Skip to contentSkip to conversation

time: Remove CONFIG_TIMER_STATS

PATCH In review Kees Cook 6 participants 6 messages Feb 8, 2017 – Feb 10, 2017

Subsystems (5)
lore
Share & tools
lore

Kees Cook

author
#0 lore raw

PATCH v2 1/1 time: Remove CONFIG_TIMER_STATS

Currently CONFIG_TIMER_STATS exposes process information across namespaces:

kernel/time/timer_list.c print_timer():

        SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);

/proc/timer_list:

#11: <0000000000000000>, hrtimer_wakeup, S:01, do_nanosleep, cron/2570

Given that the tracer can give the same information, this patch entirely removes CONFIG_TIMER_STATS.

---
v2:
- dropped doc comments for removed structure elements; thx 0-day builder.

---
 Documentation/timers/timer_stats.txt |  73 ------
 include/linux/hrtimer.h              |  11 -
 include/linux/timer.h                |  45 ----
 kernel/kthread.c                     |   1 -
 kernel/time/Makefile                 |   1 -
 kernel/time/hrtimer.c                |  38 ----
 kernel/time/timer.c                  |  48 +---
 kernel/time/timer_list.c             |  10 -
 kernel/time/timer_stats.c            | 425 -----------------------------------
 kernel/workqueue.c                   |   2 -
 lib/Kconfig.debug                    |  14 --
 11 files changed, 2 insertions(+), 666 deletions(-)
 delete mode 100644 Documentation/timers/timer_stats.txt
 delete mode 100644 kernel/time/timer_stats.c
Documentation/timers/timer_stats.txt deleted +0 -73
deleted file mode 100644
index de835ee97455..000000000000
@@ -1,73 +0,0 @@
-timer_stats - timer usage statistics
-------------------------------------
-
-timer_stats is a debugging facility to make the timer (ab)usage in a Linux
-system visible to kernel and userspace developers. If enabled in the config
-but not used it has almost zero runtime overhead, and a relatively small
-data structure overhead. Even if collection is enabled runtime all the
-locking is per-CPU and lookup is hashed.
-
-timer_stats should be used by kernel and userspace developers to verify that
-their code does not make unduly use of timers. This helps to avoid unnecessary
-wakeups, which should be avoided to optimize power consumption.
-
-It can be enabled by CONFIG_TIMER_STATS in the "Kernel hacking" configuration
-section.
-
-timer_stats collects information about the timer events which are fired in a
-Linux system over a sample period:
-
-- the pid of the task(process) which initialized the timer
-- the name of the process which initialized the timer
-- the function where the timer was initialized
-- the callback function which is associated to the timer
-- the number of events (callbacks)
-
-timer_stats adds an entry to /proc: /proc/timer_stats
-
-This entry is used to control the statistics functionality and to read out the
-sampled information.
-
-The timer_stats functionality is inactive on bootup.
-
-To activate a sample period issue:
-# echo 1 >/proc/timer_stats
-
-To stop a sample period issue:
-# echo 0 >/proc/timer_stats
-
-The statistics can be retrieved by:
-# cat /proc/timer_stats
-
-While sampling is enabled, each readout from /proc/timer_stats will see
-newly updated statistics. Once sampling is disabled, the sampled information
-is kept until a new sample period is started. This allows multiple readouts.
-
-Sample output of /proc/timer_stats:
-
-Timerstats sample period: 3.888770 s
-  12,     0 swapper          hrtimer_stop_sched_tick (hrtimer_sched_tick)
-  15,     1 swapper          hcd_submit_urb (rh_timer_func)
-   4,   959 kedac            schedule_timeout (process_timeout)
-   1,     0 swapper          page_writeback_init (wb_timer_fn)
-  28,     0 swapper          hrtimer_stop_sched_tick (hrtimer_sched_tick)
-  22,  2948 IRQ 4            tty_flip_buffer_push (delayed_work_timer_fn)
-   3,  3100 bash             schedule_timeout (process_timeout)
-   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          neigh_table_init_no_netlink (neigh_periodic_timer)
-   1,  2292 ip               __netdev_watchdog_up (dev_watchdog)
-   1,    23 events/1         do_cache_clean (delayed_work_timer_fn)
-90 total events, 30.0 events/sec
-
-The first column is the number of events, the second column the pid, the third
-column is the name of the process. The forth column shows the function which
-initialized the timer and in parenthesis the callback function which was
-executed on expiry.
-
-    Thomas, Ingo
-
-Added flag to indicate 'deferrable timer' in /proc/timer_stats. A deferrable
-timer will appear as follows
-  10D,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-
include/linux/hrtimer.h +0 -11
index cdab81ba29f8..e52b427223ba 100644
@@ -88,12 +88,6 @@ enum hrtimer_restart {
  * @base:	pointer to the timer base (per cpu and per clock)
  * @state:	state information (See bit values above)
  * @is_rel:	Set if the timer was armed relative
- * @start_pid:  timer statistics field to store the pid of the task which
- *		started the timer
- * @start_site:	timer statistics field to store the site where the timer
- *		was started
- * @start_comm: timer statistics field to store the name of the process which
- *		started the timer
  *
  * The hrtimer structure must be initialized by hrtimer_init()
  */
@@ -104,11 +98,6 @@ struct hrtimer {
 	struct hrtimer_clock_base	*base;
 	u8				state;
 	u8				is_rel;
-#ifdef CONFIG_TIMER_STATS
-	int				start_pid;
-	void				*start_site;
-	char				start_comm[16];
-#endif
 };
 
 /**
include/linux/timer.h +0 -45

57 lines not on this page · raw message

kernel/kthread.c +0 -1
index 2318fba86277..8461a4372e8a 100644
@@ -850,7 +850,6 @@ void __kthread_queue_delayed_work(struct kthread_worker *worker,
 
 	list_add(&work->node, &worker->delayed_work_list);
 	work->worker = worker;
-	timer_stats_timer_set_start_info(&dwork->timer);
 	timer->expires = jiffies + delay;
 	add_timer(timer);
 }
kernel/time/Makefile +0 -1
index 976840d29a71..938dbf33ef49 100644
@@ -15,6 +15,5 @@ ifeq ($(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST),y)
 endif
 obj-$(CONFIG_GENERIC_SCHED_CLOCK)		+= sched_clock.o
 obj-$(CONFIG_TICK_ONESHOT)			+= tick-oneshot.o tick-sched.o
-obj-$(CONFIG_TIMER_STATS)			+= timer_stats.o
 obj-$(CONFIG_DEBUG_FS)				+= timekeeping_debug.o
 obj-$(CONFIG_TEST_UDELAY)			+= test_udelay.o
kernel/time/hrtimer.c +0 -38
index c6ecedd3b839..edabde646e58 100644
@@ -766,34 +766,6 @@ void hrtimers_resume(void)
 	clock_was_set_delayed();
 }
 
-static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	if (timer->start_site)
-		return;
-	timer->start_site = __builtin_return_address(0);
-	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
-	timer->start_pid = current->pid;
-#endif
-}
-
-static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-#endif
-}
-
-static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	if (likely(!timer_stats_active))
-		return;
-	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
-				 timer->function, timer->start_comm, 0);
-#endif
-}
-
 /*
  * Counterpart to lock_hrtimer_base above:
  */
@@ -932,7 +904,6 @@ remove_hrtimer(struct hrtimer *timer, struct hrtimer_clock_base *base, bool rest
 		 * rare case and less expensive than a smp call.
 		 */
 		debug_deactivate(timer);
-		timer_stats_hrtimer_clear_start_info(timer);
 		reprogram = base->cpu_base == this_cpu_ptr(&hrtimer_bases);
 
 		if (!restart)
@@ -990,8 +961,6 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim,
 	/* Switch the timer base, if necessary: */
 	new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED);
 
-	timer_stats_hrtimer_set_start_info(timer);
-
 	leftmost = enqueue_hrtimer(timer, new_base);
 	if (!leftmost)
 		goto unlock;
@@ -1128,12 +1097,6 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id,
 	base = hrtimer_clockid_to_base(clock_id);
 	timer->base = &cpu_base->clock_base[base];
 	timerqueue_init(&timer->node);
-
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-	timer->start_pid = -1;
-	memset(timer->start_comm, 0, TASK_COMM_LEN);
-#endif
 }
 
 /**
@@ -1217,7 +1180,6 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base,
 	raw_write_seqcount_barrier(&cpu_base->seq);
 
 	__remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0);
-	timer_stats_account_hrtimer(timer);
 	fn = timer->function;
 
 	/*
kernel/time/timer.c +2 -46

97 lines not on this page · raw message

kernel/time/timer_list.c +0 -10
index afe6cd1944fc..387a3a5aa388 100644
@@ -62,21 +62,11 @@ static void
 print_timer(struct seq_file *m, struct hrtimer *taddr, struct hrtimer *timer,
 	    int idx, u64 now)
 {
-#ifdef CONFIG_TIMER_STATS
-	char tmp[TASK_COMM_LEN + 1];
-#endif
 	SEQ_printf(m, " #%d: ", idx);
 	print_name_offset(m, taddr);
 	SEQ_printf(m, ", ");
 	print_name_offset(m, timer->function);
 	SEQ_printf(m, ", S:%02x", timer->state);
-#ifdef CONFIG_TIMER_STATS
-	SEQ_printf(m, ", ");
-	print_name_offset(m, timer->start_site);
-	memcpy(tmp, timer->start_comm, TASK_COMM_LEN);
-	tmp[TASK_COMM_LEN] = 0;
-	SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);
-#endif
 	SEQ_printf(m, "\n");
 	SEQ_printf(m, " # expires at %Lu-%Lu nsecs [in %Ld to %Ld nsecs]\n",
 		(unsigned long long)ktime_to_ns(hrtimer_get_softexpires(timer)),
kernel/time/timer_stats.c deleted +0 -425

425 lines not on this page · raw message

kernel/workqueue.c +0 -2
index 1d9fb6543a66..072cbc9b175d 100644
@@ -1523,8 +1523,6 @@ static void __queue_delayed_work(int cpu, struct workqueue_struct *wq,
 		return;
 	}
 
-	timer_stats_timer_set_start_info(&dwork->timer);
-
 	dwork->wq = wq;
 	dwork->cpu = cpu;
 	timer->expires = jiffies + delay;
lib/Kconfig.debug +0 -14
index eb9e9a7870fa..132af338d6dd 100644
@@ -980,20 +980,6 @@ config DEBUG_TIMEKEEPING
 
 	  If unsure, say N.
 
-config TIMER_STATS
-	bool "Collect kernel timers statistics"
-	depends on DEBUG_KERNEL && PROC_FS
-	help
-	  If you say Y here, additional code will be inserted into the
-	  timer routines to collect statistics about kernel timers being
-	  reprogrammed. The statistics can be read from /proc/timer_stats.
-	  The statistics collection is started by writing 1 to /proc/timer_stats,
-	  writing 0 stops it. This feature is useful to collect information
-	  about timer usage patterns in kernel and userspace. This feature
-	  is lightweight if enabled in the kernel config but not activated
-	  (it defaults to deactivated on bootup and will only be activated
-	  if some application like powertop activates it explicitly).
-
 config DEBUG_PREEMPT
 	bool "Debug preemptible kernel"
 	depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT
Signature · 6 lines
2.7.4


-- 
Kees Cook
Pixel Security
#2 lore raw
Quoted 5 lines · Linus Torvalds
Quoted 4 lines · Kees Cook

Given that the tracer can give the same information, this patch entirely removes CONFIG_TIMER_STATS.

Quoted 8 lines
Quoted 3 lines

Looks good to me. Let's wait for the 4.11 merge window though. I'm assuming I'll get this through the tip timer tree..

From a quick look at the source, powertop uses this file, and appears to

handle it gracefully if it fails to open, but it will lose functionality to determine if a timer is deferred.

Dave

#3 lore raw
Quoted 10 lines · Dave Jones
Quoted 9 lines · Linus Torvalds
Quoted 7 lines · Kees Cook

Given that the tracer can give the same information, this patch entirely removes CONFIG_TIMER_STATS.

Quoted 6 lines
Quoted 4 lines

Looks good to me. Let's wait for the 4.11 merge window though. I'm assuming I'll get this through the tip timer tree..

From a quick look at the source, powertop uses this file, and appears to

handle it gracefully if it fails to open, but it will lose functionality to determine if a timer is deferred.

The lookup happens when evaluating a timer_expire_entry tracing event.

With current kernels the same information can be retrieved from the timer_start tracing event (flags field). And that's way more sensible than scanning timer_stats because when the trace is evaluated there is no guarantee at all that the timer is still queued and exposed there. That should be trivial to fix in powertop.

Thanks,

tglx

#4 lore raw
In reply to Kees Cook

Hi,

Don't forget to send to [email protected]

Le mercredi 08 février 2017 à 11:26 -0800, Kees Cook a écrit :

Quoted 1,027 lines

Currently CONFIG_TIMER_STATS exposes process information across namespaces:

kernel/time/timer_list.c print_timer():

        SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);

/proc/timer_list:

 #11: <0000000000000000>, hrtimer_wakeup, S:01, do_nanosleep, cron/2570

Given that the tracer can give the same information, this patch entirely
removes CONFIG_TIMER_STATS.

---
v2:
- dropped doc comments for removed structure elements; thx 0-day builder.
---
 Documentation/timers/timer_stats.txt |  73 ------ include/linux/hrtimer.h              |  11 - include/linux/timer.h                |  45 ---- kernel/kthread.c                     |   1 - kernel/time/Makefile                 |   1 - kernel/time/hrtimer.c                |  38 ---- kernel/time/timer.c                  |  48 +--- kernel/time/timer_list.c             |  10 - kernel/time/timer_stats.c            | 425 -----------------------
------------
 kernel/workqueue.c                   |   2 - lib/Kconfig.debug                    |  14 --

 11 files changed, 2 insertions(+), 666 deletions(-)

 delete mode 100644 Documentation/timers/timer_stats.txt delete mode 100644 kernel/time/timer_stats.c

Documentation/timers/timer_stats.txt deleted +0 -4
deleted file mode 100644
index de835ee97455..000000000000
@@ -1,73 +0,0 @@
-timer_stats - timer usage statistics
-------------------------------------
-
-timer_stats is a debugging facility to make the timer (ab)usage in a

Linux
-system visible to kernel and userspace developers. If enabled in the config
-but not used it has almost zero runtime overhead, and a relatively small
-data structure overhead. Even if collection is enabled runtime all the

Documentation/timers/timer_stats.txt deleted +0 -3
-locking is per-CPU and lookup is hashed.
-
-timer_stats should be used by kernel and userspace developers to

verify that
-their code does not make unduly use of timers. This helps to avoid unnecessary

Documentation/timers/timer_stats.txt deleted +0 -3
-wakeups, which should be avoided to optimize power consumption.
-
-It can be enabled by CONFIG_TIMER_STATS in the "Kernel hacking"

configuration

Documentation/timers/timer_stats.txt deleted +0 -3
-section.
-
-timer_stats collects information about the timer events which are

fired in a

Documentation/timers/timer_stats.txt deleted +0 -11
-Linux system over a sample period:
-
-- the pid of the task(process) which initialized the timer
-- the name of the process which initialized the timer
-- the function where the timer was initialized
-- the callback function which is associated to the timer
-- the number of events (callbacks)
-
-timer_stats adds an entry to /proc: /proc/timer_stats
-
-This entry is used to control the statistics functionality and to

read out the

Documentation/timers/timer_stats.txt deleted +0 -14
-sampled information.
-
-The timer_stats functionality is inactive on bootup.
-
-To activate a sample period issue:
-# echo 1 >/proc/timer_stats
-
-To stop a sample period issue:
-# echo 0 >/proc/timer_stats
-
-The statistics can be retrieved by:
-# cat /proc/timer_stats
-
-While sampling is enabled, each readout from /proc/timer_stats will

see
-newly updated statistics. Once sampling is disabled, the sampled information
-is kept until a new sample period is started. This allows multiple readouts.

Documentation/timers/timer_stats.txt deleted +0 -5
-
-Sample output of /proc/timer_stats:
-
-Timerstats sample period: 3.888770 s
-  12,     0 swapper          hrtimer_stop_sched_tick

(hrtimer_sched_tick)
-  15,     1 swapper          hcd_submit_urb (rh_timer_func) -   4,   959 kedac            schedule_timeout (process_timeout) -   1,     0 swapper          page_writeback_init (wb_timer_fn) -  28,     0 swapper          hrtimer_stop_sched_tick (hrtimer_sched_tick)
-  22,  2948 IRQ 4            tty_flip_buffer_push (delayed_work_timer_fn)
-   3,  3100 bash             schedule_timeout (process_timeout) -   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          neigh_table_init_no_netlink (neigh_periodic_timer)
-   1,  2292 ip               __netdev_watchdog_up (dev_watchdog) -   1,    23 events/1         do_cache_clean (delayed_work_timer_fn) -90 total events, 30.0 events/sec

Documentation/timers/timer_stats.txt deleted +0 -2
-
-The first column is the number of events, the second column the pid,

the third
-column is the name of the process. The forth column shows the function which
-initialized the timer and in parenthesis the callback function which was

Documentation/timers/timer_stats.txt deleted +0 -5
-executed on expiry.
-
-    Thomas, Ingo
-
-Added flag to indicate 'deferrable timer' in /proc/timer_stats. A

deferrable
-timer will appear as follows
-  10D,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)

Documentation/timers/timer_stats.txt deleted +0 -1
-
include/linux/hrtimer.h +0 -1
index cdab81ba29f8..e52b427223ba 100644
@@ -88,12 +88,6 @@ enum hrtimer_restart {
   * @base:	pointer to the timer base (per cpu and per clock)
   * @state:	state information (See bit values above)
   * @is_rel:	Set if the timer was armed relative
- * @start_pid:  timer statistics field to store the pid of the task

which
- * started the timer
- * @start_site: timer statistics field to store the site
where the timer
- * was started
- * @start_comm: timer statistics field to store the name of the process which
- * started the timer
  *
  * The hrtimer structure must be initialized by hrtimer_init() */

include/linux/hrtimer.h +0 -5
@@ -104,11 +98,6 @@ struct hrtimer {
  	struct hrtimer_clock_base	*base;
  	u8				state;
  	u8				is_rel;
-#ifdef CONFIG_TIMER_STATS
-	int				start_pid;
-	void				*start_site;
-	char				start_comm[16];
-#endif
  };


 /**

include/linux/timer.h +0 -0
index 51d601f192d4..5a209b84fd9e 100644
@@ -20,11 +20,6 @@ struct timer_list {
  	unsigned long		data;
  	u32			flags;

include/linux/timer.h +0 -5
-#ifdef CONFIG_TIMER_STATS
-	int			start_pid;
-	void			*start_site;
-	char			start_comm[16];
-#endif

 #ifdef CONFIG_LOCKDEP
  struct lockdep_map lockdep_map;
 #endif

include/linux/timer.h +0 -0
@@ -197,46 +192,6 @@ extern int mod_timer_pending(struct timer_list
 *timer, unsigned long expires);
   */
  #define NEXT_TIMER_MAX_DELTA	((1UL << 30) - 1)

include/linux/timer.h +0 -10
-/*
- * Timer-statistics info:
- */
-#ifdef CONFIG_TIMER_STATS
-
-extern int timer_stats_active;
-
-extern void init_timer_stats(void);
-
-extern void timer_stats_update_stats(void *timer, pid_t pid, void

*startf,

include/linux/timer.h +0 -1
-				     void *timerf, char *comm, u32

flags);

include/linux/timer.h +0 -2
-
-extern void __timer_stats_timer_set_start_info(struct timer_list

*timer,

include/linux/timer.h +0 -3
-					       void *addr);
-
-static inline void timer_stats_timer_set_start_info(struct

timer_list *timer)

include/linux/timer.h +0 -4
-{
-	if (likely(!timer_stats_active))
-		return;
-	__timer_stats_timer_set_start_info(timer,

__builtin_return_address(0));

include/linux/timer.h +0 -3
-}
-
-static inline void timer_stats_timer_clear_start_info(struct

timer_list *timer)

include/linux/timer.h +0 -9
-{
-	timer->start_site = NULL;
-}
-#else
-static inline void init_timer_stats(void)
-{
-}
-
-static inline void timer_stats_timer_set_start_info(struct

timer_list *timer)

include/linux/timer.h +0 -4
-{
-}
-
-static inline void timer_stats_timer_clear_start_info(struct

timer_list *timer)

include/linux/timer.h +0 -4
-{
-}
-#endif
-

 extern void add_timer(struct timer_list *timer);

 extern int try_to_del_timer_sync(struct timer_list *timer);

kernel/kthread.c +0 -0
index 2318fba86277..8461a4372e8a 100644
@@ -850,7 +850,6 @@ void __kthread_queue_delayed_work(struct
 kthread_worker *worker,


  list_add(&work->node, &worker->delayed_work_list); work->worker = worker;

kernel/kthread.c +0 -1
-	timer_stats_timer_set_start_info(&dwork->timer);

  timer->expires = jiffies + delay;
  add_timer(timer);
 }

kernel/time/Makefile +0 -0
index 976840d29a71..938dbf33ef49 100644
@@ -15,6 +15,5 @@ ifeq ($(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST),y)

 endif
 obj-$(CONFIG_GENERIC_SCHED_CLOCK) += sched_clock.o
 obj-$(CONFIG_TICK_ONESHOT) += tick-oneshot.o
tick-sched.o
-obj-$(CONFIG_TIMER_STATS) += timer_stats.o
 obj-$(CONFIG_DEBUG_FS) +=
timekeeping_debug.o
 obj-$(CONFIG_TEST_UDELAY) += test_udelay.o

kernel/time/hrtimer.c +0 -0
index c6ecedd3b839..edabde646e58 100644
@@ -766,34 +766,6 @@ void hrtimers_resume(void)
  	clock_was_set_delayed();
  }


-static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)

kernel/time/hrtimer.c +0 -11
-{
-#ifdef CONFIG_TIMER_STATS
-	if (timer->start_site)
-		return;
-	timer->start_site = __builtin_return_address(0);
-	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
-	timer->start_pid = current->pid;
-#endif
-}
-
-static inline void timer_stats_hrtimer_clear_start_info(struct

hrtimer *timer)

kernel/time/hrtimer.c +0 -7
-{
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-#endif
-}
-
-static inline void timer_stats_account_hrtimer(struct hrtimer

*timer)

kernel/time/hrtimer.c +0 -5
-{
-#ifdef CONFIG_TIMER_STATS
-	if (likely(!timer_stats_active))
-		return;
-	timer_stats_update_stats(timer, timer->start_pid, timer-
Quoted 1 line

start_site,

kernel/time/hrtimer.c +0 -1
-				 timer->function, timer->start_comm, 

0);

kernel/time/hrtimer.c +0 -3
-#endif
-}
-

 /*
  * Counterpart to lock_hrtimer_base above:
  */

kernel/time/hrtimer.c +0 -0
@@ -932,7 +904,6 @@ remove_hrtimer(struct hrtimer *timer, struct

hrtimer_clock_base *base, bool rest
   * rare case and less expensive than a smp call. */
  debug_deactivate(timer);

kernel/time/hrtimer.c +0 -1
-		timer_stats_hrtimer_clear_start_info(timer);

  reprogram = base->cpu_base ==
this_cpu_ptr(&hrtimer_bases);

  if (!restart)

kernel/time/hrtimer.c +0 -0
@@ -990,8 +961,6 @@ void hrtimer_start_range_ns(struct hrtimer
 *timer, ktime_t tim,
  	/* Switch the timer base, if necessary: */

  new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED);

kernel/time/hrtimer.c +0 -2
-	timer_stats_hrtimer_set_start_info(timer);
-

  leftmost = enqueue_hrtimer(timer, new_base); if (!leftmost)
  goto unlock;

kernel/time/hrtimer.c +0 -6
@@ -1128,12 +1097,6 @@ static void __hrtimer_init(struct hrtimer
 *timer, clockid_t clock_id,
  	base = hrtimer_clockid_to_base(clock_id);
  	timer->base = &cpu_base->clock_base[base];
  	timerqueue_init(&timer->node);
-
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-	timer->start_pid = -1;
-	memset(timer->start_comm, 0, TASK_COMM_LEN);
-#endif
  }


 /**

kernel/time/hrtimer.c +0 -0
@@ -1217,7 +1180,6 @@ static void __run_hrtimer(struct
 hrtimer_cpu_base *cpu_base,
  	raw_write_seqcount_barrier(&cpu_base->seq);


  __remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0);

kernel/time/hrtimer.c +0 -1
-	timer_stats_account_hrtimer(timer);

  fn = timer->function;

  /*

kernel/time/timer.c +0 -0
index ec33a6933eae..82a6bfa0c307 100644
@@ -571,38 +571,6 @@ internal_add_timer(struct timer_base *base,
 struct timer_list *timer)
  	trigger_dyntick_cpu(base, timer);
  }

kernel/time/timer.c +0 -2
-#ifdef CONFIG_TIMER_STATS
-void __timer_stats_timer_set_start_info(struct timer_list *timer,

void *addr)

kernel/time/timer.c +0 -30
-{
-	if (timer->start_site)
-		return;
-
-	timer->start_site = addr;
-	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
-	timer->start_pid = current->pid;
-}
-
-static void timer_stats_account_timer(struct timer_list *timer)
-{
-	void *site;
-
-	/*
-	 * start_site can be concurrently reset by
-	 * timer_stats_timer_clear_start_info()
-	 */
-	site = READ_ONCE(timer->start_site);
-	if (likely(!site))
-		return;
-
-	timer_stats_update_stats(timer, timer->start_pid, site,
-				 timer->function, timer->start_comm,
-				 timer->flags);
-}
-
-#else
-static void timer_stats_account_timer(struct timer_list *timer) {}
-#endif
-

 #ifdef CONFIG_DEBUG_OBJECTS_TIMERS

 static struct debug_obj_descr timer_debug_descr;

kernel/time/timer.c +0 -5
@@ -789,11 +757,6 @@ static void do_init_timer(struct timer_list
 *timer, unsigned int flags,
  {
  	timer->entry.pprev = NULL;
  	timer->flags = flags | raw_smp_processor_id();
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-	timer->start_pid = -1;
-	memset(timer->start_comm, 0, TASK_COMM_LEN);
-#endif
  	lockdep_init_map(&timer->lockdep_map, name, key, 0);
  }

kernel/time/timer.c +0 -0
@@ -1001,8 +964,6 @@ __mod_timer(struct timer_list *timer, unsigned
 long expires, bool pending_only)
  		base = lock_timer_base(timer, &flags);
  	}

kernel/time/timer.c +0 -2
-	timer_stats_timer_set_start_info(timer);
-

  ret = detach_if_pending(timer, base, false); if (!ret && pending_only)
  goto out_unlock;

kernel/time/timer.c +0 -0
@@ -1130,7 +1091,6 @@ void add_timer_on(struct timer_list *timer, int
 cpu)
  	struct timer_base *new_base, *base;
  	unsigned long flags;

kernel/time/timer.c +0 -1
-	timer_stats_timer_set_start_info(timer);

  BUG_ON(timer_pending(timer) || !timer->function);
  new_base = get_timer_cpu_base(timer->flags, cpu);

kernel/time/timer.c +0 -0
@@ -1176,7 +1136,6 @@ int del_timer(struct timer_list *timer)


  debug_assert_init(timer);

kernel/time/timer.c +0 -1
-	timer_stats_timer_clear_start_info(timer);

  if (timer_pending(timer)) {
  base = lock_timer_base(timer, &flags); ret = detach_if_pending(timer, base, true);

kernel/time/timer.c +0 -0
@@ -1204,10 +1163,9 @@ int try_to_del_timer_sync(struct timer_list
 *timer)


  base = lock_timer_base(timer, &flags);

kernel/time/timer.c +1 -2
-	if (base->running_timer != timer) {
-		timer_stats_timer_clear_start_info(timer);
+	if (base->running_timer != timer)

  ret = detach_if_pending(timer, base, true);

kernel/time/timer.c +1 -1
-	}
+

  spin_unlock_irqrestore(&base->lock, flags);
  return ret;

kernel/time/timer.c +0 -0
@@ -1331,7 +1289,6 @@ static void expire_timers(struct timer_base
 *base, struct hlist_head *head)
  		unsigned long data;


  timer = hlist_entry(head->first, struct timer_list, entry);

kernel/time/timer.c +0 -1
-		timer_stats_account_timer(timer);


  base->running_timer = timer;
  detach_timer(timer, true);

kernel/time/timer.c +0 -1
@@ -1868,7 +1825,6 @@ static void __init init_timer_cpus(void)
  void __init init_timers(void)
  {
  	init_timer_cpus();
-	init_timer_stats();
  	open_softirq(TIMER_SOFTIRQ, run_timer_softirq);
  }

kernel/time/timer_list.c +0 -0
index afe6cd1944fc..387a3a5aa388 100644
@@ -62,21 +62,11 @@ static void

 print_timer(struct seq_file *m, struct hrtimer *taddr, struct hrtimer *timer,
      int idx, u64 now)
 {

kernel/time/timer_list.c +0 -3
-#ifdef CONFIG_TIMER_STATS
-	char tmp[TASK_COMM_LEN + 1];
-#endif

  SEQ_printf(m, " #%d: ", idx);
  print_name_offset(m, taddr);
  SEQ_printf(m, ", ");
  print_name_offset(m, timer->function);
  SEQ_printf(m, ", S:%02x", timer->state);

kernel/time/timer_list.c +0 -7
-#ifdef CONFIG_TIMER_STATS
-	SEQ_printf(m, ", ");
-	print_name_offset(m, timer->start_site);
-	memcpy(tmp, timer->start_comm, TASK_COMM_LEN);
-	tmp[TASK_COMM_LEN] = 0;
-	SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);
-#endif

  SEQ_printf(m, "\n");
  SEQ_printf(m, " # expires at %Lu-%Lu nsecs [in %Ld to %Ld nsecs]\n",
  (unsigned long long)ktime_to_ns(hrtimer_get_softexpires(timer)),

kernel/time/timer_stats.c deleted +0 -7
deleted file mode 100644
index afddded947df..000000000000
@@ -1,425 +0,0 @@
-/*
- * kernel/time/timer_stats.c
- *
- * Collect timer usage statistics.
- *
- * Copyright(C) 2006, Red Hat, Inc., Ingo Molnar
- * Copyright(C) 2006 Timesys Corp., Thomas Gleixner <[email protected]

m>
- *
- * timer_stats is based on timer_top, a similar functionality which was part of
- * Con Kolivas dyntick patch set. It was developed by Daniel Petrini at the
- * Instituto Nokia de Tecnologia - INdT - Manaus. timer_top's design was based
- * on dynamic allocation of the statistics entries and linear search based
- * lookup combined with a global lock, rather than the static array, hash
- * and per-CPU locking which is used by timer_stats. It was written for the
- * pre hrtimer kernel code and therefore did not take hrtimers into account.
- * Nevertheless it provided the base for the timer_stats implementation and
- * was a helpful source of inspiration. Kudos to Daniel and the Nokia folks
- * for this effort.
- *
- * timer_top.c is
- * Copyright (C) 2005 Instituto Nokia de Tecnologia - INdT -
Manaus
- * Written by Daniel Petrini <[email protected]>
- * timer_top.c was released under the GNU General Public
License version 2
- *
- * We export the addresses and counting of timer functions being called,
- * the pid and cmdline from the owner process if applicable.
- *
- * Start/stop data collection:
- * # echo [1|0] >/proc/timer_stats
- *
- * Display the information collected so far:
- * # cat /proc/timer_stats
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */

kernel/time/timer_stats.c deleted +0 -12
-
-#include <linux/proc_fs.h>
-#include <linux/module.h>
-#include <linux/spinlock.h>
-#include <linux/sched.h>
-#include <linux/seq_file.h>
-#include <linux/kallsyms.h>
-
-#include <linux/uaccess.h>
-
-/*
- * This is our basic unit of interest: a timer expiry event

identified
- * by the timer, its start/expire functions and the PID of the task that
- * started the timer. We count the number of times an event happens:
- */

kernel/time/timer_stats.c deleted +0 -76
-struct entry {
-	/*
-	 * Hash list:
-	 */
-	struct entry		*next;
-
-	/*
-	 * Hash keys:
-	 */
-	void			*timer;
-	void			*start_func;
-	void			*expire_func;
-	pid_t			pid;
-
-	/*
-	 * Number of timeout events:
-	 */
-	unsigned long		count;
-	u32			flags;
-
-	/*
-	 * We save the command-line string to preserve
-	 * this information past task exit:
-	 */
-	char			comm[TASK_COMM_LEN + 1];
-
-} ____cacheline_aligned_in_smp;
-
-/*
- * Spinlock protecting the tables - not taken during lookup:
- */
-static DEFINE_RAW_SPINLOCK(table_lock);
-
-/*
- * Per-CPU lookup locks for fast hash lookup:
- */
-static DEFINE_PER_CPU(raw_spinlock_t, tstats_lookup_lock);
-
-/*
- * Mutex to serialize state changes with show-stats activities:
- */
-static DEFINE_MUTEX(show_mutex);
-
-/*
- * Collection status, active/inactive:
- */
-int __read_mostly timer_stats_active;
-
-/*
- * Beginning/end timestamps of measurement:
- */
-static ktime_t time_start, time_stop;
-
-/*
- * tstat entry structs only get allocated while collection is
- * active and never freed during that time - this simplifies
- * things quite a bit.
- *
- * They get freed when a new collection period is started.
- */
-#define MAX_ENTRIES_BITS	10
-#define MAX_ENTRIES		(1UL << MAX_ENTRIES_BITS)
-
-static unsigned long nr_entries;
-static struct entry entries[MAX_ENTRIES];
-
-static atomic_t overflow_count;
-
-/*
- * The entries are in a hash-table, for fast lookup:
- */
-#define TSTAT_HASH_BITS		(MAX_ENTRIES_BITS - 1)
-#define TSTAT_HASH_SIZE		(1UL << TSTAT_HASH_BITS)
-#define TSTAT_HASH_MASK		(TSTAT_HASH_SIZE - 1)
-
-#define __tstat_hashfn(entry)					
	\
kernel/time/timer_stats.c deleted +0 -1
-	(((unsigned long)(entry)->timer       ^			
	\
kernel/time/timer_stats.c deleted +0 -1
-	  (unsigned long)(entry)->start_func  ^			
	\
kernel/time/timer_stats.c deleted +0 -1
-	  (unsigned long)(entry)->expire_func ^			
	\
kernel/time/timer_stats.c deleted +0 -1
-	  (unsigned long)(entry)->pid		) &

TSTAT_HASH_MASK)

kernel/time/timer_stats.c deleted +0 -2
-
-#define tstat_hashentry(entry)	(tstat_hash_table +

__tstat_hashfn(entry))

kernel/time/timer_stats.c deleted +0 -2
-
-static struct entry *tstat_hash_table[TSTAT_HASH_SIZE]

__read_mostly;

kernel/time/timer_stats.c deleted +0 -74
-
-static void reset_entries(void)
-{
-	nr_entries = 0;
-	memset(entries, 0, sizeof(entries));
-	memset(tstat_hash_table, 0, sizeof(tstat_hash_table));
-	atomic_set(&overflow_count, 0);
-}
-
-static struct entry *alloc_entry(void)
-{
-	if (nr_entries >= MAX_ENTRIES)
-		return NULL;
-
-	return entries + nr_entries++;
-}
-
-static int match_entries(struct entry *entry1, struct entry *entry2)
-{
-	return entry1->timer       == entry2->timer	  &&
-	       entry1->start_func  == entry2->start_func  &&
-	       entry1->expire_func == entry2->expire_func &&
-	       entry1->pid	   == entry2->pid;
-}
-
-/*
- * Look up whether an entry matching this item is present
- * in the hash already. Must be called with irqs off and the
- * lookup lock held:
- */
-static struct entry *tstat_lookup(struct entry *entry, char *comm)
-{
-	struct entry **head, *curr, *prev;
-
-	head = tstat_hashentry(entry);
-	curr = *head;
-
-	/*
-	 * The fastpath is when the entry is already hashed,
-	 * we do this with the lookup lock held, but with the
-	 * table lock not held:
-	 */
-	while (curr) {
-		if (match_entries(curr, entry))
-			return curr;
-
-		curr = curr->next;
-	}
-	/*
-	 * Slowpath: allocate, set up and link a new hash entry:
-	 */
-	prev = NULL;
-	curr = *head;
-
-	raw_spin_lock(&table_lock);
-	/*
-	 * Make sure we have not raced with another CPU:
-	 */
-	while (curr) {
-		if (match_entries(curr, entry))
-			goto out_unlock;
-
-		prev = curr;
-		curr = curr->next;
-	}
-
-	curr = alloc_entry();
-	if (curr) {
-		*curr = *entry;
-		curr->count = 0;
-		curr->next = NULL;
-		memcpy(curr->comm, comm, TASK_COMM_LEN);
-
-		smp_mb(); /* Ensure that curr is initialized before

insert */

kernel/time/timer_stats.c deleted +0 -18
-
-		if (prev)
-			prev->next = curr;
-		else
-			*head = curr;
-	}
- out_unlock:
-	raw_spin_unlock(&table_lock);
-
-	return curr;
-}
-
-/**
- * timer_stats_update_stats - Update the statistics for a timer.
- * @timer:	pointer to either a timer_list or a hrtimer
- * @pid:	the pid of the task which set up the timer
- * @startf:	pointer to the function which did the timer setup
- * @timerf:	pointer to the timer callback function of the

timer
- * @comm: name of the process which set up the timer
- * @tflags: The flags field of the timer
- *
- * When the timer is already registered, then the event counter is
- * incremented. Otherwise the timer is registered in a free slot.
- */

kernel/time/timer_stats.c deleted +0 -36
-void timer_stats_update_stats(void *timer, pid_t pid, void *startf,
-			      void *timerf, char *comm, u32 tflags)
-{
-	/*
-	 * It doesn't matter which lock we take:
-	 */
-	raw_spinlock_t *lock;
-	struct entry *entry, input;
-	unsigned long flags;
-
-	if (likely(!timer_stats_active))
-		return;
-
-	lock = &per_cpu(tstats_lookup_lock, raw_smp_processor_id());
-
-	input.timer = timer;
-	input.start_func = startf;
-	input.expire_func = timerf;
-	input.pid = pid;
-	input.flags = tflags;
-
-	raw_spin_lock_irqsave(lock, flags);
-	if (!timer_stats_active)
-		goto out_unlock;
-
-	entry = tstat_lookup(&input, comm);
-	if (likely(entry))
-		entry->count++;
-	else
-		atomic_inc(&overflow_count);
-
- out_unlock:
-	raw_spin_unlock_irqrestore(lock, flags);
-}
-
-static void print_name_offset(struct seq_file *m, unsigned long

addr)

kernel/time/timer_stats.c deleted +0 -32
-{
-	char symname[KSYM_NAME_LEN];
-
-	if (lookup_symbol_name(addr, symname) < 0)
-		seq_printf(m, "<%p>", (void *)addr);
-	else
-		seq_printf(m, "%s", symname);
-}
-
-static int tstats_show(struct seq_file *m, void *v)
-{
-	struct timespec64 period;
-	struct entry *entry;
-	unsigned long ms;
-	long events = 0;
-	ktime_t time;
-	int i;
-
-	mutex_lock(&show_mutex);
-	/*
-	 * If still active then calculate up to now:
-	 */
-	if (timer_stats_active)
-		time_stop = ktime_get();
-
-	time = ktime_sub(time_stop, time_start);
-
-	period = ktime_to_timespec64(time);
-	ms = period.tv_nsec / 1000000;
-
-	seq_puts(m, "Timer Stats Version: v0.3\n");
-	seq_printf(m, "Sample period: %ld.%03ld s\n",

(long)period.tv_sec, ms);

kernel/time/timer_stats.c deleted +0 -2
-	if (atomic_read(&overflow_count))
-		seq_printf(m, "Overflow: %d entries\n",

atomic_read(&overflow_count));

kernel/time/timer_stats.c deleted +0 -1
-	seq_printf(m, "Collection: %s\n", timer_stats_active ?

"active" : "inactive");

kernel/time/timer_stats.c deleted +0 -6
-
-	for (i = 0; i < nr_entries; i++) {
-		entry = entries + i;
-		if (entry->flags & TIMER_DEFERRABLE) {
-			seq_printf(m, "%4luD, %5d %-16s ",
-				entry->count, entry->pid, entry-
Quoted 1 line

comm);

kernel/time/timer_stats.c deleted +0 -3
-		} else {
-			seq_printf(m, " %4lu, %5d %-16s ",
-				entry->count, entry->pid, entry-
Quoted 1 line

comm);

kernel/time/timer_stats.c deleted +0 -3
-		}
-
-		print_name_offset(m, (unsigned long)entry-
Quoted 1 line

start_func);

kernel/time/timer_stats.c deleted +0 -2
-		seq_puts(m, " (");
-		print_name_offset(m, (unsigned long)entry-
Quoted 1 line

expire_func);

kernel/time/timer_stats.c deleted +0 -11
-		seq_puts(m, ")\n");
-
-		events += entry->count;
-	}
-
-	ms += period.tv_sec * 1000;
-	if (!ms)
-		ms = 1;
-
-	if (events && period.tv_sec)
-		seq_printf(m, "%ld total events, %ld.%03ld

events/sec\n",

kernel/time/timer_stats.c deleted +0 -21
-			   events, events * 1000 / ms,
-			   (events * 1000000 / ms) % 1000);
-	else
-		seq_printf(m, "%ld total events\n", events);
-
-	mutex_unlock(&show_mutex);
-
-	return 0;
-}
-
-/*
- * After a state change, make sure all concurrent lookup/update
- * activities have stopped:
- */
-static void sync_access(void)
-{
-	unsigned long flags;
-	int cpu;
-
-	for_each_online_cpu(cpu) {
-		raw_spinlock_t *lock = &per_cpu(tstats_lookup_lock,

cpu);

kernel/time/timer_stats.c deleted +0 -8
-
-		raw_spin_lock_irqsave(lock, flags);
-		/* nothing */
-		raw_spin_unlock_irqrestore(lock, flags);
-	}
-}
-
-static ssize_t tstats_write(struct file *file, const char __user

*buf,

kernel/time/timer_stats.c deleted +0 -54
-			    size_t count, loff_t *offs)
-{
-	char ctl[2];
-
-	if (count != 2 || *offs)
-		return -EINVAL;
-
-	if (copy_from_user(ctl, buf, count))
-		return -EFAULT;
-
-	mutex_lock(&show_mutex);
-	switch (ctl[0]) {
-	case '0':
-		if (timer_stats_active) {
-			timer_stats_active = 0;
-			time_stop = ktime_get();
-			sync_access();
-		}
-		break;
-	case '1':
-		if (!timer_stats_active) {
-			reset_entries();
-			time_start = ktime_get();
-			smp_mb();
-			timer_stats_active = 1;
-		}
-		break;
-	default:
-		count = -EINVAL;
-	}
-	mutex_unlock(&show_mutex);
-
-	return count;
-}
-
-static int tstats_open(struct inode *inode, struct file *filp)
-{
-	return single_open(filp, tstats_show, NULL);
-}
-
-static const struct file_operations tstats_fops = {
-	.open		= tstats_open,
-	.read		= seq_read,
-	.write		= tstats_write,
-	.llseek		= seq_lseek,
-	.release	= single_release,
-};
-
-void __init init_timer_stats(void)
-{
-	int cpu;
-
-	for_each_possible_cpu(cpu)
-		raw_spin_lock_init(&per_cpu(tstats_lookup_lock,

cpu));

kernel/time/timer_stats.c deleted +0 -12
-}
-
-static int __init init_tstats_procfs(void)
-{
-	struct proc_dir_entry *pe;
-
-	pe = proc_create("timer_stats", 0644, NULL, &tstats_fops);
-	if (!pe)
-		return -ENOMEM;
-	return 0;
-}
-__initcall(init_tstats_procfs);
kernel/workqueue.c +0 -0
index 1d9fb6543a66..072cbc9b175d 100644
@@ -1523,8 +1523,6 @@ static void __queue_delayed_work(int cpu,
 struct workqueue_struct *wq,
  		return;
  	}

kernel/workqueue.c +0 -2
-	timer_stats_timer_set_start_info(&dwork->timer);
-

  dwork->wq = wq;
  dwork->cpu = cpu;
  timer->expires = jiffies + delay;

lib/Kconfig.debug +0 -0
index eb9e9a7870fa..132af338d6dd 100644
@@ -980,20 +980,6 @@ config DEBUG_TIMEKEEPING


    If unsure, say N.

lib/Kconfig.debug +0 -5
-config TIMER_STATS
-	bool "Collect kernel timers statistics"
-	depends on DEBUG_KERNEL && PROC_FS
-	help
-	  If you say Y here, additional code will be inserted into

the

lib/Kconfig.debug +0 -1
-	  timer routines to collect statistics about kernel timers

being

lib/Kconfig.debug +0 -1
-	  reprogrammed. The statistics can be read from

/proc/timer_stats.

lib/Kconfig.debug +0 -1
-	  The statistics collection is started by writing 1 to

/proc/timer_stats,

lib/Kconfig.debug +0 -1
-	  writing 0 stops it. This feature is useful to collect

information

lib/Kconfig.debug +0 -1
-	  about timer usage patterns in kernel and userspace. This

feature

lib/Kconfig.debug +0 -1
-	  is lightweight if enabled in the kernel config but not

activated

lib/Kconfig.debug +0 -1
-	  (it defaults to deactivated on bootup and will only be

activated

lib/Kconfig.debug +0 -1
-	  if some application like powertop activates it

explicitly).
-
 config DEBUG_PREEMPT
  bool "Debug preemptible kernel"
  depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT --
2.7.4

#5 lore raw
In reply to Kees Cook
time: Remove CONFIG_TIMER_STATS
Commit-ID:  dfb4357da6ddbdf57d583ba64361c9d792b0e0b1
Gitweb:     http://git.kernel.org/tip/dfb4357da6ddbdf57d583ba64361c9d792b0e0b1
Author:     Kees Cook <[email protected]>
AuthorDate: Wed, 8 Feb 2017 11:26:59 -0800
Committer:  Thomas Gleixner <[email protected]>
CommitDate: Fri, 10 Feb 2017 11:15:08 +0100

time: Remove CONFIG_TIMER_STATS

Currently CONFIG_TIMER_STATS exposes process information across namespaces:

kernel/time/timer_list.c print_timer():

        SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);

/proc/timer_list:

#11: <0000000000000000>, hrtimer_wakeup, S:01, do_nanosleep, cron/2570

Given that the tracer can give the same information, this patch entirely removes CONFIG_TIMER_STATS.

Cc Nicolas Pitre <[email protected]>
Cc Lai Jiangshan <[email protected]>
Cc Shuah Khan <[email protected]>
Cc Xing Gao <[email protected]>
Cc Jonathan Corbet <[email protected]>
Cc Jessica Frazelle <[email protected]>
Cc Nicolas Iooss <[email protected]>
Cc "Paul E. McKenney" <[email protected]>
Cc Petr Mladek <[email protected]>
Cc Richard Cochran <[email protected]>
Cc Tejun Heo <[email protected]>
Cc Michal Marek <[email protected]>
Cc Josh Poimboeuf <[email protected]>
Cc Dmitry Vyukov <[email protected]>
Cc Oleg Nesterov <[email protected]>
Cc "Eric W. Biederman" <[email protected]>
Cc Olof Johansson <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Arjan van de Ven <[email protected]>
---
 Documentation/timers/timer_stats.txt |  73 ------
 include/linux/hrtimer.h              |  11 -
 include/linux/timer.h                |  45 ----
 kernel/kthread.c                     |   1 -
 kernel/time/Makefile                 |   1 -
 kernel/time/hrtimer.c                |  38 ----
 kernel/time/timer.c                  |  48 +---
 kernel/time/timer_list.c             |  10 -
 kernel/time/timer_stats.c            | 425 -----------------------------------
 kernel/workqueue.c                   |   2 -
 lib/Kconfig.debug                    |  14 --
 11 files changed, 2 insertions(+), 666 deletions(-)
Documentation/timers/timer_stats.txt deleted +0 -73
deleted file mode 100644
index de835ee..0000000
@@ -1,73 +0,0 @@
-timer_stats - timer usage statistics
-------------------------------------
-
-timer_stats is a debugging facility to make the timer (ab)usage in a Linux
-system visible to kernel and userspace developers. If enabled in the config
-but not used it has almost zero runtime overhead, and a relatively small
-data structure overhead. Even if collection is enabled runtime all the
-locking is per-CPU and lookup is hashed.
-
-timer_stats should be used by kernel and userspace developers to verify that
-their code does not make unduly use of timers. This helps to avoid unnecessary
-wakeups, which should be avoided to optimize power consumption.
-
-It can be enabled by CONFIG_TIMER_STATS in the "Kernel hacking" configuration
-section.
-
-timer_stats collects information about the timer events which are fired in a
-Linux system over a sample period:
-
-- the pid of the task(process) which initialized the timer
-- the name of the process which initialized the timer
-- the function where the timer was initialized
-- the callback function which is associated to the timer
-- the number of events (callbacks)
-
-timer_stats adds an entry to /proc: /proc/timer_stats
-
-This entry is used to control the statistics functionality and to read out the
-sampled information.
-
-The timer_stats functionality is inactive on bootup.
-
-To activate a sample period issue:
-# echo 1 >/proc/timer_stats
-
-To stop a sample period issue:
-# echo 0 >/proc/timer_stats
-
-The statistics can be retrieved by:
-# cat /proc/timer_stats
-
-While sampling is enabled, each readout from /proc/timer_stats will see
-newly updated statistics. Once sampling is disabled, the sampled information
-is kept until a new sample period is started. This allows multiple readouts.
-
-Sample output of /proc/timer_stats:
-
-Timerstats sample period: 3.888770 s
-  12,     0 swapper          hrtimer_stop_sched_tick (hrtimer_sched_tick)
-  15,     1 swapper          hcd_submit_urb (rh_timer_func)
-   4,   959 kedac            schedule_timeout (process_timeout)
-   1,     0 swapper          page_writeback_init (wb_timer_fn)
-  28,     0 swapper          hrtimer_stop_sched_tick (hrtimer_sched_tick)
-  22,  2948 IRQ 4            tty_flip_buffer_push (delayed_work_timer_fn)
-   3,  3100 bash             schedule_timeout (process_timeout)
-   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-   1,     1 swapper          neigh_table_init_no_netlink (neigh_periodic_timer)
-   1,  2292 ip               __netdev_watchdog_up (dev_watchdog)
-   1,    23 events/1         do_cache_clean (delayed_work_timer_fn)
-90 total events, 30.0 events/sec
-
-The first column is the number of events, the second column the pid, the third
-column is the name of the process. The forth column shows the function which
-initialized the timer and in parenthesis the callback function which was
-executed on expiry.
-
-    Thomas, Ingo
-
-Added flag to indicate 'deferrable timer' in /proc/timer_stats. A deferrable
-timer will appear as follows
-  10D,     1 swapper          queue_delayed_work_on (delayed_work_timer_fn)
-
include/linux/hrtimer.h +0 -11
index cdab81b..e52b427 100644
@@ -88,12 +88,6 @@ enum hrtimer_restart {
  * @base:	pointer to the timer base (per cpu and per clock)
  * @state:	state information (See bit values above)
  * @is_rel:	Set if the timer was armed relative
- * @start_pid:  timer statistics field to store the pid of the task which
- *		started the timer
- * @start_site:	timer statistics field to store the site where the timer
- *		was started
- * @start_comm: timer statistics field to store the name of the process which
- *		started the timer
  *
  * The hrtimer structure must be initialized by hrtimer_init()
  */
@@ -104,11 +98,6 @@ struct hrtimer {
 	struct hrtimer_clock_base	*base;
 	u8				state;
 	u8				is_rel;
-#ifdef CONFIG_TIMER_STATS
-	int				start_pid;
-	void				*start_site;
-	char				start_comm[16];
-#endif
 };
 
 /**
include/linux/timer.h +0 -45

57 lines not on this page · raw message

kernel/kthread.c +0 -1
index 2318fba..8461a43 100644
@@ -850,7 +850,6 @@ void __kthread_queue_delayed_work(struct kthread_worker *worker,
 
 	list_add(&work->node, &worker->delayed_work_list);
 	work->worker = worker;
-	timer_stats_timer_set_start_info(&dwork->timer);
 	timer->expires = jiffies + delay;
 	add_timer(timer);
 }
kernel/time/Makefile +0 -1
index 976840d..938dbf3 100644
@@ -15,6 +15,5 @@ ifeq ($(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST),y)
 endif
 obj-$(CONFIG_GENERIC_SCHED_CLOCK)		+= sched_clock.o
 obj-$(CONFIG_TICK_ONESHOT)			+= tick-oneshot.o tick-sched.o
-obj-$(CONFIG_TIMER_STATS)			+= timer_stats.o
 obj-$(CONFIG_DEBUG_FS)				+= timekeeping_debug.o
 obj-$(CONFIG_TEST_UDELAY)			+= test_udelay.o
kernel/time/hrtimer.c +0 -38
index c6ecedd..edabde6 100644
@@ -766,34 +766,6 @@ void hrtimers_resume(void)
 	clock_was_set_delayed();
 }
 
-static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	if (timer->start_site)
-		return;
-	timer->start_site = __builtin_return_address(0);
-	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
-	timer->start_pid = current->pid;
-#endif
-}
-
-static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-#endif
-}
-
-static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
-{
-#ifdef CONFIG_TIMER_STATS
-	if (likely(!timer_stats_active))
-		return;
-	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
-				 timer->function, timer->start_comm, 0);
-#endif
-}
-
 /*
  * Counterpart to lock_hrtimer_base above:
  */
@@ -932,7 +904,6 @@ remove_hrtimer(struct hrtimer *timer, struct hrtimer_clock_base *base, bool rest
 		 * rare case and less expensive than a smp call.
 		 */
 		debug_deactivate(timer);
-		timer_stats_hrtimer_clear_start_info(timer);
 		reprogram = base->cpu_base == this_cpu_ptr(&hrtimer_bases);
 
 		if (!restart)
@@ -990,8 +961,6 @@ void hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim,
 	/* Switch the timer base, if necessary: */
 	new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED);
 
-	timer_stats_hrtimer_set_start_info(timer);
-
 	leftmost = enqueue_hrtimer(timer, new_base);
 	if (!leftmost)
 		goto unlock;
@@ -1128,12 +1097,6 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id,
 	base = hrtimer_clockid_to_base(clock_id);
 	timer->base = &cpu_base->clock_base[base];
 	timerqueue_init(&timer->node);
-
-#ifdef CONFIG_TIMER_STATS
-	timer->start_site = NULL;
-	timer->start_pid = -1;
-	memset(timer->start_comm, 0, TASK_COMM_LEN);
-#endif
 }
 
 /**
@@ -1217,7 +1180,6 @@ static void __run_hrtimer(struct hrtimer_cpu_base *cpu_base,
 	raw_write_seqcount_barrier(&cpu_base->seq);
 
 	__remove_hrtimer(timer, base, HRTIMER_STATE_INACTIVE, 0);
-	timer_stats_account_hrtimer(timer);
 	fn = timer->function;
 
 	/*
kernel/time/timer.c +2 -46

97 lines not on this page · raw message

kernel/time/timer_list.c +0 -10
index afe6cd1..387a3a5 100644
@@ -62,21 +62,11 @@ static void
 print_timer(struct seq_file *m, struct hrtimer *taddr, struct hrtimer *timer,
 	    int idx, u64 now)
 {
-#ifdef CONFIG_TIMER_STATS
-	char tmp[TASK_COMM_LEN + 1];
-#endif
 	SEQ_printf(m, " #%d: ", idx);
 	print_name_offset(m, taddr);
 	SEQ_printf(m, ", ");
 	print_name_offset(m, timer->function);
 	SEQ_printf(m, ", S:%02x", timer->state);
-#ifdef CONFIG_TIMER_STATS
-	SEQ_printf(m, ", ");
-	print_name_offset(m, timer->start_site);
-	memcpy(tmp, timer->start_comm, TASK_COMM_LEN);
-	tmp[TASK_COMM_LEN] = 0;
-	SEQ_printf(m, ", %s/%d", tmp, timer->start_pid);
-#endif
 	SEQ_printf(m, "\n");
 	SEQ_printf(m, " # expires at %Lu-%Lu nsecs [in %Ld to %Ld nsecs]\n",
 		(unsigned long long)ktime_to_ns(hrtimer_get_softexpires(timer)),
kernel/time/timer_stats.c deleted +0 -425

425 lines not on this page · raw message

kernel/workqueue.c +0 -2
index 1d9fb65..072cbc9 100644
@@ -1523,8 +1523,6 @@ static void __queue_delayed_work(int cpu, struct workqueue_struct *wq,
 		return;
 	}
 
-	timer_stats_timer_set_start_info(&dwork->timer);
-
 	dwork->wq = wq;
 	dwork->cpu = cpu;
 	timer->expires = jiffies + delay;
lib/Kconfig.debug +0 -14
index eb9e9a7..132af33 100644
@@ -980,20 +980,6 @@ config DEBUG_TIMEKEEPING
 
 	  If unsure, say N.
 
-config TIMER_STATS
-	bool "Collect kernel timers statistics"
-	depends on DEBUG_KERNEL && PROC_FS
-	help
-	  If you say Y here, additional code will be inserted into the
-	  timer routines to collect statistics about kernel timers being
-	  reprogrammed. The statistics can be read from /proc/timer_stats.
-	  The statistics collection is started by writing 1 to /proc/timer_stats,
-	  writing 0 stops it. This feature is useful to collect information
-	  about timer usage patterns in kernel and userspace. This feature
-	  is lightweight if enabled in the kernel config but not activated
-	  (it defaults to deactivated on bootup and will only be activated
-	  if some application like powertop activates it explicitly).
-
 config DEBUG_PREEMPT
 	bool "Debug preemptible kernel"
 	depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT