tjh.dev
Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1.TH CPUPOWER\-MONITOR "1" "22/02/2011" "" "cpupower Manual"
2.SH NAME
3cpupower\-monitor \- Report processor frequency and idle statistics
4.SH SYNOPSIS
5.ft B
6.B cpupower monitor
7.RB "\-l"
8
9.B cpupower monitor
10.RB [ "\-m <mon1>," [ "<mon2>,..." ] ]
11.RB [ "\-i seconds" ]
12.br
13.B cpupower monitor
14.RB [ "\-m <mon1>," [ "<mon2>,..." ] ]
15.RB command
16.br
17.SH DESCRIPTION
18\fBcpupower-monitor \fP reports processor topology, frequency and idle power
19state statistics. Either \fBcommand\fP is forked and
20statistics are printed upon its completion, or statistics are printed periodically.
21
22\fBcpupower-monitor \fP implements independent processor sleep state and
23frequency counters. Some are retrieved from kernel statistics, some are
24directly reading out hardware registers. Use \-l to get an overview which are
25supported on your system.
26
27.SH Options
28.PP
29\-l
30.RS 4
31List available monitors on your system. Additional details about each monitor
32are shown:
33.RS 2
34.IP \(bu
35The name in quotation marks which can be passed to the \-m parameter.
36.IP \(bu
37The number of different counters the monitor supports in brackets.
38.IP \(bu
39The amount of time in seconds the counters might overflow, due to
40implementation constraints.
41.IP \(bu
42The name and a description of each counter and its processor hierarchy level
43coverage in square brackets:
44.RS 4
45.IP \(bu
46[T] \-> Thread
47.IP \(bu
48[C] \-> Core
49.IP \(bu
50[P] \-> Processor Package (Socket)
51.IP \(bu
52[M] \-> Machine/Platform wide counter
53.RE
54.RE
55.RE
56.PP
57\-m <mon1>,<mon2>,...
58.RS 4
59Only display specific monitors. Use the monitor string(s) provided by \-l option.
60.RE
61.PP
62\-i seconds
63.RS 4
64Measure intervall.
65.RE
66.PP
67command
68.RS 4
69Measure idle and frequency characteristics of an arbitrary command/workload.
70The executable \fBcommand\fP is forked and upon its exit, statistics gathered since it was
71forked are displayed.
72.RE
73.PP
74\-v
75.RS 4
76Increase verbosity if the binary was compiled with the DEBUG option set.
77.RE
78
79.SH MONITOR DESCRIPTIONS
80.SS "Idle_Stats"
81Shows statistics of the cpuidle kernel subsystem. Values are retrieved from
82/sys/devices/system/cpu/cpu*/cpuidle/state*/.
83The kernel updates these values every time an idle state is entered or
84left. Therefore there can be some inaccuracy when cores are in an idle
85state for some time when the measure starts or ends. In worst case it can happen
86that one core stayed in an idle state for the whole measure time and the idle
87state usage time as exported by the kernel did not get updated. In this case
88a state residency of 0 percent is shown while it was 100.
89
90.SS "Mperf"
91The name comes from the aperf/mperf (average and maximum) MSR registers used
92which are available on recent X86 processors. It shows the average frequency
93(including boost frequencies).
94The fact that on all recent hardware the mperf timer stops ticking in any idle
95state it is also used to show C0 (processor is active) and Cx (processor is in
96any sleep state) times. These counters do not have the inaccuracy restrictions
97the "Idle_Stats" counters may show.
98May work poorly on Linux-2.6.20 through 2.6.29, as the \fBacpi-cpufreq \fP
99kernel frequency driver periodically cleared aperf/mperf registers in those
100kernels.
101
102.SS "Nehalem" "SandyBridge"
103Intel Core and Package sleep state counters.
104Threads (hyperthreaded cores) may not be able to enter deeper core states if
105its sibling is utilized.
106Deepest package sleep states may in reality show up as machine/platform wide
107sleep states and can only be entered if all cores are idle. Look up Intel
108manuals (some are provided in the References section) for further details.
109
110.SS "Ontario" "Liano"
111AMD laptop and desktop processor (family 12h and 14h) sleep state counters.
112The registers are accessed via PCI and therefore can still be read out while
113cores have been offlined.
114
115There is one special counter: NBP1 (North Bridge P1).
116This one always returns 0 or 1, depending on whether the North Bridge P1
117power state got entered at least once during measure time.
118Being able to enter NBP1 state also depends on graphics power management.
119Therefore this counter can be used to verify whether the graphics' driver
120power management is working as expected.
121
122.SH EXAMPLES
123
124cpupower monitor -l" may show:
125.RS 4
126Monitor "Mperf" (3 states) \- Might overflow after 922000000 s
127
128 ...
129
130Monitor "Idle_Stats" (3 states) \- Might overflow after 4294967295 s
131
132 ...
133
134.RE
135cpupower monitor \-m "Idle_Stats,Mperf" scp /tmp/test /nfs/tmp
136
137Monitor the scp command, show both Mperf and Idle_Stats states counter
138statistics, but in exchanged order.
139
140
141
142.RE
143Be careful that the typical command to fully utilize one CPU by doing:
144
145cpupower monitor cat /dev/zero >/dev/null
146
147Does not work as expected, because the measured output is redirected to
148/dev/null. This could get workarounded by putting the line into an own, tiny
149shell script. Hit CTRL\-c to terminate the command and get the measure output
150displayed.
151
152.SH REFERENCES
153"BIOS and Kernel Developer’s Guide (BKDG) for AMD Family 14h Processors"
154http://support.amd.com/us/Processor_TechDocs/43170.pdf
155
156"Intel® Turbo Boost Technology
157in Intel® Core™ Microarchitecture (Nehalem) Based Processors"
158http://download.intel.com/design/processor/applnots/320354.pdf
159
160"Intel® 64 and IA-32 Architectures Software Developer's Manual
161Volume 3B: System Programming Guide"
162http://www.intel.com/products/processor/manuals
163
164.SH FILES
165.ta
166.nf
167/dev/cpu/*/msr
168/sys/devices/system/cpu/cpu*/cpuidle/state*/.
169.fi
170
171.SH "SEE ALSO"
172powertop(8), msr(4), vmstat(8)
173.PP
174.SH AUTHORS
175.nf
176Written by Thomas Renninger <trenn@suse.de>
177
178Nehalem, SandyBridge monitors and command passing
179based on turbostat.8 from Len Brown <len.brown@intel.com>