Commit fb346fd9 authored by Waiman Long's avatar Waiman Long Committed by Ingo Molnar
Browse files

locking/lock_events: Make lock_events available for all archs & other locks 



The QUEUED_LOCK_STAT option to report queued spinlocks event counts
was previously allowed only on x86 architecture. To make the locking
event counting code more useful, it is now renamed to a more generic
LOCK_EVENT_COUNTS config option. This new option will be available to
all the architectures that use qspinlock at the moment.

Other locking code can now start to use the generic locking event
counting code by including lock_events.h and put the new locking event
names into the lock_events_list.h header file.

My experience with lock event counting is that it gives valuable insight
on how the locking code works and what can be done to make it better. I
would like to extend this benefit to other locking code like mutex and
rwsem in the near future.

The PV qspinlock specific code will stay in qspinlock_stat.h. The
locking event counters will now reside in the <debugfs>/lock_event_counts
directory.

Signed-off-by: default avatarWaiman Long <longman@redhat.com>
Acked-by: default avatarPeter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: default avatarDavidlohr Bueso <dbueso@suse.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/20190404174320.22416-9-longman@redhat.com


Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
parent ad53fa10

arch/Kconfig

+10 −0
Original line number Diff line number Diff line
@@ -901,6 +901,16 @@ config HAVE_ARCH_PREL32_RELOCATIONS 
config ARCH_USE_MEMREMAP_PROT

	bool



config LOCK_EVENT_COUNTS

	bool "Locking event counts collection"

	depends on DEBUG_FS

	depends on QUEUED_SPINLOCKS

	---help---

	  Enable light-weight counting of various locking related events

	  in the system with minimal performance impact. This reduces

	  the chance of application behavior change because of timing

	  differences. The counts are reported via debugfs.



source "kernel/gcov/Kconfig"



source "scripts/gcc-plugins/Kconfig"

arch/x86/Kconfig

+0 −8
Original line number Diff line number Diff line
@@ -780,14 +780,6 @@ config PARAVIRT_SPINLOCKS 


	  If you are unsure how to answer this question, answer Y.



config QUEUED_LOCK_STAT

	bool "Paravirt queued spinlock statistics"

	depends on PARAVIRT_SPINLOCKS && DEBUG_FS

	---help---

	  Enable the collection of statistical data on the slowpath

	  behavior of paravirtualized queued spinlocks and report

	  them on debugfs.



source "arch/x86/xen/Kconfig"



config KVM_GUEST

kernel/locking/Makefile

+1 −0
Original line number Diff line number Diff line
@@ -28,3 +28,4 @@ obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o 
obj-$(CONFIG_QUEUED_RWLOCKS) += qrwlock.o

obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o

obj-$(CONFIG_WW_MUTEX_SELFTEST) += test-ww_mutex.o

obj-$(CONFIG_LOCK_EVENT_COUNTS) += lock_events.o

kernel/locking/lock_events.c

0 → 100644
+153 −0
Original line number Diff line number Diff line 
/* SPDX-License-Identifier: GPL-2.0 */

/*

 * This program is free software; you can redistribute it and/or modify

 * it under the terms of the GNU General Public License as published by

 * the Free Software Foundation; either version 2 of the License, or

 * (at your option) any later version.

 *

 * This program is distributed in the hope that it will be useful,

 * but WITHOUT ANY WARRANTY; without even the implied warranty of

 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

 * GNU General Public License for more details.

 *

 * Authors: Waiman Long <waiman.long@hpe.com>

 */



/*

 * Collect locking event counts

 */

#include <linux/debugfs.h>

#include <linux/sched.h>

#include <linux/sched/clock.h>

#include <linux/fs.h>



#include "lock_events.h"



#undef  LOCK_EVENT

#define LOCK_EVENT(name)	[LOCKEVENT_ ## name] = #name,



#define LOCK_EVENTS_DIR		"lock_event_counts"



/*

 * When CONFIG_LOCK_EVENT_COUNTS is enabled, event counts of different

 * types of locks will be reported under the <debugfs>/lock_event_counts/

 * directory. See lock_events_list.h for the list of available locking

 * events.

 *

 * Writing to the special ".reset_counts" file will reset all the above

 * locking event counts. This is a very slow operation and so should not

 * be done frequently.

 *

 * These event counts are implemented as per-cpu variables which are

 * summed and computed whenever the corresponding debugfs files are read. This

 * minimizes added overhead making the counts usable even in a production

 * environment.

 */

static const char * const lockevent_names[lockevent_num + 1] = {



#include "lock_events_list.h"



	[LOCKEVENT_reset_cnts] = ".reset_counts",

};



/*

 * Per-cpu counts

 */

DEFINE_PER_CPU(unsigned long, lockevents[lockevent_num]);



/*

 * The lockevent_read() function can be overridden.

 */

ssize_t __weak lockevent_read(struct file *file, char __user *user_buf,

			      size_t count, loff_t *ppos)

{

	char buf[64];

	int cpu, id, len;

	u64 sum = 0;



	/*

	 * Get the counter ID stored in file->f_inode->i_private

	 */

	id = (long)file_inode(file)->i_private;



	if (id >= lockevent_num)

		return -EBADF;



	for_each_possible_cpu(cpu)

		sum += per_cpu(lockevents[id], cpu);

	len = snprintf(buf, sizeof(buf) - 1, "%llu\n", sum);



	return simple_read_from_buffer(user_buf, count, ppos, buf, len);

}



/*

 * Function to handle write request

 *

 * When idx = reset_cnts, reset all the counts.

 */

static ssize_t lockevent_write(struct file *file, const char __user *user_buf,

			   size_t count, loff_t *ppos)

{

	int cpu;



	/*

	 * Get the counter ID stored in file->f_inode->i_private

	 */

	if ((long)file_inode(file)->i_private != LOCKEVENT_reset_cnts)

		return count;



	for_each_possible_cpu(cpu) {

		int i;

		unsigned long *ptr = per_cpu_ptr(lockevents, cpu);



		for (i = 0 ; i < lockevent_num; i++)

			WRITE_ONCE(ptr[i], 0);

	}

	return count;

}



/*

 * Debugfs data structures

 */

static const struct file_operations fops_lockevent = {

	.read = lockevent_read,

	.write = lockevent_write,

	.llseek = default_llseek,

};



/*

 * Initialize debugfs for the locking event counts.

 */

static int __init init_lockevent_counts(void)

{

	struct dentry *d_counts = debugfs_create_dir(LOCK_EVENTS_DIR, NULL);

	int i;



	if (!d_counts)

		goto out;



	/*

	 * Create the debugfs files

	 *

	 * As reading from and writing to the stat files can be slow, only

	 * root is allowed to do the read/write to limit impact to system

	 * performance.

	 */

	for (i = 0; i < lockevent_num; i++)

		if (!debugfs_create_file(lockevent_names[i], 0400, d_counts,

					 (void *)(long)i, &fops_lockevent))

			goto fail_undo;



	if (!debugfs_create_file(lockevent_names[LOCKEVENT_reset_cnts], 0200,

				 d_counts, (void *)(long)LOCKEVENT_reset_cnts,

				 &fops_lockevent))

		goto fail_undo;



	return 0;

fail_undo:

	debugfs_remove_recursive(d_counts);

out:

	pr_warn("Could not create '%s' debugfs entries\n", LOCK_EVENTS_DIR);

	return -ENOMEM;

}

fs_initcall(init_lockevent_counts);

kernel/locking/lock_events.h

+7 −3
Original line number Diff line number Diff line
@@ -13,6 +13,9 @@ 
 * Authors: Waiman Long <longman@redhat.com>

 */



#ifndef __LOCKING_LOCK_EVENTS_H

#define __LOCKING_LOCK_EVENTS_H



enum lock_events {



#include "lock_events_list.h"
@@ -21,7 +24,7 @@ enum lock_events { 
	LOCKEVENT_reset_cnts = lockevent_num,

};



#ifdef CONFIG_QUEUED_LOCK_STAT

#ifdef CONFIG_LOCK_EVENT_COUNTS

/*

 * Per-cpu counters

 */
@@ -46,10 +49,11 @@ static inline void __lockevent_add(enum lock_events event, int inc) 


#define lockevent_add(ev, c)	__lockevent_add(LOCKEVENT_ ##ev, c)



#else  /* CONFIG_QUEUED_LOCK_STAT */

#else  /* CONFIG_LOCK_EVENT_COUNTS */



#define lockevent_inc(ev)

#define lockevent_add(ev, c)

#define lockevent_cond_inc(ev, c)



#endif /* CONFIG_QUEUED_LOCK_STAT */
 
#endif /* CONFIG_LOCK_EVENT_COUNTS */

#endif /* __LOCKING_LOCK_EVENTS_H */