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
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
|
//po4a: entry man manual
////
uclampset(1) manpage
Copyright (C) 2020-2021 Qais Yousef <qais.yousef@arm.com>
Copyright (C) 2020-2021 Arm Ltd
This is free documentation; 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.
The GNU General Public License's references to "object code"
and "executables" are to be interpreted as the output of any
document formatting or typesetting system, including
intermediate and printed output.
This manual 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.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
////
= uclampset(1)
:doctype: manpage
:man manual: User Commands
:man source: util-linux {release-version}
:page-layout: base
:command: uclampset
:colon: :
== NAME
uclampset -
manipulate the utilization clamping attributes of the system or a process
== SYNOPSIS
*uclampset* [options] [*-m* _uclamp_min_] [*-M* _uclamp_max] _command argument_
*uclampset* [options] [*-m* _uclamp_min_] [*-M* _uclamp_max_] *-p* _PID_
== DESCRIPTION
*uclampset* sets or retrieves the utilization clamping attributes of an existing _PID_, or runs _command_ with the given attributes.
Utilization clamping is a new feature added in v5.3. It gives a hint to the scheduler about the allowed range of utilization the task should be operating at.
The utilization of the task affects frequency selection and task placement. Only schedutil cpufreq governor understands handling util clamp hints at the time of writing. Consult your kernel docs for further info about other cpufreq governors support.
If you're running on asymmetric heterogeneous system like Arm's big.LITTLE. Utilization clamping can help bias task placement. If the task is boosted such that _util_min_ value is higher than the little cores' capacity, then the scheduler will do its best to place it on a big core.
Similarly, if _util_max_ is smaller than or equal the capacity of the little cores, then the scheduler can still choose to place it there even if the actual utilization of the task is at max.
Setting a task's _uclamp_min_ to a none zero value will effectively boost the task as when it runs it'll always start from this utilization value.
By setting a task's _uclamp_max_ below 1024, this will effectively cap the task as when it runs it'll never be able to go above this utilization value.
The full utilization range is: [0:1024]. The special value -1 is used to reset to system's default.
== OPTIONS
*-m*::
Set _util_min_ value.
*-M*::
Set _util_max_ value.
*-a*, *--all-tasks*::
Set or retrieve the utilization clamping attributes of all the tasks (threads) for a given PID.
*-p*, *--pid*::
Operate on an existing PID and do not launch a new task.
*-s*, *--system*::
Set or retrieve the system-wide utilization clamping attributes.
*-R*, *--reset-on-fork*::
Set *SCHED_FLAG_RESET_ON_FORK* flag.
*-v*, *--verbose*::
Show status information.
include::man-common/help-version.adoc[]
== USAGE
//TRANSLATORS: Keep {colon} untranslated.
The default behavior is to run a new command{colon}::
*uclampset* _[-m uclamp_min]_ _[-M uclamp_max]_ _command_ [_arguments_]
//TRANSLATORS: Keep {colon} untranslated.
You can also retrieve the utilization clamping attributes of an existing task{colon}::
*uclampset -p* _PID_
//TRANSLATORS: Keep {colon} untranslated.
Or set them{colon}::
*uclampset -p* _PID_ _[-m uclamp_min]_ _[-M uclamp_max]_
//TRANSLATORS: Keep {colon} untranslated.
Or control the system-wide attributes{colon}::
*uclampset -s* _[-m uclamp_min]_ _[-M uclamp_max]_
== PERMISSIONS
A user must possess *CAP_SYS_NICE* to change the scheduling attributes of a process. Any user can retrieve the scheduling information.
== NOTES
The system wide utilization clamp attributes are there to control the _allowed_ range the tasks can use. By default both _uclamp_min_ and _uclamp_max_ are set to 1024. This means users can set the utilization clamp values for their task across the full range [0:1024].
//TRANSLATORS: Keep {colon} untranslated.
For example{colon}::
*uclampset -s* `-m 512` `-M 700`
will prevent any task from being boosted higher than 512. And all tasks in the systems are capped to a utilization of 700. Effectively rendering the maximum performance of the system to 700.
Consult your kernel docs for the exact expected behavior on that kernel.
== AUTHORS
mailto:qais.yousef@arm.com[Qais Yousef]
== SEE ALSO
*nice*(1),
*renice*(1),
*taskset*(1),
*sched*(7)
See *sched_setscheduler*(2) and *sched_setattr*(2) for a description of the Linux scheduling scheme.
include::man-common/bugreports.adoc[]
include::man-common/footer.adoc[]
ifdef::translation[]
include::man-common/translation.adoc[]
endif::[]
|
