret = trace_define_field(call, meta->types[i],
					 meta->args[i], offset,
					 sizeof(unsigned long), 0,
					 FILTER_OTHER);
		offset += sizeof(unsigned long);
	}

	return ret;
}

int syscall_exit_define_fields(struct ftrace_event_call *call)
{
	struct syscall_trace_exit trace;
	int ret;

	ret = trace_define_field(call, SYSCALL_FIELD(int, nr), FILTER_OTHER);
	if (ret)
		return ret;

	ret = trace_define_field(call, SYSCALL_FIELD(long, ret),
				 FILTER_OTHER);

	return ret;
}

void ftrace_syscall_enter(void *ignore, struct pt_regs *regs, long id)
{
	struct syscall_trace_enter *entry;
	struct syscall_metadata *sys_data;
	struct ring_buffer_event *event;
	struct ring_buffer *buffer;
	int size;
	int syscall_nr;

	syscall_nr = syscall_get_nr(current, regs);
	if (syscall_nr < 0)
		return;
	if (!test_bit(syscall_nr, enabled_enter_syscalls))
		return;

	sys_data = syscall_nr_to_meta(syscall_nr);
	if (!sys_data)
		return;

	size = sizeof(*entry) + sizeof(unsigned long) * sys_data->nb_args;

	event = trace_current_buffer_lock_reserve(&buffer,
			sys_data->enter_event->id, size, 0, 0);
	if (!event)
		return;

	entry = ring_buffer_event_data(event);
	entry->nr = syscall_nr;
	syscall_get_arguments(current, regs, 0, sys_data->nb_args, entry->args);

	if (!filter_current_check_discard(buffer, sys_data->enter_event,
					  entry, event))
		trace_current_buffer_unlock_commit(buffer, event, 0, 0);
}

void ftrace_syscall_exit(void *ignore, struct pt_regs *regs, long ret)
{
	struct syscall_trace_exit *entry;
	struct syscall_metadata *sys_data;
	struct ring_buffer_event *event;
	struct ring_buffer *buffer;
	int syscall_nr;

	syscall_nr = syscall_get_nr(current, regs);
	if (syscall_nr < 0)
		return;
	if (!test_bit(syscall_nr, enabled_exit_syscalls))
		return;

	sys_data = syscall_nr_to_meta(syscall_nr);
	if (!sys_data)
		return;

	event = trace_current_buffer_lock_reserve(&buffer,
			sys_data->exit_event->id, sizeof(*entry), 0, 0);
	if (!event)
		return;

	entry = ring_buffer_event_data(event);
	entry->nr = syscall_nr;
	entry->ret = syscall_get_return_value(current, regs);

	if (!filter_current_check_discard(buffer, sys_data->exit_event,
					  entry, event))
		trace_current_buffer_unlock_commit(buffer, event, 0, 0);
}

int reg_event_syscall_enter(struct ftrace_event_call *call)
{
	int ret = 0;
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;
	if (num < 0 || num >= NR_syscalls)
		return -ENOSYS;
	mutex_lock(&syscall_trace_lock);
	if (!sys_refcount_enter)
		ret = register_trace_sys_enter(ftrace_syscall_enter, NULL);
	if (!ret) {
		set_bit(num, enabled_enter_syscalls);
		sys_refcount_enter++;
	}
	mutex_unlock(&syscall_trace_lock);
	return ret;
}

void unreg_event_syscall_enter(struct ftrace_event_call *call)
{
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;
	if (num < 0 || num >= NR_syscalls)
		return;
	mutex_lock(&syscall_trace_lock);
	sys_refcount_enter--;
	clear_bit(num, enabled_enter_syscalls);
	if (!sys_refcount_enter)
		unregister_trace_sys_enter(ftrace_syscall_enter, NULL);
	mutex_unlock(&syscall_trace_lock);
}

int reg_event_syscall_exit(struct ftrace_event_call *call)
{
	int ret = 0;
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;
	if (num < 0 || num >= NR_syscalls)
		return -ENOSYS;
	mutex_lock(&syscall_trace_lock);
	if (!sys_refcount_exit)
		ret = register_trace_sys_exit(ftrace_syscall_exit, NULL);
	if (!ret) {
		set_bit(num, enabled_exit_syscalls);
		sys_refcount_exit++;
	}
	mutex_unlock(&syscall_trace_lock);
	return ret;
}

void unreg_event_syscall_exit(struct ftrace_event_call *call)
{
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;
	if (num < 0 || num >= NR_syscalls)
		return;
	mutex_lock(&syscall_trace_lock);
	sys_refcount_exit--;
	clear_bit(num, enabled_exit_syscalls);
	if (!sys_refcount_exit)
		unregister_trace_sys_exit(ftrace_syscall_exit, NULL);
	mutex_unlock(&syscall_trace_lock);
}

int init_syscall_trace(struct ftrace_event_call *call)
{
	int id;

	if (set_syscall_print_fmt(call) < 0)
		return -ENOMEM;

	id = trace_event_raw_init(call);

	if (id < 0) {
		free_syscall_print_fmt(call);
		return id;
	}

	return id;
}

unsigned long __init arch_syscall_addr(int nr)
{
	return (unsigned long)sys_call_table[nr];
}

int __init init_ftrace_syscalls(void)
{
	struct syscall_metadata *meta;
	unsigned long addr;
	int i;

	syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) *
					NR_syscalls, GFP_KERNEL);
	if (!syscalls_metadata) {
		WARN_ON(1);
		return -ENOMEM;
	}

	for (i = 0; i < NR_syscalls; i++) {
		addr = arch_syscall_addr(i);
		meta = find_syscall_meta(addr);
		if (!meta)
			continue;

		meta->syscall_nr = i;
		syscalls_metadata[i] = meta;
	}

	return 0;
}
core_initcall(init_ftrace_syscalls);

#ifdef CONFIG_PERF_EVENTS

static DECLARE_BITMAP(enabled_perf_enter_syscalls, NR_syscalls);
static DECLARE_BITMAP(enabled_perf_exit_syscalls, NR_syscalls);
static int sys_perf_refcount_enter;
static int sys_perf_refcount_exit;

static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id)
{
	struct syscall_metadata *sys_data;
	struct syscall_trace_enter *rec;
	unsigned long flags;
	int syscall_nr;
	int rctx;
	int size;

	syscall_nr = syscall_get_nr(current, regs);
	if (!test_bit(syscall_nr, enabled_perf_enter_syscalls))
		return;

	sys_data = syscall_nr_to_meta(syscall_nr);
	if (!sys_data)
		return;

	/* get the size after alignment with the u32 buffer size field */
	size = sizeof(unsigned long) * sys_data->nb_args + sizeof(*rec);
	size = ALIGN(size + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);

	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
		      "perf buffer not large enough"))
		return;

	rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size,
				sys_data->enter_event->id, &rctx, &flags);
	if (!rec)
		return;

	rec->nr = syscall_nr;
	syscall_get_arguments(current, regs, 0, sys_data->nb_args,
			       (unsigned long *)&rec->args);
	perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs);
}

int perf_sysenter_enable(struct ftrace_event_call *call)
{
	int ret = 0;
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;

	mutex_lock(&syscall_trace_lock);
	if (!sys_perf_refcount_enter)
		ret = register_trace_sys_enter(perf_syscall_enter, NULL);
	if (ret) {
		pr_info("event trace: Could not activate"
				"syscall entry trace point");
	} else {
		set_bit(num, enabled_perf_enter_syscalls);
		sys_perf_refcount_enter++;
	}
	mutex_unlock(&syscall_trace_lock);
	return ret;
}

void perf_sysenter_disable(struct ftrace_event_call *call)
{
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;

	mutex_lock(&syscall_trace_lock);
	sys_perf_refcount_enter--;
	clear_bit(num, enabled_perf_enter_syscalls);
	if (!sys_perf_refcount_enter)
		unregister_trace_sys_enter(perf_syscall_enter, NULL);
	mutex_unlock(&syscall_trace_lock);
}

static void perf_syscall_exit(void *ignore, struct pt_regs *regs, long ret)
{
	struct syscall_metadata *sys_data;
	struct syscall_trace_exit *rec;
	unsigned long flags;
	int syscall_nr;
	int rctx;
	int size;

	syscall_nr = syscall_get_nr(current, regs);
	if (!test_bit(syscall_nr, enabled_perf_exit_syscalls))
		return;

	sys_data = syscall_nr_to_meta(syscall_nr);
	if (!sys_data)
		return;

	/* We can probably do that at build time */
	size = ALIGN(sizeof(*rec) + sizeof(u32), sizeof(u64));
	size -= sizeof(u32);

	/*
	 * Impossible, but be paranoid with the future
	 * How to put this check outside runtime?
	 */
	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
		"exit event has grown above perf buffer size"))
		return;

	rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size,
				sys_data->exit_event->id, &rctx, &flags);
	if (!rec)
		return;

	rec->nr = syscall_nr;
	rec->ret = syscall_get_return_value(current, regs);

	perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs);
}

int perf_sysexit_enable(struct ftrace_event_call *call)
{
	int ret = 0;
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;

	mutex_lock(&syscall_trace_lock);
	if (!sys_perf_refcount_exit)
		ret = register_trace_sys_exit(perf_syscall_exit, NULL);
	if (ret) {
		pr_info("event trace: Could not activate"
				"syscall exit trace point");
	} else {
		set_bit(num, enabled_perf_exit_syscalls);
		sys_perf_refcount_exit++;
	}
	mutex_unlock(&syscall_trace_lock);
	return ret;
}

void perf_sysexit_disable(struct ftrace_event_call *call)
{
	int num;

	num = ((struct syscall_metadata *)call->data)->syscall_nr;

	mutex_lock(&syscall_trace_lock);
	sys_perf_refcount_exit--;
	clear_bit(num, enabled_perf_exit_syscalls);
	if (!sys_perf_refcount_exit)
		unregister_trace_sys_exit(perf_syscall_exit, NULL);
	mutex_unlock(&syscall_trace_lock);
}

#endif /* CONFIG_PERF_EVENTS */

static int syscall_enter_register(struct ftrace_event_call *event,
				 enum trace_reg type)
{
	switch (type) {
	case TRACE_REG_REGISTER:
		return reg_event_syscall_enter(event);
	case TRACE_REG_UNREGISTER:
		unreg_event_syscall_enter(event);
		return 0;

#ifdef CONFIG_PERF_EVENTS
	case TRACE_REG_PERF_REGISTER:
		return perf_sysenter_enable(event);
	case TRACE_REG_PERF_UNREGISTER:
		perf_sysenter_disable(event);
		return 0;
#endif
	}
	return 0;
}

static int syscall_exit_register(struct ftrace_event_call *event,
				 enum trace_reg type)
{
	switch (type) {
	case TRACE_REG_REGISTER:
		return reg_event_syscall_exit(event);
	case TRACE_REG_UNREGISTER:
		unreg_event_syscall_exit(event);
		return 0;

#ifdef CONFIG_PERF_EVENTS
	case TRACE_REG_PERF_REGISTER:
		return perf_sysexit_enable(event);
	case TRACE_REG_PERF_UNREGISTER:
		perf_sysexit_disable(event);
		return 0;
#endif
	}
	return 0;
}
Commit	Line	Data
47788c58	1	#include <trace/syscall.h>
1c569f02	2	#include <trace/events/syscalls.h>
5a0e3ad6	3	#include <linux/slab.h>
ee08c6ec	4	#include <linux/kernel.h>
fb34a08c	5	#include <linux/ftrace.h>
cdd6c482	6	#include <linux/perf_event.h>
ee08c6ec FW	7	#include <asm/syscall.h>
	8
	9	#include "trace_output.h"
	10	#include "trace.h"
	11
5be71b61	12	static DEFINE_MUTEX(syscall_trace_lock);
fb34a08c JB	13	static int sys_refcount_enter;
fb34a08c JB	14	static int sys_refcount_exit;
57421dbb JB	15	static DECLARE_BITMAP(enabled_enter_syscalls, NR_syscalls);
57421dbb JB	16	static DECLARE_BITMAP(enabled_exit_syscalls, NR_syscalls);
ee08c6ec	17
2239291a SR	18	static int syscall_enter_register(struct ftrace_event_call *event,
	19	enum trace_reg type);
	20	static int syscall_exit_register(struct ftrace_event_call *event,
	21	enum trace_reg type);
	22
	23	struct ftrace_event_class event_class_syscall_enter = {
	24	.system = "syscalls",
	25	.reg = syscall_enter_register
	26	};
	27
	28	struct ftrace_event_class event_class_syscall_exit = {
	29	.system = "syscalls",
	30	.reg = syscall_exit_register
8f082018 SR	31	};
8f082018 SR	32
c44fc770 FW	33	extern unsigned long __start_syscalls_metadata[];
	34	extern unsigned long __stop_syscalls_metadata[];
	35
	36	static struct syscall_metadata **syscalls_metadata;
	37
	38	static struct syscall_metadata *find_syscall_meta(unsigned long syscall)
	39	{
	40	struct syscall_metadata *start;
	41	struct syscall_metadata *stop;
	42	char str[KSYM_SYMBOL_LEN];
	43
	44
	45	start = (struct syscall_metadata *)__start_syscalls_metadata;
	46	stop = (struct syscall_metadata *)__stop_syscalls_metadata;
	47	kallsyms_lookup(syscall, NULL, NULL, NULL, str);
	48
	49	for ( ; start < stop; start++) {
	50	/*
	51	* Only compare after the "sys" prefix. Archs that use
	52	* syscall wrappers may have syscalls symbols aliases prefixed
	53	* with "SyS" instead of "sys", leading to an unwanted
	54	* mismatch.
	55	*/
	56	if (start->name && !strcmp(start->name + 3, str + 3))
	57	return start;
	58	}
	59	return NULL;
	60	}
	61
	62	static struct syscall_metadata *syscall_nr_to_meta(int nr)
	63	{
	64	if (!syscalls_metadata \|\| nr >= NR_syscalls \|\| nr < 0)
	65	return NULL;
	66
	67	return syscalls_metadata[nr];
	68	}
	69
bed1ffca FW	70	enum print_line_t
	71	print_syscall_enter(struct trace_iterator *iter, int flags)
	72	{
	73	struct trace_seq *s = &iter->seq;
	74	struct trace_entry *ent = iter->ent;
	75	struct syscall_trace_enter *trace;
	76	struct syscall_metadata *entry;
	77	int i, ret, syscall;
	78
64c12e04	79	trace = (typeof(trace))ent;
bed1ffca	80	syscall = trace->nr;
bed1ffca	81	entry = syscall_nr_to_meta(syscall);
64c12e04	82
bed1ffca FW	83	if (!entry)
	84	goto end;
	85
fcc19438	86	if (entry->enter_event->id != ent->type) {
64c12e04 JB	87	WARN_ON_ONCE(1);
	88	goto end;
	89	}
	90
bed1ffca FW	91	ret = trace_seq_printf(s, "%s(", entry->name);
	92	if (!ret)
	93	return TRACE_TYPE_PARTIAL_LINE;
	94
	95	for (i = 0; i < entry->nb_args; i++) {
	96	/* parameter types */
ba8b3a40	97	if (trace_flags & TRACE_ITER_VERBOSE) {
bed1ffca FW	98	ret = trace_seq_printf(s, "%s ", entry->types[i]);
	99	if (!ret)
	100	return TRACE_TYPE_PARTIAL_LINE;
	101	}
	102	/* parameter values */
4539f077	103	ret = trace_seq_printf(s, "%s: %lx%s", entry->args[i],
bed1ffca	104	trace->args[i],
4539f077	105	i == entry->nb_args - 1 ? "" : ", ");
bed1ffca FW	106	if (!ret)
	107	return TRACE_TYPE_PARTIAL_LINE;
	108	}
	109
4539f077 LZ	110	ret = trace_seq_putc(s, ')');
	111	if (!ret)
	112	return TRACE_TYPE_PARTIAL_LINE;
	113
bed1ffca	114	end:
4539f077 LZ	115	ret = trace_seq_putc(s, '\n');
	116	if (!ret)
	117	return TRACE_TYPE_PARTIAL_LINE;
	118
bed1ffca FW	119	return TRACE_TYPE_HANDLED;
	120	}
	121
	122	enum print_line_t
	123	print_syscall_exit(struct trace_iterator *iter, int flags)
	124	{
	125	struct trace_seq *s = &iter->seq;
	126	struct trace_entry *ent = iter->ent;
	127	struct syscall_trace_exit *trace;
	128	int syscall;
	129	struct syscall_metadata *entry;
	130	int ret;
	131
64c12e04	132	trace = (typeof(trace))ent;
bed1ffca	133	syscall = trace->nr;
bed1ffca	134	entry = syscall_nr_to_meta(syscall);
64c12e04	135
bed1ffca FW	136	if (!entry) {
	137	trace_seq_printf(s, "\n");
	138	return TRACE_TYPE_HANDLED;
	139	}
	140
fcc19438	141	if (entry->exit_event->id != ent->type) {
64c12e04 JB	142	WARN_ON_ONCE(1);
	143	return TRACE_TYPE_UNHANDLED;
	144	}
	145
bed1ffca FW	146	ret = trace_seq_printf(s, "%s -> 0x%lx\n", entry->name,
	147	trace->ret);
	148	if (!ret)
	149	return TRACE_TYPE_PARTIAL_LINE;
	150
	151	return TRACE_TYPE_HANDLED;
	152	}
	153
e6971969 LZ	154	extern char *__bad_type_size(void);
	155
	156	#define SYSCALL_FIELD(type, name) \
	157	sizeof(type) != sizeof(trace.name) ? \
	158	__bad_type_size() : \
26a50744 TZ	159	#type, #name, offsetof(typeof(trace), name), \
26a50744 TZ	160	sizeof(trace.name), is_signed_type(type)
e6971969	161
50307a45 LJ	162	static
	163	int __set_enter_print_fmt(struct syscall_metadata entry, char buf, int len)
	164	{
	165	int i;
	166	int pos = 0;
	167
	168	/* When len=0, we just calculate the needed length */
	169	#define LEN_OR_ZERO (len ? len - pos : 0)
	170
	171	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"");
	172	for (i = 0; i < entry->nb_args; i++) {
	173	pos += snprintf(buf + pos, LEN_OR_ZERO, "%s: 0x%%0%zulx%s",
	174	entry->args[i], sizeof(unsigned long),
	175	i == entry->nb_args - 1 ? "" : ", ");
	176	}
	177	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"");
	178
	179	for (i = 0; i < entry->nb_args; i++) {
	180	pos += snprintf(buf + pos, LEN_OR_ZERO,
	181	", ((unsigned long)(REC->%s))", entry->args[i]);
	182	}
	183
	184	#undef LEN_OR_ZERO
	185
	186	/* return the length of print_fmt */
	187	return pos;
	188	}
	189
	190	static int set_syscall_print_fmt(struct ftrace_event_call *call)
	191	{
	192	char *print_fmt;
	193	int len;
	194	struct syscall_metadata *entry = call->data;
	195
	196	if (entry->enter_event != call) {
	197	call->print_fmt = "\"0x%lx\", REC->ret";
	198	return 0;
	199	}
	200
	201	/* First: called with 0 length to calculate the needed length */
	202	len = __set_enter_print_fmt(entry, NULL, 0);
	203
	204	print_fmt = kmalloc(len + 1, GFP_KERNEL);
	205	if (!print_fmt)
	206	return -ENOMEM;
	207
	208	/* Second: actually write the @print_fmt */
	209	__set_enter_print_fmt(entry, print_fmt, len + 1);
	210	call->print_fmt = print_fmt;
	211
	212	return 0;
	213	}
	214
	215	static void free_syscall_print_fmt(struct ftrace_event_call *call)
	216	{
	217	struct syscall_metadata *entry = call->data;
	218
	219	if (entry->enter_event == call)
	220	kfree(call->print_fmt);
	221	}
	222
540b7b8d LZ	223	int syscall_enter_define_fields(struct ftrace_event_call *call)
	224	{
	225	struct syscall_trace_enter trace;
31c16b13	226	struct syscall_metadata *meta = call->data;
540b7b8d	227	int ret;
540b7b8d LZ	228	int i;
	229	int offset = offsetof(typeof(trace), args);
	230
0f1ef51d LJ	231	ret = trace_define_field(call, SYSCALL_FIELD(int, nr), FILTER_OTHER);
	232	if (ret)
	233	return ret;
	234
540b7b8d	235	for (i = 0; i < meta->nb_args; i++) {
aeaeae11 FW	236	ret = trace_define_field(call, meta->types[i],
aeaeae11 FW	237	meta->args[i], offset,
43b51ead LZ	238	sizeof(unsigned long), 0,
43b51ead LZ	239	FILTER_OTHER);
540b7b8d LZ	240	offset += sizeof(unsigned long);
	241	}
	242
	243	return ret;
	244	}
	245
	246	int syscall_exit_define_fields(struct ftrace_event_call *call)
	247	{
	248	struct syscall_trace_exit trace;
	249	int ret;
	250
0f1ef51d LJ	251	ret = trace_define_field(call, SYSCALL_FIELD(int, nr), FILTER_OTHER);
	252	if (ret)
	253	return ret;
	254
26a50744	255	ret = trace_define_field(call, SYSCALL_FIELD(long, ret),
43b51ead	256	FILTER_OTHER);
540b7b8d LZ	257
	258	return ret;
	259	}
	260
38516ab5	261	void ftrace_syscall_enter(void ignore, struct pt_regs regs, long id)
ee08c6ec	262	{
bed1ffca FW	263	struct syscall_trace_enter *entry;
	264	struct syscall_metadata *sys_data;
	265	struct ring_buffer_event *event;
e77405ad	266	struct ring_buffer *buffer;
bed1ffca	267	int size;
ee08c6ec FW	268	int syscall_nr;
	269
	270	syscall_nr = syscall_get_nr(current, regs);
cd0980fc HB	271	if (syscall_nr < 0)
cd0980fc HB	272	return;
fb34a08c JB	273	if (!test_bit(syscall_nr, enabled_enter_syscalls))
fb34a08c JB	274	return;
ee08c6ec	275
bed1ffca FW	276	sys_data = syscall_nr_to_meta(syscall_nr);
	277	if (!sys_data)
	278	return;
	279
	280	size = sizeof(entry) + sizeof(unsigned long) sys_data->nb_args;
	281
fcc19438 LJ	282	event = trace_current_buffer_lock_reserve(&buffer,
fcc19438 LJ	283	sys_data->enter_event->id, size, 0, 0);
bed1ffca FW	284	if (!event)
	285	return;
	286
	287	entry = ring_buffer_event_data(event);
	288	entry->nr = syscall_nr;
	289	syscall_get_arguments(current, regs, 0, sys_data->nb_args, entry->args);
	290
e77405ad SR	291	if (!filter_current_check_discard(buffer, sys_data->enter_event,
	292	entry, event))
	293	trace_current_buffer_unlock_commit(buffer, event, 0, 0);
ee08c6ec FW	294	}
ee08c6ec FW	295
38516ab5	296	void ftrace_syscall_exit(void ignore, struct pt_regs regs, long ret)
ee08c6ec	297	{
bed1ffca FW	298	struct syscall_trace_exit *entry;
	299	struct syscall_metadata *sys_data;
	300	struct ring_buffer_event *event;
e77405ad	301	struct ring_buffer *buffer;
ee08c6ec FW	302	int syscall_nr;
	303
	304	syscall_nr = syscall_get_nr(current, regs);
cd0980fc HB	305	if (syscall_nr < 0)
cd0980fc HB	306	return;
fb34a08c JB	307	if (!test_bit(syscall_nr, enabled_exit_syscalls))
fb34a08c JB	308	return;
ee08c6ec	309
bed1ffca FW	310	sys_data = syscall_nr_to_meta(syscall_nr);
	311	if (!sys_data)
	312	return;
	313
fcc19438 LJ	314	event = trace_current_buffer_lock_reserve(&buffer,
fcc19438 LJ	315	sys_data->exit_event->id, sizeof(*entry), 0, 0);
bed1ffca FW	316	if (!event)
	317	return;
	318
	319	entry = ring_buffer_event_data(event);
	320	entry->nr = syscall_nr;
	321	entry->ret = syscall_get_return_value(current, regs);
	322
e77405ad SR	323	if (!filter_current_check_discard(buffer, sys_data->exit_event,
	324	entry, event))
	325	trace_current_buffer_unlock_commit(buffer, event, 0, 0);
ee08c6ec FW	326	}
ee08c6ec FW	327
bd1a5c84	328	int reg_event_syscall_enter(struct ftrace_event_call *call)
ee08c6ec	329	{
fb34a08c JB	330	int ret = 0;
fb34a08c JB	331	int num;
fb34a08c	332
c252f657	333	num = ((struct syscall_metadata *)call->data)->syscall_nr;
57421dbb	334	if (num < 0 \|\| num >= NR_syscalls)
fb34a08c JB	335	return -ENOSYS;
	336	mutex_lock(&syscall_trace_lock);
	337	if (!sys_refcount_enter)
38516ab5	338	ret = register_trace_sys_enter(ftrace_syscall_enter, NULL);
3b8e4273	339	if (!ret) {
fb34a08c JB	340	set_bit(num, enabled_enter_syscalls);
	341	sys_refcount_enter++;
	342	}
	343	mutex_unlock(&syscall_trace_lock);
	344	return ret;
ee08c6ec FW	345	}
ee08c6ec FW	346
bd1a5c84	347	void unreg_event_syscall_enter(struct ftrace_event_call *call)
ee08c6ec	348	{
fb34a08c	349	int num;
ee08c6ec	350
c252f657	351	num = ((struct syscall_metadata *)call->data)->syscall_nr;
57421dbb	352	if (num < 0 \|\| num >= NR_syscalls)
fb34a08c JB	353	return;
	354	mutex_lock(&syscall_trace_lock);
	355	sys_refcount_enter--;
	356	clear_bit(num, enabled_enter_syscalls);
	357	if (!sys_refcount_enter)
38516ab5	358	unregister_trace_sys_enter(ftrace_syscall_enter, NULL);
fb34a08c JB	359	mutex_unlock(&syscall_trace_lock);
fb34a08c JB	360	}
ee08c6ec	361
bd1a5c84	362	int reg_event_syscall_exit(struct ftrace_event_call *call)
ee08c6ec	363	{
fb34a08c JB	364	int ret = 0;
fb34a08c JB	365	int num;
fb34a08c	366
c252f657	367	num = ((struct syscall_metadata *)call->data)->syscall_nr;
57421dbb	368	if (num < 0 \|\| num >= NR_syscalls)
fb34a08c JB	369	return -ENOSYS;
	370	mutex_lock(&syscall_trace_lock);
	371	if (!sys_refcount_exit)
38516ab5	372	ret = register_trace_sys_exit(ftrace_syscall_exit, NULL);
3b8e4273	373	if (!ret) {
fb34a08c JB	374	set_bit(num, enabled_exit_syscalls);
fb34a08c JB	375	sys_refcount_exit++;
ee08c6ec	376	}
fb34a08c JB	377	mutex_unlock(&syscall_trace_lock);
	378	return ret;
	379	}
ee08c6ec	380
bd1a5c84	381	void unreg_event_syscall_exit(struct ftrace_event_call *call)
fb34a08c JB	382	{
fb34a08c JB	383	int num;
ee08c6ec	384
c252f657	385	num = ((struct syscall_metadata *)call->data)->syscall_nr;
57421dbb	386	if (num < 0 \|\| num >= NR_syscalls)
fb34a08c JB	387	return;
	388	mutex_lock(&syscall_trace_lock);
	389	sys_refcount_exit--;
	390	clear_bit(num, enabled_exit_syscalls);
	391	if (!sys_refcount_exit)
38516ab5	392	unregister_trace_sys_exit(ftrace_syscall_exit, NULL);
fb34a08c	393	mutex_unlock(&syscall_trace_lock);
ee08c6ec	394	}
fb34a08c	395
a1301da0 LJ	396	int init_syscall_trace(struct ftrace_event_call *call)
	397	{
	398	int id;
	399
50307a45 LJ	400	if (set_syscall_print_fmt(call) < 0)
	401	return -ENOMEM;
	402
c7ef3a90 SR	403	id = trace_event_raw_init(call);
	404
	405	if (id < 0) {
50307a45	406	free_syscall_print_fmt(call);
c7ef3a90	407	return id;
50307a45	408	}
c7ef3a90 SR	409
c7ef3a90 SR	410	return id;
a1301da0 LJ	411	}
a1301da0 LJ	412
e7b8e675 MF	413	unsigned long __init arch_syscall_addr(int nr)
	414	{
	415	return (unsigned long)sys_call_table[nr];
	416	}
	417
c44fc770 FW	418	int __init init_ftrace_syscalls(void)
	419	{
	420	struct syscall_metadata *meta;
	421	unsigned long addr;
	422	int i;
	423
	424	syscalls_metadata = kzalloc(sizeof(syscalls_metadata)
	425	NR_syscalls, GFP_KERNEL);
	426	if (!syscalls_metadata) {
	427	WARN_ON(1);
	428	return -ENOMEM;
	429	}
	430
	431	for (i = 0; i < NR_syscalls; i++) {
	432	addr = arch_syscall_addr(i);
	433	meta = find_syscall_meta(addr);
c252f657 LJ	434	if (!meta)
	435	continue;
	436
	437	meta->syscall_nr = i;
c44fc770 FW	438	syscalls_metadata[i] = meta;
	439	}
	440
	441	return 0;
	442	}
	443	core_initcall(init_ftrace_syscalls);
	444
07b139c8	445	#ifdef CONFIG_PERF_EVENTS
19007a67	446
97d5a220 FW	447	static DECLARE_BITMAP(enabled_perf_enter_syscalls, NR_syscalls);
	448	static DECLARE_BITMAP(enabled_perf_exit_syscalls, NR_syscalls);
	449	static int sys_perf_refcount_enter;
	450	static int sys_perf_refcount_exit;
f4b5ffcc	451
38516ab5	452	static void perf_syscall_enter(void ignore, struct pt_regs regs, long id)
f4b5ffcc JB	453	{
f4b5ffcc JB	454	struct syscall_metadata *sys_data;
20ab4425 FW	455	struct syscall_trace_enter *rec;
20ab4425 FW	456	unsigned long flags;
f4b5ffcc	457	int syscall_nr;
4ed7c92d	458	int rctx;
19007a67	459	int size;
f4b5ffcc JB	460
f4b5ffcc JB	461	syscall_nr = syscall_get_nr(current, regs);
97d5a220	462	if (!test_bit(syscall_nr, enabled_perf_enter_syscalls))
f4b5ffcc JB	463	return;
	464
	465	sys_data = syscall_nr_to_meta(syscall_nr);
	466	if (!sys_data)
	467	return;
	468
19007a67 FW	469	/* get the size after alignment with the u32 buffer size field */
	470	size = sizeof(unsigned long) * sys_data->nb_args + sizeof(*rec);
	471	size = ALIGN(size + sizeof(u32), sizeof(u64));
	472	size -= sizeof(u32);
	473
97d5a220 FW	474	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
97d5a220 FW	475	"perf buffer not large enough"))
20ab4425 FW	476	return;
20ab4425 FW	477
97d5a220	478	rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size,
430ad5a6 XG	479	sys_data->enter_event->id, &rctx, &flags);
	480	if (!rec)
	481	return;
20ab4425	482
20ab4425 FW	483	rec->nr = syscall_nr;
	484	syscall_get_arguments(current, regs, 0, sys_data->nb_args,
	485	(unsigned long *)&rec->args);
97d5a220	486	perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs);
f4b5ffcc JB	487	}
f4b5ffcc JB	488
97d5a220	489	int perf_sysenter_enable(struct ftrace_event_call *call)
f4b5ffcc JB	490	{
	491	int ret = 0;
	492	int num;
	493
3bbe84e9	494	num = ((struct syscall_metadata *)call->data)->syscall_nr;
f4b5ffcc JB	495
f4b5ffcc JB	496	mutex_lock(&syscall_trace_lock);
97d5a220	497	if (!sys_perf_refcount_enter)
38516ab5	498	ret = register_trace_sys_enter(perf_syscall_enter, NULL);
f4b5ffcc JB	499	if (ret) {
	500	pr_info("event trace: Could not activate"
	501	"syscall entry trace point");
	502	} else {
97d5a220 FW	503	set_bit(num, enabled_perf_enter_syscalls);
97d5a220 FW	504	sys_perf_refcount_enter++;
f4b5ffcc JB	505	}
	506	mutex_unlock(&syscall_trace_lock);
	507	return ret;
	508	}
	509
97d5a220	510	void perf_sysenter_disable(struct ftrace_event_call *call)
f4b5ffcc JB	511	{
	512	int num;
	513
3bbe84e9	514	num = ((struct syscall_metadata *)call->data)->syscall_nr;
f4b5ffcc JB	515
f4b5ffcc JB	516	mutex_lock(&syscall_trace_lock);
97d5a220 FW	517	sys_perf_refcount_enter--;
	518	clear_bit(num, enabled_perf_enter_syscalls);
	519	if (!sys_perf_refcount_enter)
38516ab5	520	unregister_trace_sys_enter(perf_syscall_enter, NULL);
f4b5ffcc JB	521	mutex_unlock(&syscall_trace_lock);
	522	}
	523
38516ab5	524	static void perf_syscall_exit(void ignore, struct pt_regs regs, long ret)
f4b5ffcc JB	525	{
f4b5ffcc JB	526	struct syscall_metadata *sys_data;
20ab4425 FW	527	struct syscall_trace_exit *rec;
20ab4425 FW	528	unsigned long flags;
f4b5ffcc	529	int syscall_nr;
4ed7c92d	530	int rctx;
20ab4425	531	int size;
f4b5ffcc JB	532
f4b5ffcc JB	533	syscall_nr = syscall_get_nr(current, regs);
97d5a220	534	if (!test_bit(syscall_nr, enabled_perf_exit_syscalls))
f4b5ffcc JB	535	return;
	536
	537	sys_data = syscall_nr_to_meta(syscall_nr);
	538	if (!sys_data)
	539	return;
	540
20ab4425 FW	541	/* We can probably do that at build time */
	542	size = ALIGN(sizeof(*rec) + sizeof(u32), sizeof(u64));
	543	size -= sizeof(u32);
19007a67	544
20ab4425 FW	545	/*
	546	* Impossible, but be paranoid with the future
	547	* How to put this check outside runtime?
	548	*/
97d5a220 FW	549	if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
97d5a220 FW	550	"exit event has grown above perf buffer size"))
20ab4425 FW	551	return;
20ab4425 FW	552
97d5a220	553	rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size,
430ad5a6 XG	554	sys_data->exit_event->id, &rctx, &flags);
	555	if (!rec)
	556	return;
20ab4425	557
20ab4425 FW	558	rec->nr = syscall_nr;
	559	rec->ret = syscall_get_return_value(current, regs);
	560
97d5a220	561	perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs);
f4b5ffcc JB	562	}
f4b5ffcc JB	563
97d5a220	564	int perf_sysexit_enable(struct ftrace_event_call *call)
f4b5ffcc JB	565	{
	566	int ret = 0;
	567	int num;
	568
3bbe84e9	569	num = ((struct syscall_metadata *)call->data)->syscall_nr;
f4b5ffcc JB	570
f4b5ffcc JB	571	mutex_lock(&syscall_trace_lock);
97d5a220	572	if (!sys_perf_refcount_exit)
38516ab5	573	ret = register_trace_sys_exit(perf_syscall_exit, NULL);
f4b5ffcc JB	574	if (ret) {
f4b5ffcc JB	575	pr_info("event trace: Could not activate"
6574658b	576	"syscall exit trace point");
f4b5ffcc	577	} else {
97d5a220 FW	578	set_bit(num, enabled_perf_exit_syscalls);
97d5a220 FW	579	sys_perf_refcount_exit++;
f4b5ffcc JB	580	}
	581	mutex_unlock(&syscall_trace_lock);
	582	return ret;
	583	}
	584
97d5a220	585	void perf_sysexit_disable(struct ftrace_event_call *call)
f4b5ffcc JB	586	{
	587	int num;
	588
3bbe84e9	589	num = ((struct syscall_metadata *)call->data)->syscall_nr;
f4b5ffcc JB	590
f4b5ffcc JB	591	mutex_lock(&syscall_trace_lock);
97d5a220 FW	592	sys_perf_refcount_exit--;
	593	clear_bit(num, enabled_perf_exit_syscalls);
	594	if (!sys_perf_refcount_exit)
38516ab5	595	unregister_trace_sys_exit(perf_syscall_exit, NULL);
f4b5ffcc JB	596	mutex_unlock(&syscall_trace_lock);
	597	}
	598
07b139c8	599	#endif /* CONFIG_PERF_EVENTS */
f4b5ffcc	600
2239291a SR	601	static int syscall_enter_register(struct ftrace_event_call *event,
	602	enum trace_reg type)
	603	{
	604	switch (type) {
	605	case TRACE_REG_REGISTER:
	606	return reg_event_syscall_enter(event);
	607	case TRACE_REG_UNREGISTER:
	608	unreg_event_syscall_enter(event);
	609	return 0;
	610
	611	#ifdef CONFIG_PERF_EVENTS
	612	case TRACE_REG_PERF_REGISTER:
	613	return perf_sysenter_enable(event);
	614	case TRACE_REG_PERF_UNREGISTER:
	615	perf_sysenter_disable(event);
	616	return 0;
	617	#endif
	618	}
	619	return 0;
	620	}
	621
	622	static int syscall_exit_register(struct ftrace_event_call *event,
	623	enum trace_reg type)
	624	{
	625	switch (type) {
	626	case TRACE_REG_REGISTER:
	627	return reg_event_syscall_exit(event);
	628	case TRACE_REG_UNREGISTER:
	629	unreg_event_syscall_exit(event);
	630	return 0;
	631
	632	#ifdef CONFIG_PERF_EVENTS
	633	case TRACE_REG_PERF_REGISTER:
	634	return perf_sysexit_enable(event);
	635	case TRACE_REG_PERF_UNREGISTER:
	636	perf_sysexit_disable(event);
	637	return 0;
	638	#endif
	639	}
	640	return 0;
	641	}