perf/trace: Add ability to set a target task for events

A few events are interesting not only for a current task.
For example, sched_stat_* events are interesting for a task
which wakes up. For this reason, it will be good if such
events will be delivered to a target task too.

Now a target task can be set by using __perf_task().

The original idea and a draft patch belongs to Peter Zijlstra.

I need these events for profiling sleep times. sched_switch is used for
getting callchains and sched_stat_* is used for getting time periods.
These events are combined in user space, then it can be analyzed by
perf tools.
Inspired-by: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Arun Sharma <asharma@fb.com>
Signed-off-by: Andrew Vagin <avagin@openvz.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1342016098-213063-1-git-send-email-avagin@openvz.orgSigned-off-by: Ingo Molnar <mingo@kernel.org>

include/linux/ftrace_event.h

@@ -306,9 +306,10 @@ extern void *perf_trace_buf_prepare(int size, unsigned short type,
 
 static inline void
 perf_trace_buf_submit(void *raw_data, int size, int rctx, u64 addr,
-		       u64 count, struct pt_regs *regs, void *head)
+		       u64 count, struct pt_regs *regs, void *head,
+		       struct task_struct *task)
 {
-	perf_tp_event(addr, count, raw_data, size, regs, head, rctx);
+	perf_tp_event(addr, count, raw_data, size, regs, head, rctx, task);
 }
 #endif
 

include/linux/perf_event.h

@@ -1272,7 +1272,8 @@ static inline bool perf_paranoid_kernel(void)
 extern void perf_event_init(void);
 extern void perf_tp_event(u64 addr, u64 count, void *record,
 			  int entry_size, struct pt_regs *regs,
-			  struct hlist_head *head, int rctx);
+			  struct hlist_head *head, int rctx,
+			  struct task_struct *task);
 extern void perf_bp_event(struct perf_event *event, void *data);
 
 #ifndef perf_misc_flags

include/trace/events/sched.h

@@ -73,6 +73,9 @@ DECLARE_EVENT_CLASS(sched_wakeup_template,
 		__entry->prio		= p->prio;
 		__entry->success	= success;
 		__entry->target_cpu	= task_cpu(p);
+	)
+	TP_perf_assign(
+		__perf_task(p);
 	),
 
 	TP_printk("comm=%s pid=%d prio=%d success=%d target_cpu=%03d",
@@ -325,6 +328,7 @@ DECLARE_EVENT_CLASS(sched_stat_template,
 	)
 	TP_perf_assign(
 		__perf_count(delay);
+		__perf_task(tsk);
 	),
 
 	TP_printk("comm=%s pid=%d delay=%Lu [ns]",

include/trace/ftrace.h

@@ -712,6 +712,9 @@ __attribute__((section("_ftrace_events"))) *__event_##call = &event_##call
 #undef __perf_count
 #define __perf_count(c) __count = (c)
 
+#undef __perf_task
+#define __perf_task(t) __task = (t)
+
 #undef TP_perf_assign
 #define TP_perf_assign(args...) args
 
@@ -725,6 +728,7 @@ perf_trace_##call(void *__data, proto)					\
 	struct ftrace_raw_##call *entry;				\
 	struct pt_regs __regs;						\
 	u64 __addr = 0, __count = 1;					\
+	struct task_struct *__task = NULL;				\
 	struct hlist_head *head;					\
 	int __entry_size;						\
 	int __data_size;						\
@@ -752,7 +756,7 @@ perf_trace_##call(void *__data, proto)					\
 									\
 	head = this_cpu_ptr(event_call->perf_events);			\
 	perf_trace_buf_submit(entry, __entry_size, rctx, __addr,	\
-		__count, &__regs, head);				\
+		__count, &__regs, head, __task);			\
 }
 
 /*

kernel/events/callchain.c

@@ -153,7 +153,8 @@ put_callchain_entry(int rctx)
 	put_recursion_context(__get_cpu_var(callchain_recursion), rctx);
 }
 
-struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
+struct perf_callchain_entry *
+perf_callchain(struct perf_event *event, struct pt_regs *regs)
 {
 	int rctx;
 	struct perf_callchain_entry *entry;
@@ -178,6 +179,12 @@ struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
 	}
 
 	if (regs) {
+		/*
+		 * Disallow cross-task user callchains.
+		 */
+		if (event->ctx->task && event->ctx->task != current)
+			goto exit_put;
+
 		perf_callchain_store(entry, PERF_CONTEXT_USER);
 		perf_callchain_user(entry, regs);
 	}

kernel/events/core.c

@@ -4039,7 +4039,7 @@ void perf_prepare_sample(struct perf_event_header *header,
 	if (sample_type & PERF_SAMPLE_CALLCHAIN) {
 		int size = 1;
 
-		data->callchain = perf_callchain(regs);
+		data->callchain = perf_callchain(event, regs);
 
 		if (data->callchain)
 			size += data->callchain->nr;
@@ -5209,7 +5209,8 @@ static int perf_tp_event_match(struct perf_event *event,
 }
 
 void perf_tp_event(u64 addr, u64 count, void *record, int entry_size,
-		   struct pt_regs *regs, struct hlist_head *head, int rctx)
+		   struct pt_regs *regs, struct hlist_head *head, int rctx,
+		   struct task_struct *task)
 {
 	struct perf_sample_data data;
 	struct perf_event *event;
@@ -5228,6 +5229,31 @@ void perf_tp_event(u64 addr, u64 count, void *record, int entry_size,
 			perf_swevent_event(event, count, &data, regs);
 	}
 
+	/*
+	 * If we got specified a target task, also iterate its context and
+	 * deliver this event there too.
+	 */
+	if (task && task != current) {
+		struct perf_event_context *ctx;
+		struct trace_entry *entry = record;
+
+		rcu_read_lock();
+		ctx = rcu_dereference(task->perf_event_ctxp[perf_sw_context]);
+		if (!ctx)
+			goto unlock;
+
+		list_for_each_entry_rcu(event, &ctx->event_list, event_entry) {
+			if (event->attr.type != PERF_TYPE_TRACEPOINT)
+				continue;
+			if (event->attr.config != entry->type)
+				continue;
+			if (perf_tp_event_match(event, &data, regs))
+				perf_swevent_event(event, count, &data, regs);
+		}
+unlock:
+		rcu_read_unlock();
+	}
+
 	perf_swevent_put_recursion_context(rctx);
 }
 EXPORT_SYMBOL_GPL(perf_tp_event);

kernel/events/internal.h

@@ -101,7 +101,8 @@ __output_copy(struct perf_output_handle *handle,
 }
 
 /* Callchain handling */
-extern struct perf_callchain_entry *perf_callchain(struct pt_regs *regs);
+extern struct perf_callchain_entry *
+perf_callchain(struct perf_event *event, struct pt_regs *regs);
 extern int get_callchain_buffers(void);
 extern void put_callchain_buffers(void);
 

kernel/trace/trace_event_perf.c

@@ -281,7 +281,7 @@ perf_ftrace_function_call(unsigned long ip, unsigned long parent_ip)
 
 	head = this_cpu_ptr(event_function.perf_events);
 	perf_trace_buf_submit(entry, ENTRY_SIZE, rctx, 0,
-			      1, &regs, head);
+			      1, &regs, head, NULL);
 
 #undef ENTRY_SIZE
 }

kernel/trace/trace_kprobe.c

@@ -1002,7 +1002,8 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp,
 	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
 
 	head = this_cpu_ptr(call->perf_events);
-	perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head);
+	perf_trace_buf_submit(entry, size, rctx,
+					entry->ip, 1, regs, head, NULL);
 }
 
 /* Kretprobe profile handler */
@@ -1033,7 +1034,8 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri,
 	store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize);
 
 	head = this_cpu_ptr(call->perf_events);
-	perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head);
+	perf_trace_buf_submit(entry, size, rctx,
+					entry->ret_ip, 1, regs, head, NULL);
 }
 #endif	/* CONFIG_PERF_EVENTS */
 

kernel/trace/trace_syscalls.c

@@ -532,7 +532,7 @@ static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id)
 			       (unsigned long *)&rec->args);
 
 	head = this_cpu_ptr(sys_data->enter_event->perf_events);
-	perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head);
+	perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head, NULL);
 }
 
 int perf_sysenter_enable(struct ftrace_event_call *call)
@@ -608,7 +608,7 @@ static void perf_syscall_exit(void *ignore, struct pt_regs *regs, long ret)
 	rec->ret = syscall_get_return_value(current, regs);
 
 	head = this_cpu_ptr(sys_data->exit_event->perf_events);
-	perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head);
+	perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head, NULL);
 }
 
 int perf_sysexit_enable(struct ftrace_event_call *call)

kernel/trace/trace_uprobe.c

@@ -670,7 +670,7 @@ static void uprobe_perf_func(struct trace_uprobe *tu, struct pt_regs *regs)
 		call_fetch(&tu->args[i].fetch, regs, data + tu->args[i].offset);
 
 	head = this_cpu_ptr(call->perf_events);
-	perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head);
+	perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head, NULL);
 
  out:
 	preempt_enable();