1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
|
.\" Copyright (C) 1994, 1995, Daniel Quinlan <quinlan@yggdrasil.com>
.\" Copyright (C) 2002-2008, 2017, Michael Kerrisk <mtk.manpages@gmail.com>
.\" Copyright (C) 2023, Alejandro Colomar <alx@kernel.org>
.\"
.\" SPDX-License-Identifier: GPL-3.0-or-later
.\"
.TH proc_timer_stats 5 2024-05-02 "Linux man-pages (unreleased)"
.SH NAME
/proc/timer_stats \- timer statistics
.SH DESCRIPTION
.TP
.IR /proc/timer_stats " (from Linux 2.6.21 until Linux 4.10)"
.\" commit 82f67cd9fca8c8762c15ba7ed0d5747588c1e221
.\" Date: Fri Feb 16 01:28:13 2007 -0800
.\" Text largely derived from Documentation/timers/timer_stats.txt
.\" removed in commit dfb4357da6ddbdf57d583ba64361c9d792b0e0b1
.\" Date: Wed Feb 8 11:26:59 2017 -0800
This is a debugging facility to make timer (ab)use in a Linux
system visible to kernel and user-space developers.
It can be used by kernel and user-space developers to verify that
their code does not make undue use of timers.
The goal is to avoid unnecessary wakeups,
thereby optimizing power consumption.
.IP
If enabled in the kernel
.RB ( CONFIG_TIMER_STATS ),
but not used,
it has almost zero run-time overhead and a relatively small
data-structure overhead.
Even if collection is enabled at run time, overhead is low:
all the locking is per-CPU and lookup is hashed.
.IP
The
.I /proc/timer_stats
file is used both to control sampling facility and to read out the
sampled information.
.IP
The
.I timer_stats
functionality is inactive on bootup.
A sampling period can be started using the following command:
.IP
.in +4n
.EX
# echo 1 > /proc/timer_stats
.EE
.in
.IP
The following command stops a sampling period:
.IP
.in +4n
.EX
# echo 0 > /proc/timer_stats
.EE
.in
.IP
The statistics can be retrieved by:
.IP
.in +4n
.EX
$ cat /proc/timer_stats
.EE
.in
.IP
While sampling is enabled, each readout from
.I /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.
.IP
Sample output from
.IR /proc/timer_stats :
.IP
.in +4n
.EX
.RB $ " cat /proc/timer_stats"
Timer Stats Version: v0.3
Sample period: 1.764 s
Collection: active
255, 0 swapper/3 hrtimer_start_range_ns (tick_sched_timer)
71, 0 swapper/1 hrtimer_start_range_ns (tick_sched_timer)
58, 0 swapper/0 hrtimer_start_range_ns (tick_sched_timer)
4, 1694 gnome\-shell mod_delayed_work_on (delayed_work_timer_fn)
17, 7 rcu_sched rcu_gp_kthread (process_timeout)
\&...
1, 4911 kworker/u16:0 mod_delayed_work_on (delayed_work_timer_fn)
1D, 2522 kworker/0:0 queue_delayed_work_on (delayed_work_timer_fn)
1029 total events, 583.333 events/sec
.EE
.in
.IP
The output columns are:
.RS
.IP [1] 5
a count of the number of events,
optionally (since Linux 2.6.23) followed by the letter \[aq]D\[aq]
.\" commit c5c061b8f9726bc2c25e19dec227933a13d1e6b7 deferrable timers
if this is a deferrable timer;
.IP [2]
the PID of the process that initialized the timer;
.IP [3]
the name of the process that initialized the timer;
.IP [4]
the function where the timer was initialized; and
(in parentheses)
the callback function that is associated with the timer.
.RE
.IP
During the Linux 4.11 development cycle,
this file was removed because of security concerns,
as it exposes information across namespaces.
Furthermore, it is possible to obtain
the same information via in-kernel tracing facilities such as ftrace.
.SH SEE ALSO
.BR proc (5)
|
