Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1620673

[for-next][PATCH 7/7] tracing: Make sure rcu_irq_enter() can work for trace_*_rcuidle() trace events

From Steven Rostedt <rostedt@goodmis.org>
Newsgroups linux.kernel
Subject [for-next][PATCH 7/7] tracing: Make sure rcu_irq_enter() can work for trace_*_rcuidle() trace events
Date 2017-04-10 22:00 +0200
Message-ID <tuNDI-5ev-13@gated-at.bofh.it> (permalink)
References <tuNDI-5ev-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


From: "Steven Rostedt (VMware)" <rostedt@goodmis.org>

Stack tracing discovered that there's a small location inside the RCU
infrastructure where calling rcu_irq_enter() does not work. As trace events
use rcu_irq_enter() it must make sure that it is functionable. A check
against rcu_irq_enter_disabled() is added with a WARN_ON_ONCE() as no trace
event should ever be used in that part of RCU. If the warning is triggered,
then the trace event is ignored.

Restructure the __DO_TRACE() a bit to get rid of the prercu and postrcu,
and just have an rcucheck that does the work from within the _DO_TRACE()
macro. gcc optimization will compile out the rcucheck=0 case.

Link: http://lkml.kernel.org/r/20170405093207.404f8deb@gandalf.local.home

Acked-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
---
 include/linux/tracepoint.h | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h
index f72fcfe0e66a..cc48cb2ce209 100644
--- a/include/linux/tracepoint.h
+++ b/include/linux/tracepoint.h
@@ -128,7 +128,7 @@ extern void syscall_unregfunc(void);
  * as "(void *, void)". The DECLARE_TRACE_NOARGS() will pass in just
  * "void *data", where as the DECLARE_TRACE() will pass in "void *data, proto".
  */
-#define __DO_TRACE(tp, proto, args, cond, prercu, postrcu)		\
+#define __DO_TRACE(tp, proto, args, cond, rcucheck)			\
 	do {								\
 		struct tracepoint_func *it_func_ptr;			\
 		void *it_func;						\
@@ -136,7 +136,11 @@ extern void syscall_unregfunc(void);
 									\
 		if (!(cond))						\
 			return;						\
-		prercu;							\
+		if (rcucheck) {						\
+			if (WARN_ON_ONCE(rcu_irq_enter_disabled()))	\
+				return;					\
+			rcu_irq_enter_irqson();				\
+		}							\
 		rcu_read_lock_sched_notrace();				\
 		it_func_ptr = rcu_dereference_sched((tp)->funcs);	\
 		if (it_func_ptr) {					\
@@ -147,20 +151,19 @@ extern void syscall_unregfunc(void);
 			} while ((++it_func_ptr)->func);		\
 		}							\
 		rcu_read_unlock_sched_notrace();			\
-		postrcu;						\
+		if (rcucheck)						\
+			rcu_irq_exit_irqson();				\
 	} while (0)
 
 #ifndef MODULE
-#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)	\
+#define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args) \
 	static inline void trace_##name##_rcuidle(proto)		\
 	{								\
 		if (static_key_false(&__tracepoint_##name.key))		\
 			__DO_TRACE(&__tracepoint_##name,		\
 				TP_PROTO(data_proto),			\
 				TP_ARGS(data_args),			\
-				TP_CONDITION(cond),			\
-				rcu_irq_enter_irqson(),			\
-				rcu_irq_exit_irqson());			\
+				TP_CONDITION(cond), 1);			\
 	}
 #else
 #define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args)
@@ -186,7 +189,7 @@ extern void syscall_unregfunc(void);
 			__DO_TRACE(&__tracepoint_##name,		\
 				TP_PROTO(data_proto),			\
 				TP_ARGS(data_args),			\
-				TP_CONDITION(cond),,);			\
+				TP_CONDITION(cond), 0);			\
 		if (IS_ENABLED(CONFIG_LOCKDEP) && (cond)) {		\
 			rcu_read_lock_sched_notrace();			\
 			rcu_dereference_sched(__tracepoint_##name.funcs);\
-- 
2.10.2

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[for-next][PATCH 0/7] tracing: Stack tracing and rcu modification updates Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 6/7] rcu/tracing: Add rcu_disabled to denote when rcu_irq_enter() will not  work Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 4/7] tracing: Rename trace_active to disable_stack_tracer and inline its  modification Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
    Re: [for-next][PATCH 4/7] tracing: Rename trace_active to  disable_stack_tracer and inline its modification Valentin Rothberg <valentinrothberg@gmail.com> - 2017-04-13 10:00 +0200
      Re: [for-next][PATCH 4/7] tracing: Rename trace_active to  disable_stack_tracer and inline its modification Steven Rostedt <rostedt@goodmis.org> - 2017-04-13 16:10 +0200
  [for-next][PATCH 7/7] tracing: Make sure rcu_irq_enter() can work for trace_*_rcuidle()  trace events Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 5/7] rcu: Fix dyntick-idle tracing Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 1/7] ftrace: Add use of synchronize_rcu_tasks() with dynamic trampolines Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 3/7] tracing: Add stack_tracer_disable/enable() functions Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200
  [for-next][PATCH 2/7] tracing: Replace the per_cpu() with __this_cpu*() in trace_stack.c Steven Rostedt <rostedt@goodmis.org> - 2017-04-10 22:00 +0200

csiph-web