linux/kernel/trace/trace_event_perf.c

385 lines
8.8 KiB
C
Raw Normal View History

/*
* trace event based perf event profiling/tracing
*
* Copyright (C) 2009 Red Hat Inc, Peter Zijlstra <pzijlstr@redhat.com>
perf: Take a hot regs snapshot for trace events We are taking a wrong regs snapshot when a trace event triggers. Either we use get_irq_regs(), which gives us the interrupted registers if we are in an interrupt, or we use task_pt_regs() which gives us the state before we entered the kernel, assuming we are lucky enough to be no kernel thread, in which case task_pt_regs() returns the initial set of regs when the kernel thread was started. What we want is different. We need a hot snapshot of the regs, so that we can get the instruction pointer to record in the sample, the frame pointer for the callchain, and some other things. Let's use the new perf_fetch_caller_regs() for that. Comparison with perf record -e lock: -R -a -f -g Before: perf [kernel] [k] __do_softirq | --- __do_softirq | |--55.16%-- __open | --44.84%-- __write_nocancel After: perf [kernel] [k] perf_tp_event | --- perf_tp_event | |--41.07%-- lock_acquire | | | |--39.36%-- _raw_spin_lock | | | | | |--7.81%-- hrtimer_interrupt | | | smp_apic_timer_interrupt | | | apic_timer_interrupt The old case was producing unreliable callchains. Now having right frame and instruction pointers, we have the trace we want. Also syscalls and kprobe events already have the right regs, let's use them instead of wasting a retrieval. v2: Follow the rename perf_save_regs() -> perf_fetch_caller_regs() Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Archs <linux-arch@vger.kernel.org>
2010-03-03 09:16:16 +03:00
* Copyright (C) 2009-2010 Frederic Weisbecker <fweisbec@gmail.com>
*/
#include <linux/module.h>
#include <linux/kprobes.h>
#include "trace.h"
static char __percpu *perf_trace_buf[PERF_NR_CONTEXTS];
/*
* Force it to be aligned to unsigned long to avoid misaligned accesses
* suprises
*/
typedef typeof(unsigned long [PERF_MAX_TRACE_SIZE / sizeof(unsigned long)])
perf_trace_t;
/* Count the events in use (per event id, not per instance) */
static int total_ref_count;
static int perf_trace_event_perm(struct ftrace_event_call *tp_event,
struct perf_event *p_event)
{
if (tp_event->perf_perm) {
int ret = tp_event->perf_perm(tp_event, p_event);
if (ret)
return ret;
}
/*
* We checked and allowed to create parent,
* allow children without checking.
*/
if (p_event->parent)
return 0;
/*
* It's ok to check current process (owner) permissions in here,
* because code below is called only via perf_event_open syscall.
*/
/* The ftrace function trace is allowed only for root. */
if (ftrace_event_is_function(tp_event)) {
if (perf_paranoid_tracepoint_raw() && !capable(CAP_SYS_ADMIN))
return -EPERM;
/*
* We don't allow user space callchains for function trace
* event, due to issues with page faults while tracing page
* fault handler and its overall trickiness nature.
*/
if (!p_event->attr.exclude_callchain_user)
return -EINVAL;
/*
* Same reason to disable user stack dump as for user space
* callchains above.
*/
if (p_event->attr.sample_type & PERF_SAMPLE_STACK_USER)
return -EINVAL;
}
/* No tracing, just counting, so no obvious leak */
if (!(p_event->attr.sample_type & PERF_SAMPLE_RAW))
return 0;
/* Some events are ok to be traced by non-root users... */
if (p_event->attach_state == PERF_ATTACH_TASK) {
if (tp_event->flags & TRACE_EVENT_FL_CAP_ANY)
return 0;
}
/*
* ...otherwise raw tracepoint data can be a severe data leak,
* only allow root to have these.
*/
if (perf_paranoid_tracepoint_raw() && !capable(CAP_SYS_ADMIN))
return -EPERM;
return 0;
}
static int perf_trace_event_reg(struct ftrace_event_call *tp_event,
struct perf_event *p_event)
{
struct hlist_head __percpu *list;
int ret = -ENOMEM;
int cpu;
p_event->tp_event = tp_event;
if (tp_event->perf_refcount++ > 0)
return 0;
list = alloc_percpu(struct hlist_head);
if (!list)
goto fail;
for_each_possible_cpu(cpu)
INIT_HLIST_HEAD(per_cpu_ptr(list, cpu));
tp_event->perf_events = list;
if (!total_ref_count) {
char __percpu *buf;
int i;
for (i = 0; i < PERF_NR_CONTEXTS; i++) {
buf = (char __percpu *)alloc_percpu(perf_trace_t);
if (!buf)
goto fail;
perf_trace_buf[i] = buf;
}
}
ret = tp_event->class->reg(tp_event, TRACE_REG_PERF_REGISTER, NULL);
if (ret)
goto fail;
total_ref_count++;
return 0;
fail:
if (!total_ref_count) {
int i;
for (i = 0; i < PERF_NR_CONTEXTS; i++) {
free_percpu(perf_trace_buf[i]);
perf_trace_buf[i] = NULL;
}
}
if (!--tp_event->perf_refcount) {
free_percpu(tp_event->perf_events);
tp_event->perf_events = NULL;
}
return ret;
}
static void perf_trace_event_unreg(struct perf_event *p_event)
{
struct ftrace_event_call *tp_event = p_event->tp_event;
int i;
if (--tp_event->perf_refcount > 0)
goto out;
tp_event->class->reg(tp_event, TRACE_REG_PERF_UNREGISTER, NULL);
/*
* Ensure our callback won't be called anymore. The buffers
* will be freed after that.
*/
tracepoint_synchronize_unregister();
free_percpu(tp_event->perf_events);
tp_event->perf_events = NULL;
if (!--total_ref_count) {
for (i = 0; i < PERF_NR_CONTEXTS; i++) {
free_percpu(perf_trace_buf[i]);
perf_trace_buf[i] = NULL;
}
}
out:
module_put(tp_event->mod);
}
static int perf_trace_event_open(struct perf_event *p_event)
{
struct ftrace_event_call *tp_event = p_event->tp_event;
return tp_event->class->reg(tp_event, TRACE_REG_PERF_OPEN, p_event);
}
static void perf_trace_event_close(struct perf_event *p_event)
{
struct ftrace_event_call *tp_event = p_event->tp_event;
tp_event->class->reg(tp_event, TRACE_REG_PERF_CLOSE, p_event);
}
static int perf_trace_event_init(struct ftrace_event_call *tp_event,
struct perf_event *p_event)
{
int ret;
ret = perf_trace_event_perm(tp_event, p_event);
if (ret)
return ret;
ret = perf_trace_event_reg(tp_event, p_event);
if (ret)
return ret;
ret = perf_trace_event_open(p_event);
if (ret) {
perf_trace_event_unreg(p_event);
return ret;
}
return 0;
}
int perf_trace_init(struct perf_event *p_event)
{
struct ftrace_event_call *tp_event;
u64 event_id = p_event->attr.config;
int ret = -EINVAL;
mutex_lock(&event_mutex);
list_for_each_entry(tp_event, &ftrace_events, list) {
if (tp_event->event.type == event_id &&
tp_event->class && tp_event->class->reg &&
try_module_get(tp_event->mod)) {
ret = perf_trace_event_init(tp_event, p_event);
if (ret)
module_put(tp_event->mod);
break;
}
}
mutex_unlock(&event_mutex);
return ret;
}
void perf_trace_destroy(struct perf_event *p_event)
{
mutex_lock(&event_mutex);
perf_trace_event_close(p_event);
perf_trace_event_unreg(p_event);
mutex_unlock(&event_mutex);
}
2010-06-16 16:37:10 +04:00
int perf_trace_add(struct perf_event *p_event, int flags)
{
struct ftrace_event_call *tp_event = p_event->tp_event;
struct hlist_head __percpu *pcpu_list;
struct hlist_head *list;
pcpu_list = tp_event->perf_events;
if (WARN_ON_ONCE(!pcpu_list))
return -EINVAL;
2010-06-16 16:37:10 +04:00
if (!(flags & PERF_EF_START))
p_event->hw.state = PERF_HES_STOPPED;
list = this_cpu_ptr(pcpu_list);
hlist_add_head_rcu(&p_event->hlist_entry, list);
return tp_event->class->reg(tp_event, TRACE_REG_PERF_ADD, p_event);
}
2010-06-16 16:37:10 +04:00
void perf_trace_del(struct perf_event *p_event, int flags)
{
struct ftrace_event_call *tp_event = p_event->tp_event;
hlist_del_rcu(&p_event->hlist_entry);
tp_event->class->reg(tp_event, TRACE_REG_PERF_DEL, p_event);
}
void *perf_trace_buf_prepare(int size, unsigned short type,
perf: Avoid horrible stack usage Both Linus (most recent) and Steve (a while ago) reported that perf related callbacks have massive stack bloat. The problem is that software events need a pt_regs in order to properly report the event location and unwind stack. And because we could not assume one was present we allocated one on stack and filled it with minimal bits required for operation. Now, pt_regs is quite large, so this is undesirable. Furthermore it turns out that most sites actually have a pt_regs pointer available, making this even more onerous, as the stack space is pointless waste. This patch addresses the problem by observing that software events have well defined nesting semantics, therefore we can use static per-cpu storage instead of on-stack. Linus made the further observation that all but the scheduler callers of perf_sw_event() have a pt_regs available, so we change the regular perf_sw_event() to require a valid pt_regs (where it used to be optional) and add perf_sw_event_sched() for the scheduler. We have a scheduler specific call instead of a more generic _noregs() like construct because we can assume non-recursion from the scheduler and thereby simplify the code further (_noregs would have to put the recursion context call inline in order to assertain which __perf_regs element to use). One last note on the implementation of perf_trace_buf_prepare(); we allow .regs = NULL for those cases where we already have a pt_regs pointer available and do not need another. Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Reported-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Javi Merino <javi.merino@arm.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Petr Mladek <pmladek@suse.cz> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Tom Zanussi <tom.zanussi@linux.intel.com> Cc: Vaibhav Nagarnaik <vnagarnaik@google.com> Link: http://lkml.kernel.org/r/20141216115041.GW3337@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-12-16 14:47:34 +03:00
struct pt_regs **regs, int *rctxp)
{
struct trace_entry *entry;
unsigned long flags;
char *raw_data;
int pc;
BUILD_BUG_ON(PERF_MAX_TRACE_SIZE % sizeof(unsigned long));
if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE,
"perf buffer not large enough"))
return NULL;
pc = preempt_count();
*rctxp = perf_swevent_get_recursion_context();
if (*rctxp < 0)
return NULL;
perf: Avoid horrible stack usage Both Linus (most recent) and Steve (a while ago) reported that perf related callbacks have massive stack bloat. The problem is that software events need a pt_regs in order to properly report the event location and unwind stack. And because we could not assume one was present we allocated one on stack and filled it with minimal bits required for operation. Now, pt_regs is quite large, so this is undesirable. Furthermore it turns out that most sites actually have a pt_regs pointer available, making this even more onerous, as the stack space is pointless waste. This patch addresses the problem by observing that software events have well defined nesting semantics, therefore we can use static per-cpu storage instead of on-stack. Linus made the further observation that all but the scheduler callers of perf_sw_event() have a pt_regs available, so we change the regular perf_sw_event() to require a valid pt_regs (where it used to be optional) and add perf_sw_event_sched() for the scheduler. We have a scheduler specific call instead of a more generic _noregs() like construct because we can assume non-recursion from the scheduler and thereby simplify the code further (_noregs would have to put the recursion context call inline in order to assertain which __perf_regs element to use). One last note on the implementation of perf_trace_buf_prepare(); we allow .regs = NULL for those cases where we already have a pt_regs pointer available and do not need another. Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Reported-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Javi Merino <javi.merino@arm.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Petr Mladek <pmladek@suse.cz> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Tom Zanussi <tom.zanussi@linux.intel.com> Cc: Vaibhav Nagarnaik <vnagarnaik@google.com> Link: http://lkml.kernel.org/r/20141216115041.GW3337@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-12-16 14:47:34 +03:00
if (regs)
*regs = this_cpu_ptr(&__perf_regs[*rctxp]);
raw_data = this_cpu_ptr(perf_trace_buf[*rctxp]);
/* zero the dead bytes from align to not leak stack to user */
memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64));
entry = (struct trace_entry *)raw_data;
local_save_flags(flags);
tracing_generic_entry_update(entry, flags, pc);
entry->type = type;
return raw_data;
}
EXPORT_SYMBOL_GPL(perf_trace_buf_prepare);
NOKPROBE_SYMBOL(perf_trace_buf_prepare);
#ifdef CONFIG_FUNCTION_TRACER
static void
perf_ftrace_function_call(unsigned long ip, unsigned long parent_ip,
struct ftrace_ops *ops, struct pt_regs *pt_regs)
{
struct ftrace_entry *entry;
struct hlist_head *head;
struct pt_regs regs;
int rctx;
head = this_cpu_ptr(event_function.perf_events);
if (hlist_empty(head))
return;
#define ENTRY_SIZE (ALIGN(sizeof(struct ftrace_entry) + sizeof(u32), \
sizeof(u64)) - sizeof(u32))
BUILD_BUG_ON(ENTRY_SIZE > PERF_MAX_TRACE_SIZE);
perf_fetch_caller_regs(&regs);
entry = perf_trace_buf_prepare(ENTRY_SIZE, TRACE_FN, NULL, &rctx);
if (!entry)
return;
entry->ip = ip;
entry->parent_ip = parent_ip;
perf_trace_buf_submit(entry, ENTRY_SIZE, rctx, 0,
1, &regs, head, NULL);
#undef ENTRY_SIZE
}
static int perf_ftrace_function_register(struct perf_event *event)
{
struct ftrace_ops *ops = &event->ftrace_ops;
ops->flags |= FTRACE_OPS_FL_CONTROL;
ops->func = perf_ftrace_function_call;
return register_ftrace_function(ops);
}
static int perf_ftrace_function_unregister(struct perf_event *event)
{
struct ftrace_ops *ops = &event->ftrace_ops;
ftrace, perf: Add filter support for function trace event Adding support to filter function trace event via perf interface. It is now possible to use filter interface in the perf tool like: perf record -e ftrace:function --filter="(ip == mm_*)" ls The filter syntax is restricted to the the 'ip' field only, and following operators are accepted '==' '!=' '||', ending up with the filter strings like: ip == f1[, ]f2 ... || ip != f3[, ]f4 ... with comma ',' or space ' ' as a function separator. If the space ' ' is used as a separator, the right side of the assignment needs to be enclosed in double quotes '"', e.g.: perf record -e ftrace:function --filter '(ip == do_execve,sys_*,ext*)' ls perf record -e ftrace:function --filter '(ip == "do_execve,sys_*,ext*")' ls perf record -e ftrace:function --filter '(ip == "do_execve sys_* ext*")' ls The '==' operator adds trace filter with same effect as would be added via set_ftrace_filter file. The '!=' operator adds trace filter with same effect as would be added via set_ftrace_notrace file. The right side of the '!=', '==' operators is list of functions or regexp. to be added to filter separated by space. The '||' operator is used for connecting multiple filter definitions together. It is possible to have more than one '==' and '!=' operators within one filter string. Link: http://lkml.kernel.org/r/1329317514-8131-8-git-send-email-jolsa@redhat.com Signed-off-by: Jiri Olsa <jolsa@redhat.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2012-02-15 18:51:54 +04:00
int ret = unregister_ftrace_function(ops);
ftrace_free_filter(ops);
return ret;
}
static void perf_ftrace_function_enable(struct perf_event *event)
{
ftrace_function_local_enable(&event->ftrace_ops);
}
static void perf_ftrace_function_disable(struct perf_event *event)
{
ftrace_function_local_disable(&event->ftrace_ops);
}
int perf_ftrace_event_register(struct ftrace_event_call *call,
enum trace_reg type, void *data)
{
switch (type) {
case TRACE_REG_REGISTER:
case TRACE_REG_UNREGISTER:
break;
case TRACE_REG_PERF_REGISTER:
case TRACE_REG_PERF_UNREGISTER:
return 0;
case TRACE_REG_PERF_OPEN:
return perf_ftrace_function_register(data);
case TRACE_REG_PERF_CLOSE:
return perf_ftrace_function_unregister(data);
case TRACE_REG_PERF_ADD:
perf_ftrace_function_enable(data);
return 0;
case TRACE_REG_PERF_DEL:
perf_ftrace_function_disable(data);
return 0;
}
return -EINVAL;
}
#endif /* CONFIG_FUNCTION_TRACER */