perf_counter: Rework the perf counter disable/enable

The current disable/enable mechanism is:

	token = hw_perf_save_disable();
	...
	/* do bits */
	...
	hw_perf_restore(token);

This works well, provided that the use nests properly. Except we don't.

x86 NMI/INT throttling has non-nested use of this, breaking things. Therefore
provide a reference counter disable/enable interface, where the first disable
disables the hardware, and the last enable enables the hardware again.

[ Impact: refactor, simplify the PMU disable/enable logic ]

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h
index 614f921..e543ecc 100644
--- a/include/linux/perf_counter.h
+++ b/include/linux/perf_counter.h
@@ -544,8 +544,10 @@
 extern void perf_counter_do_pending(void);
 extern void perf_counter_print_debug(void);
 extern void perf_counter_unthrottle(void);
-extern u64 hw_perf_save_disable(void);
-extern void hw_perf_restore(u64 ctrl);
+extern void __perf_disable(void);
+extern bool __perf_enable(void);
+extern void perf_disable(void);
+extern void perf_enable(void);
 extern int perf_counter_task_disable(void);
 extern int perf_counter_task_enable(void);
 extern int hw_perf_group_sched_in(struct perf_counter *group_leader,
@@ -600,8 +602,8 @@
 static inline void perf_counter_do_pending(void)			{ }
 static inline void perf_counter_print_debug(void)			{ }
 static inline void perf_counter_unthrottle(void)			{ }
-static inline void hw_perf_restore(u64 ctrl)				{ }
-static inline u64 hw_perf_save_disable(void)		      { return 0; }
+static inline void perf_disable(void)					{ }
+static inline void perf_enable(void)					{ }
 static inline int perf_counter_task_disable(void)	{ return -EINVAL; }
 static inline int perf_counter_task_enable(void)	{ return -EINVAL; }