1 .TH TURBOSTAT 8 1 .TH TURBOSTAT 8
2 .SH NAME 2 .SH NAME
3 turbostat \- Report processor frequency and id 3 turbostat \- Report processor frequency and idle statistics
4 .SH SYNOPSIS 4 .SH SYNOPSIS
5 .ft B 5 .ft B
6 .B turbostat 6 .B turbostat
7 .RB [ Options ] 7 .RB [ Options ]
8 .RB command 8 .RB command
9 .br 9 .br
10 .B turbostat 10 .B turbostat
11 .RB [ Options ] 11 .RB [ Options ]
12 .RB [ "\--interval seconds" ] 12 .RB [ "\--interval seconds" ]
13 .SH DESCRIPTION 13 .SH DESCRIPTION
14 \fBturbostat \fP reports processor topology, f 14 \fBturbostat \fP reports processor topology, frequency,
15 idle power-state statistics, temperature and p 15 idle power-state statistics, temperature and power on X86 processors.
16 There are two ways to invoke turbostat. 16 There are two ways to invoke turbostat.
17 The first method is to supply a 17 The first method is to supply a
18 \fBcommand\fP, which is forked and statistics 18 \fBcommand\fP, which is forked and statistics are printed
19 in one-shot upon its completion. 19 in one-shot upon its completion.
20 The second method is to omit the command, 20 The second method is to omit the command,
21 and turbostat displays statistics every 5 seco 21 and turbostat displays statistics every 5 seconds interval.
22 The 5-second interval can be changed using the 22 The 5-second interval can be changed using the --interval option.
23 .PP 23 .PP
24 Some information is not available on older pro 24 Some information is not available on older processors.
25 .SS Options 25 .SS Options
26 Options can be specified with a single or doub 26 Options can be specified with a single or double '-', and only as much of the option
27 name as necessary to disambiguate it from othe 27 name as necessary to disambiguate it from others is necessary. Note that options are case-sensitive.
28 .PP 28 .PP
29 \fB--add attributes\fP add column with counter 29 \fB--add attributes\fP add column with counter having specified 'attributes'. The 'location' attribute is required, all others are optional.
30 .nf 30 .nf
31 location: {\fBmsrDDD\fP | \fBmsr0xXXX\ !! 31 location: {\fBmsrDDD\fP | \fBmsr0xXXX\fP | \fB/sys/path...\fP}
32 msrDDD is a decimal offset, eg 32 msrDDD is a decimal offset, eg. msr16
33 msr0xXXX is a hex offset, eg. 33 msr0xXXX is a hex offset, eg. msr0x10
34 /sys/path... is an absolute pa 34 /sys/path... is an absolute path to a sysfs attribute
35 <device> is a perf device from <<
36 <event> is a perf event for gi <<
37 perf/cstate_core/c1-re <<
38 35
39 scope: {\fBcpu\fP | \fBcore\fP | \fBpa 36 scope: {\fBcpu\fP | \fBcore\fP | \fBpackage\fP}
40 sample and print the counter f 37 sample and print the counter for every cpu, core, or package.
41 default: cpu 38 default: cpu
42 39
43 size: {\fBu32\fP | \fBu64\fP } 40 size: {\fBu32\fP | \fBu64\fP }
44 MSRs are read as 64-bits, u32 41 MSRs are read as 64-bits, u32 truncates the displayed value to 32-bits.
45 default: u64 42 default: u64
46 43
47 format: {\fBraw\fP | \fBdelta\fP | \fB 44 format: {\fBraw\fP | \fBdelta\fP | \fBpercent\fP}
48 'raw' shows the MSR contents i 45 'raw' shows the MSR contents in hex.
49 'delta' shows the difference i 46 'delta' shows the difference in values during the measurement interval.
50 'percent' shows the delta as a 47 'percent' shows the delta as a percentage of the cycles elapsed.
51 default: delta 48 default: delta
52 49
53 name: "name_string" 50 name: "name_string"
54 Any string that does not match 51 Any string that does not match a key-word above is used
55 as the column header. 52 as the column header.
56 .fi 53 .fi
57 .PP 54 .PP
58 \fB--add pmt,[attr_name=attr_value, ...]\fP ad <<
59 .nf <<
60 name="name_string" <<
61 For column header. <<
62 <<
63 type={\fBraw\fP} <<
64 'raw' shows the counter conten <<
65 default: raw <<
66 <<
67 format={\fBraw\fP | \fBdelta\fP} <<
68 'raw' shows the counter conten <<
69 'delta' shows the difference i <<
70 default: raw <<
71 <<
72 domain={\fBcpu%u\fP | \fBcore%u\fP | \ <<
73 'cpu' per cpu/thread counter. <<
74 'core' per core counter. <<
75 'package' per package counter. <<
76 '%u' denotes id of the domain <<
77 <<
78 offset=\fB%u\fP <<
79 '%u' offset within the PMT MMI <<
80 <<
81 lsb=\fB%u\fP <<
82 '%u' least significant bit wit <<
83 <<
84 msb=\fB%u\fP <<
85 '%u' most significant bit with <<
86 <<
87 guid=\fB%x\fP <<
88 '%x' hex identifier of the PMT <<
89 .fi <<
90 .PP <<
91 \fB--cpu cpu-set\fP limit output to system sum 55 \fB--cpu cpu-set\fP limit output to system summary plus the specified cpu-set. If cpu-set is the string "core", then the system summary plus the first CPU in each core are printed -- eg. subsequent HT siblings are not printed. Or if cpu-set is the string "package", then the system summary plus the first CPU in each package is printed. Otherwise, the system summary plus the specified set of CPUs are printed. The cpu-set is ordered from low to high, comma delimited with ".." and "-" permitted to denote a range. eg. 1,2,8,14..17,21-44
92 .PP 56 .PP
93 \fB--hide column\fP do not show the specified !! 57 \fB--hide column\fP do not show the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. Use "--hide sysfs" to hide the sysfs statistics columns as a group.
94 .PP 58 .PP
95 \fB--enable column\fP show the specified built 59 \fB--enable column\fP show the specified built-in columns, which are otherwise disabled, by default. Currently the only built-in counters disabled by default are "usec", "Time_Of_Day_Seconds", "APIC" and "X2APIC".
96 The column name "all" can be used to enable al 60 The column name "all" can be used to enable all disabled-by-default built-in counters.
97 .PP 61 .PP
98 \fB--show column\fP show only the specified bu !! 62 \fB--show column\fP show only the specified built-in columns. May be invoked multiple times, or with a comma-separated list of column names. Use "--show sysfs" to show the sysfs statistics columns as a group.
99 .PP <<
100 \fB--show CATEGORY --hide CATEGORY\fP Show an <<
101 .PP 63 .PP
102 \fB--Dump\fP displays the raw counter values. 64 \fB--Dump\fP displays the raw counter values.
103 .PP 65 .PP
104 \fB--quiet\fP Do not decode and print the syst 66 \fB--quiet\fP Do not decode and print the system configuration header information.
105 .PP 67 .PP
106 \fB--no-msr\fP Disable all the uses of the MSR <<
107 .PP <<
108 \fB--no-perf\fP Disable all the uses of the pe <<
109 .PP <<
110 \fB--interval seconds\fP overrides the default 68 \fB--interval seconds\fP overrides the default 5.0 second measurement interval.
111 .PP 69 .PP
112 \fB--num_iterations num\fP number of the measu 70 \fB--num_iterations num\fP number of the measurement iterations.
113 .PP 71 .PP
114 \fB--out output_file\fP turbostat output is wr 72 \fB--out output_file\fP turbostat output is written to the specified output_file.
115 The file is truncated if it already exists, an 73 The file is truncated if it already exists, and it is created if it does not exist.
116 .PP 74 .PP
117 \fB--help\fP displays usage for the most commo 75 \fB--help\fP displays usage for the most common parameters.
118 .PP 76 .PP
119 \fB--Joules\fP displays energy in Joules, rath 77 \fB--Joules\fP displays energy in Joules, rather than dividing Joules by time to print power in Watts.
120 .PP 78 .PP
121 \fB--list\fP display column header names avail 79 \fB--list\fP display column header names available for use by --show and --hide, then exit.
122 .PP 80 .PP
123 \fB--Summary\fP limits output to a 1-line Syst 81 \fB--Summary\fP limits output to a 1-line System Summary for each interval.
124 .PP 82 .PP
125 \fB--TCC temperature\fP sets the Thermal Contr 83 \fB--TCC temperature\fP sets the Thermal Control Circuit temperature for systems which do not export that value. This is used for making sense of the Digital Thermal Sensor outputs, as they return degrees Celsius below the TCC activation temperature.
126 .PP 84 .PP
127 \fB--version\fP displays the version. 85 \fB--version\fP displays the version.
128 .PP 86 .PP
129 The \fBcommand\fP parameter forks \fBcommand\f 87 The \fBcommand\fP parameter forks \fBcommand\fP, and upon its exit,
130 displays the statistics gathered since it was 88 displays the statistics gathered since it was forked.
131 .PP 89 .PP
132 .SH ROW DESCRIPTIONS 90 .SH ROW DESCRIPTIONS
133 The system configuration dump (if --quiet is n 91 The system configuration dump (if --quiet is not used) is followed by statistics. The first row of the statistics labels the content of each column (below). The second row of statistics is the system summary line. The system summary line has a '-' in the columns for the Package, Core, and CPU. The contents of the system summary line depends on the type of column. Columns that count items (eg. IRQ) show the sum across all CPUs in the system. Columns that show a percentage show the average across all CPUs in the system. Columns that dump raw MSR values simply show 0 in the summary. After the system summary row, each row describes a specific Package/Core/CPU. Note that if the --cpu parameter is used to limit which specific CPUs are displayed, turbostat will still collect statistics for all CPUs in the system and will still show the system summary for all CPUs in the system.
134 .SH COLUMN DESCRIPTIONS 92 .SH COLUMN DESCRIPTIONS
135 .PP !! 93 .nf
136 \fBusec\fP For each CPU, the number of microse 94 \fBusec\fP For each CPU, the number of microseconds elapsed during counter collection, including thread migration -- if any. This counter is disabled by default, and is enabled with "--enable usec", or --debug. On the summary row, usec refers to the total elapsed time to collect the counters on all cpus.
137 .PP <<
138 \fBTime_Of_Day_Seconds\fP For each CPU, the ge 95 \fBTime_Of_Day_Seconds\fP For each CPU, the gettimeofday(2) value (seconds.subsec since Epoch) when the counters ending the measurement interval were collected. This column is disabled by default, and can be enabled with "--enable Time_Of_Day_Seconds" or "--debug". On the summary row, Time_Of_Day_Seconds refers to the timestamp following collection of counters on the last CPU.
139 .PP <<
140 \fBCore\fP processor core number. Note that m 96 \fBCore\fP processor core number. Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology (HT).
141 .PP <<
142 \fBCPU\fP Linux CPU (logical processor) number 97 \fBCPU\fP Linux CPU (logical processor) number. Yes, it is okay that on many systems the CPUs are not listed in numerical order -- for efficiency reasons, turbostat runs in topology order, so HT siblings appear together.
143 .PP <<
144 \fBPackage\fP processor package number -- not 98 \fBPackage\fP processor package number -- not present on systems with a single processor package.
145 .PP <<
146 \fBAvg_MHz\fP number of cycles executed divide 99 \fBAvg_MHz\fP number of cycles executed divided by time elapsed. Note that this includes idle-time when 0 instructions are executed.
147 .PP <<
148 \fBBusy%\fP percent of the measurement interva 100 \fBBusy%\fP percent of the measurement interval that the CPU executes instructions, aka. % of time in "C0" state.
149 .PP <<
150 \fBBzy_MHz\fP average clock rate while the CPU 101 \fBBzy_MHz\fP average clock rate while the CPU was not idle (ie. in "c0" state).
151 .PP <<
152 \fBTSC_MHz\fP average MHz that the TSC ran dur 102 \fBTSC_MHz\fP average MHz that the TSC ran during the entire interval.
153 .PP <<
154 \fBIRQ\fP The number of interrupts serviced by 103 \fBIRQ\fP The number of interrupts serviced by that CPU during the measurement interval. The system total line is the sum of interrupts serviced across all CPUs. turbostat parses /proc/interrupts to generate this summary.
155 .PP <<
156 \fBSMI\fP The number of System Management Inte 104 \fBSMI\fP The number of System Management Interrupts serviced CPU during the measurement interval. While this counter is actually per-CPU, SMI are triggered on all processors, so the number should be the same for all CPUs.
157 .PP <<
158 \fBC1, C2, C3...\fP The number times Linux req 105 \fBC1, C2, C3...\fP The number times Linux requested the C1, C2, C3 idle state during the measurement interval. The system summary line shows the sum for all CPUs. These are C-state names as exported in /sys/devices/system/cpu/cpu*/cpuidle/state*/name. While their names are generic, their attributes are processor specific. They the system description section of output shows what MWAIT sub-states they are mapped to on each system.
159 .PP <<
160 \fBC1%, C2%, C3%\fP The residency percentage t 106 \fBC1%, C2%, C3%\fP The residency percentage that Linux requested C1, C2, C3.... The system summary is the average of all CPUs in the system. Note that these are software, reflecting what was requested. The hardware counters reflect what was actually achieved.
161 .PP <<
162 \fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the 107 \fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the percentage residency in hardware core idle states. These numbers are from hardware residency counters.
163 .PP <<
164 \fBCoreTmp\fP Degrees Celsius reported by the 108 \fBCoreTmp\fP Degrees Celsius reported by the per-core Digital Thermal Sensor.
165 .PP <<
166 \fBPkgTmp\fP Degrees Celsius reported by the p 109 \fBPkgTmp\fP Degrees Celsius reported by the per-package Package Thermal Monitor.
167 .PP !! 110 \fBGFX%rc6\fP The percentage of time the GPU is in the "render C6" state, rc6, during the measurement interval. From /sys/class/drm/card0/power/rc6_residency_ms.
168 \fBGFX%rc6\fP The percentage of time the GPU i !! 111 \fBGFXMHz\fP Instantaneous snapshot of what sysfs presents at the end of the measurement interval. From /sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz.
169 .PP <<
170 \fBGFXMHz\fP Instantaneous snapshot of what sy <<
171 .PP <<
172 \fBGFXAMHz\fP Instantaneous snapshot of what s <<
173 .PP <<
174 \fBSAM%mc6\fP The percentage of time the SA Me <<
175 .PP <<
176 \fBSAMMHz\fP Instantaneous snapshot of what sy <<
177 .PP <<
178 \fBSAMAMHz\fP Instantaneous snapshot of what s <<
179 .PP <<
180 \fBPkg%pc2, Pkg%pc3, Pkg%pc6, Pkg%pc7\fP perce 112 \fBPkg%pc2, Pkg%pc3, Pkg%pc6, Pkg%pc7\fP percentage residency in hardware package idle states. These numbers are from hardware residency counters.
181 .PP <<
182 \fBPkgWatt\fP Watts consumed by the whole pack 113 \fBPkgWatt\fP Watts consumed by the whole package.
183 .PP <<
184 \fBCorWatt\fP Watts consumed by the core part 114 \fBCorWatt\fP Watts consumed by the core part of the package.
185 .PP <<
186 \fBGFXWatt\fP Watts consumed by the Graphics p 115 \fBGFXWatt\fP Watts consumed by the Graphics part of the package -- available only on client processors.
187 .PP <<
188 \fBRAMWatt\fP Watts consumed by the DRAM DIMMS 116 \fBRAMWatt\fP Watts consumed by the DRAM DIMMS -- available only on server processors.
189 .PP <<
190 \fBPKG_%\fP percent of the interval that RAPL 117 \fBPKG_%\fP percent of the interval that RAPL throttling was active on the Package. Note that the system summary is the sum of the package throttling time, and thus may be higher than 100% on a multi-package system. Note that the meaning of this field is model specific. For example, some hardware increments this counter when RAPL responds to thermal limits, but does not increment this counter when RAPL responds to power limits. Comparing PkgWatt and PkgTmp to system limits is necessary.
191 .PP <<
192 \fBRAM_%\fP percent of the interval that RAPL 118 \fBRAM_%\fP percent of the interval that RAPL throttling was active on DRAM.
193 .PP !! 119 .fi
194 \fBUncMHz\fP per-package uncore MHz, instantan <<
195 .PP <<
196 \fBUMHz1.0\fP per-package uncore MHz for domai <<
197 .SH TOO MUCH INFORMATION EXAMPLE 120 .SH TOO MUCH INFORMATION EXAMPLE
198 By default, turbostat dumps all possible infor 121 By default, turbostat dumps all possible information -- a system configuration header, followed by columns for all counters.
199 This is ideal for remote debugging, use the "- 122 This is ideal for remote debugging, use the "--out" option to save everything to a text file, and get that file to the expert helping you debug.
200 .PP 123 .PP
201 When you are not interested in all that inform 124 When you are not interested in all that information, and there are several ways to see only what you want. First the "--quiet" option will skip the configuration information, and turbostat will show only the counter columns. Second, you can reduce the columns with the "--hide" and "--show" options. If you use the "--show" option, then turbostat will show only the columns you list. If you use the "--hide" option, turbostat will show all columns, except the ones you list.
202 .PP 125 .PP
203 To find out what columns are available for --s !! 126 To find out what columns are available for --show and --hide, the "--list" option is available. For convenience, the special strings "sysfs" can be used to refer to all of the sysfs C-state counters at once:
204 .PP <<
205 .nf 127 .nf
206 sudo ./turbostat --show sysfs --quiet sleep 10 128 sudo ./turbostat --show sysfs --quiet sleep 10
207 10.003837 sec 129 10.003837 sec
208 C1 C1E C3 C6 C7s 130 C1 C1E C3 C6 C7s C1% C1E% C3% C6% C7s%
209 4 21 2 2 459 131 4 21 2 2 459 0.14 0.82 0.00 0.00 98.93
210 1 17 2 2 130 132 1 17 2 2 130 0.00 0.02 0.00 0.00 99.80
211 0 0 0 0 31 133 0 0 0 0 31 0.00 0.00 0.00 0.00 99.95
212 2 1 0 0 52 134 2 1 0 0 52 1.14 6.49 0.00 0.00 92.21
213 1 2 0 0 52 135 1 2 0 0 52 0.00 0.08 0.00 0.00 99.86
214 0 0 0 0 71 136 0 0 0 0 71 0.00 0.00 0.00 0.00 99.89
215 0 0 0 0 25 137 0 0 0 0 25 0.00 0.00 0.00 0.00 99.96
216 0 0 0 0 74 138 0 0 0 0 74 0.00 0.00 0.00 0.00 99.94
217 0 1 0 0 24 139 0 1 0 0 24 0.00 0.00 0.00 0.00 99.84
218 .fi 140 .fi
219 .PP 141 .PP
220 .SH ONE SHOT COMMAND EXAMPLE 142 .SH ONE SHOT COMMAND EXAMPLE
221 If turbostat is invoked with a command, it wil 143 If turbostat is invoked with a command, it will fork that command
222 and output the statistics gathered after the c 144 and output the statistics gathered after the command exits.
223 In this case, turbostat output goes to stderr, 145 In this case, turbostat output goes to stderr, by default.
224 Output can instead be saved to a file using th 146 Output can instead be saved to a file using the --out option.
225 In this example, the "sleep 10" command is for 147 In this example, the "sleep 10" command is forked, and turbostat waits for it to complete before saving all statistics into "ts.out". Note that "sleep 10" is not part of turbostat, but is simply an example of a command that turbostat can fork. The "ts.out" file is what you want to edit in a very wide window, paste into a spreadsheet, or attach to a bugzilla entry.
226 148
227 .nf 149 .nf
228 [root@hsw]# ./turbostat -o ts.out sleep 10 150 [root@hsw]# ./turbostat -o ts.out sleep 10
229 [root@hsw]# 151 [root@hsw]#
230 .fi 152 .fi
231 153
232 .SH PERIODIC INTERVAL EXAMPLE 154 .SH PERIODIC INTERVAL EXAMPLE
233 Without a command to fork, turbostat displays 155 Without a command to fork, turbostat displays statistics ever 5 seconds.
234 Periodic output goes to stdout, by default, un 156 Periodic output goes to stdout, by default, unless --out is used to specify an output file.
235 The 5-second interval can be changed with the 157 The 5-second interval can be changed with the "-i sec" option.
236 .nf 158 .nf
237 sudo turbostat --quiet --show CPU,frequency !! 159 sudo ./turbostat --quiet --hide sysfs,IRQ,SMI,CoreTmp,PkgTmp,GFX%rc6,GFXMHz,PkgWatt,CorWatt,GFXWatt
238 Core CPU Avg_MHz Busy% Bzy_MH !! 160 Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7
239 - - 524 12.48 4198 !! 161 - - 488 12.52 3900 3498 12.50 0.00 0.00 74.98
240 0 0 4 0.09 4081 !! 162 0 0 5 0.13 3900 3498 99.87 0.00 0.00 0.00
241 0 4 1 0.02 4063 !! 163 0 4 3897 99.99 3900 3498 0.01
242 1 1 2 0.06 4063 !! 164 1 1 0 0.00 3856 3498 0.01 0.00 0.00 99.98
243 1 5 2 0.05 4070 !! 165 1 5 0 0.00 3861 3498 0.01
244 2 2 4178 99.52 4199 !! 166 2 2 1 0.02 3889 3498 0.03 0.00 0.00 99.95
245 2 6 3 0.08 4159 !! 167 2 6 0 0.00 3863 3498 0.05
246 3 3 1 0.04 4046 !! 168 3 3 0 0.01 3869 3498 0.02 0.00 0.00 99.97
247 3 7 0 0.01 3989 !! 169 3 7 0 0.00 3878 3498 0.03
248 Core CPU Avg_MHz Busy% Bzy_MH !! 170 Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7
249 - - 525 12.52 4198 !! 171 - - 491 12.59 3900 3498 12.42 0.00 0.00 74.99
250 0 0 4 0.10 4051 !! 172 0 0 27 0.69 3900 3498 99.31 0.00 0.00 0.00
251 0 4 2 0.04 3993 !! 173 0 4 3898 99.99 3900 3498 0.01
252 1 1 3 0.07 4054 !! 174 1 1 0 0.00 3883 3498 0.01 0.00 0.00 99.99
253 1 5 4 0.10 4018 !! 175 1 5 0 0.00 3898 3498 0.01
254 2 2 4178 99.51 4199 !! 176 2 2 0 0.01 3889 3498 0.02 0.00 0.00 99.98
255 2 6 4 0.09 4143 !! 177 2 6 0 0.00 3889 3498 0.02
256 3 3 2 0.06 4026 !! 178 3 3 0 0.00 3856 3498 0.01 0.00 0.00 99.99
257 3 7 7 0.17 4074 !! 179 3 7 0 0.00 3897 3498 0.01
258 .fi 180 .fi
259 This example also shows the use of the --show !! 181 This example also shows the use of the --hide option to skip columns that are not wanted.
>> 182 Note that cpu4 in this example is 99.99% busy, while the other CPUs are all under 1% busy.
>> 183 Notice that cpu4's HT sibling is cpu0, which is under 1% busy, but can get into CPU%c1 only,
>> 184 because its cpu4's activity on shared hardware keeps it from entering a deeper C-state.
260 185
261 .SH SYSTEM CONFIGURATION INFORMATION EXAMPLE 186 .SH SYSTEM CONFIGURATION INFORMATION EXAMPLE
262 187
263 By default, turbostat always dumps system conf 188 By default, turbostat always dumps system configuration information
264 before taking measurements. In the example ab 189 before taking measurements. In the example above, "--quiet" is used
265 to suppress that output. Here is an example o 190 to suppress that output. Here is an example of the configuration information:
266 .nf 191 .nf
267 turbostat version 2022.04.16 - Len Brown <lenb@ !! 192 turbostat version 2017.02.15 - Len Brown <lenb@kernel.org>
268 Kernel command line: BOOT_IMAGE=/boot/vmlinuz- !! 193 CPUID(0): GenuineIntel 13 CPUID levels; family:model:stepping 0x6:3c:3 (6:60:3)
269 CPUID(0): GenuineIntel 0x16 CPUID levels !! 194 CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI-TM TM
270 CPUID(1): family:model:stepping 0x6:9e:9 (6:15 !! 195 CPUID(6): APERF, TURBO, DTS, PTM, No-HWP, No-HWPnotify, No-HWPwindow, No-HWPepp, No-HWPpkg, EPB
271 CPUID(0x80000000): max_extended_levels: 0x8000 !! 196 cpu4: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST No-MWAIT PREFETCH TURBO)
272 CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI !! 197 CPUID(7): No-SGX
273 CPUID(6): APERF, TURBO, DTS, PTM, HWP, HWPnoti !! 198 cpu4: MSR_MISC_PWR_MGMT: 0x00400000 (ENable-EIST_Coordination DISable-EPB DISable-OOB)
274 cpu7: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EI !! 199 RAPL: 3121 sec. Joule Counter Range, at 84 Watts
275 CPUID(7): SGX !! 200 cpu4: MSR_PLATFORM_INFO: 0x80838f3012300
276 cpu7: MSR_IA32_FEATURE_CONTROL: 0x00000005 (Lo <<
277 CPUID(0x15): eax_crystal: 2 ebx_tsc: 258 ecx_c <<
278 TSC: 3096 MHz (24000000 Hz * 258 / 2 / 1000000 <<
279 CPUID(0x16): base_mhz: 3100 max_mhz: 4200 bus_ <<
280 cpu7: MSR_MISC_PWR_MGMT: 0x00401cc0 (ENable-EI <<
281 RAPL: 5825 sec. Joule Counter Range, at 45 Wat <<
282 cpu7: MSR_PLATFORM_INFO: 0x80839f1011f00 <<
283 8 * 100.0 = 800.0 MHz max efficiency frequency 201 8 * 100.0 = 800.0 MHz max efficiency frequency
284 31 * 100.0 = 3100.0 MHz base frequency !! 202 35 * 100.0 = 3500.0 MHz base frequency
285 cpu7: MSR_IA32_POWER_CTL: 0x002c005d (C1E auto !! 203 cpu4: MSR_IA32_POWER_CTL: 0x0004005d (C1E auto-promotion: DISabled)
286 cpu7: MSR_TURBO_RATIO_LIMIT: 0x2728292a !! 204 cpu4: MSR_TURBO_RATIO_LIMIT: 0x25262727
287 39 * 100.0 = 3900.0 MHz max turbo 4 active cor !! 205 37 * 100.0 = 3700.0 MHz max turbo 4 active cores
288 40 * 100.0 = 4000.0 MHz max turbo 3 active cor !! 206 38 * 100.0 = 3800.0 MHz max turbo 3 active cores
289 41 * 100.0 = 4100.0 MHz max turbo 2 active cor !! 207 39 * 100.0 = 3900.0 MHz max turbo 2 active cores
290 42 * 100.0 = 4200.0 MHz max turbo 1 active cor !! 208 39 * 100.0 = 3900.0 MHz max turbo 1 active cores
291 cpu7: MSR_CONFIG_TDP_NOMINAL: 0x0000001f (base !! 209 cpu4: MSR_CONFIG_TDP_NOMINAL: 0x00000023 (base_ratio=35)
292 cpu7: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 () !! 210 cpu4: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 ()
293 cpu7: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 () !! 211 cpu4: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 ()
294 cpu7: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( loc !! 212 cpu4: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1)
295 cpu7: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 ( !! 213 cpu4: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 (MAX_NON_TURBO_RATIO=0 lock=0)
296 cpu7: MSR_PKG_CST_CONFIG_CONTROL: 0x1e008008 ( !! 214 cpu4: MSR_PKG_CST_CONFIG_CONTROL: 0x1e000400 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, UNlocked: pkg-cstate-limit=0: pc0)
297 Uncore Frequency pkg0 die0: 800 - 3900 MHz (80 !! 215 cpu4: POLL: CPUIDLE CORE POLL IDLE
298 /dev/cpu_dma_latency: 2000000000 usec (default !! 216 cpu4: C1: MWAIT 0x00
299 current_driver: intel_idle !! 217 cpu4: C1E: MWAIT 0x01
300 current_governor: menu !! 218 cpu4: C3: MWAIT 0x10
301 current_governor_ro: menu !! 219 cpu4: C6: MWAIT 0x20
302 cpu7: POLL: CPUIDLE CORE POLL IDLE !! 220 cpu4: C7s: MWAIT 0x32
303 cpu7: C1: MWAIT 0x00 !! 221 cpu4: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2-Prefetch L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch)
304 cpu7: C1E: MWAIT 0x01 !! 222 cpu0: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced)
305 cpu7: C3: MWAIT 0x10 !! 223 cpu0: MSR_CORE_PERF_LIMIT_REASONS, 0x31200000 (Active: ) (Logged: Transitions, MultiCoreTurbo, Amps, Auto-HWP, )
306 cpu7: C6: MWAIT 0x20 !! 224 cpu0: MSR_GFX_PERF_LIMIT_REASONS, 0x00000000 (Active: ) (Logged: )
307 cpu7: C7s: MWAIT 0x33 !! 225 cpu0: MSR_RING_PERF_LIMIT_REASONS, 0x0d000000 (Active: ) (Logged: Amps, PkgPwrL1, PkgPwrL2, )
308 cpu7: C8: MWAIT 0x40 <<
309 cpu7: C9: MWAIT 0x50 <<
310 cpu7: C10: MWAIT 0x60 <<
311 cpu7: cpufreq driver: intel_pstate <<
312 cpu7: cpufreq governor: performance <<
313 cpufreq intel_pstate no_turbo: 0 <<
314 cpu7: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2 <<
315 cpu0: MSR_PM_ENABLE: 0x00000001 (HWP) <<
316 cpu0: MSR_HWP_CAPABILITIES: 0x01101f53 (high 8 <<
317 cpu0: MSR_HWP_REQUEST: 0x00005353 (min 83 max <<
318 cpu0: MSR_HWP_INTERRUPT: 0x00000001 (EN_Guaran <<
319 cpu0: MSR_HWP_STATUS: 0x00000004 (No-Guarantee <<
320 cpu0: EPB: 6 (balanced) <<
321 cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.12500 226 cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061 Joules, 0.000977 sec.)
322 cpu0: MSR_PKG_POWER_INFO: 0x00000168 (45 W TDP !! 227 cpu0: MSR_PKG_POWER_INFO: 0x000002a0 (84 W TDP, RAPL 0 - 0 W, 0.000000 sec.)
323 cpu0: MSR_PKG_POWER_LIMIT: 0x42820800218208 (U !! 228 cpu0: MSR_PKG_POWER_LIMIT: 0x428348001a82a0 (UNlocked)
324 cpu0: PKG Limit #1: ENabled (65.000 Watts, 64. !! 229 cpu0: PKG Limit #1: ENabled (84.000000 Watts, 8.000000 sec, clamp DISabled)
325 cpu0: PKG Limit #2: ENabled (65.000 Watts, 0.0 !! 230 cpu0: PKG Limit #2: ENabled (105.000000 Watts, 0.002441* sec, clamp DISabled)
326 cpu0: MSR_VR_CURRENT_CONFIG: 0x00000000 <<
327 cpu0: PKG Limit #4: 0.000000 Watts (UNlocked) <<
328 cpu0: MSR_DRAM_POWER_LIMIT: 0x5400de00000000 ( <<
329 cpu0: DRAM Limit: DISabled (0.000 Watts, 0.000 <<
330 cpu0: MSR_PP0_POLICY: 0 231 cpu0: MSR_PP0_POLICY: 0
331 cpu0: MSR_PP0_POWER_LIMIT: 0x00000000 (UNlocke 232 cpu0: MSR_PP0_POWER_LIMIT: 0x00000000 (UNlocked)
332 cpu0: Cores Limit: DISabled (0.000 Watts, 0.00 !! 233 cpu0: Cores Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled)
333 cpu0: MSR_PP1_POLICY: 0 234 cpu0: MSR_PP1_POLICY: 0
334 cpu0: MSR_PP1_POWER_LIMIT: 0x00000000 (UNlocke 235 cpu0: MSR_PP1_POWER_LIMIT: 0x00000000 (UNlocked)
335 cpu0: GFX Limit: DISabled (0.000 Watts, 0.0009 !! 236 cpu0: GFX Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled)
336 cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00640000 !! 237 cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00641400 (100 C)
337 cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x8820080 !! 238 cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x884c0800 (24 C)
338 cpu0: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x0000 !! 239 cpu0: MSR_IA32_THERM_STATUS: 0x884c0000 (24 C +/- 1)
339 cpu7: MSR_PKGC3_IRTL: 0x0000884e (valid, 79872 !! 240 cpu1: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1)
340 cpu7: MSR_PKGC6_IRTL: 0x00008876 (valid, 12083 !! 241 cpu2: MSR_IA32_THERM_STATUS: 0x884e0000 (22 C +/- 1)
341 cpu7: MSR_PKGC7_IRTL: 0x00008894 (valid, 15155 !! 242 cpu3: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1)
342 cpu7: MSR_PKGC8_IRTL: 0x000088fa (valid, 25600 !! 243 cpu4: MSR_PKGC3_IRTL: 0x00008842 (valid, 67584 ns)
343 cpu7: MSR_PKGC9_IRTL: 0x0000894c (valid, 33996 !! 244 cpu4: MSR_PKGC6_IRTL: 0x00008873 (valid, 117760 ns)
344 cpu7: MSR_PKGC10_IRTL: 0x00008bf2 (valid, 1034 !! 245 cpu4: MSR_PKGC7_IRTL: 0x00008891 (valid, 148480 ns)
345 .fi 246 .fi
346 .PP <<
347 The \fBmax efficiency\fP frequency, a.k.a. Low 247 The \fBmax efficiency\fP frequency, a.k.a. Low Frequency Mode, is the frequency
348 available at the minimum package voltage. The 248 available at the minimum package voltage. The \fBTSC frequency\fP is the base
349 frequency of the processor -- this should matc 249 frequency of the processor -- this should match the brand string
350 in /proc/cpuinfo. This base frequency 250 in /proc/cpuinfo. This base frequency
351 should be sustainable on all CPUs indefinitely 251 should be sustainable on all CPUs indefinitely, given nominal power and cooling.
352 The remaining rows show what maximum turbo fre 252 The remaining rows show what maximum turbo frequency is possible
353 depending on the number of idle cores. Note t 253 depending on the number of idle cores. Note that not all information is
354 available on all processors. 254 available on all processors.
355 .SH ADD COUNTER EXAMPLE 255 .SH ADD COUNTER EXAMPLE
356 Here we limit turbostat to showing just the CP 256 Here we limit turbostat to showing just the CPU number for cpu0 - cpu3.
357 We add a counter showing the 32-bit raw value 257 We add a counter showing the 32-bit raw value of MSR 0x199 (MSR_IA32_PERF_CTL),
358 labeling it with the column header, "PRF_CTRL" 258 labeling it with the column header, "PRF_CTRL", and display it only once,
359 after the conclusion of a 0.1 second sleep. !! 259 afte the conclusion of a 0.1 second sleep.
360 .nf 260 .nf
361 sudo ./turbostat --quiet --cpu 0-3 --show CPU 261 sudo ./turbostat --quiet --cpu 0-3 --show CPU --add msr0x199,u32,raw,PRF_CTRL sleep .1
362 0.101604 sec 262 0.101604 sec
363 CPU PRF_CTRL 263 CPU PRF_CTRL
364 - 0x00000000 264 - 0x00000000
365 0 0x00000c00 265 0 0x00000c00
366 1 0x00000800 266 1 0x00000800
367 2 0x00000a00 267 2 0x00000a00
368 3 0x00000800 268 3 0x00000800
369 269
370 .fi 270 .fi
371 271
372 .SH ADD PERF COUNTER EXAMPLE <<
373 Here we limit turbostat to showing just the CP <<
374 We add a counter showing time spent in C1 core <<
375 labeling it with the column header, "pCPU%c1", <<
376 after the conclusion of 0.1 second sleep. <<
377 We also show CPU%c1 built-in counter that shou <<
378 .nf <<
379 sudo ./turbostat --quiet --cpu 0-3 --show CPU, <<
380 0.102448 sec <<
381 CPU pCPU%c1 CPU%c1 <<
382 - 34.89 34.89 <<
383 0 45.99 45.99 <<
384 1 45.94 45.94 <<
385 2 23.83 23.83 <<
386 3 23.84 23.84 <<
387 <<
388 .fi <<
389 <<
390 .SH ADD PMT COUNTER EXAMPLE <<
391 Here we limit turbostat to showing just the CP <<
392 We add two counters, showing crystal clock cou <<
393 All the parameters passed are based on the met <<
394 <<
395 For the crystal clock count, we <<
396 label it with the column header, "XTAL", <<
397 we set the type to 'raw', to read the number o <<
398 we set the format to 'delta', to display the d <<
399 we set the domain to 'package0', to collect it <<
400 we set the offset to '0', which is a offset of <<
401 we set the lsb and msb to cover all 64 bits of <<
402 and finally we set the guid to '0x1a067102', t <<
403 <<
404 For the DC6 residency counter, we <<
405 label it with the column header, "Die%c6", <<
406 we set the type to 'txtal_time', to obtain the <<
407 we set the format to 'delta', to display the d <<
408 we set the domain to 'package0', to collect it <<
409 we set the offset to '0', which is a offset of <<
410 we set the lsb and msb to cover all 64 bits of <<
411 and finally we set the guid to '0x1a067102', t <<
412 <<
413 .nf <<
414 sudo ./turbostat --quiet --cpu 0 --show CPU -- <<
415 0.104352 sec <<
416 CPU XTAL Die%c6 <<
417 - 0x0000006d4d957ca7 0.00 <<
418 0 0x0000006d4d957ca7 0.00 <<
419 0.102448 sec <<
420 .fi <<
421 <<
422 .SH INPUT 272 .SH INPUT
423 273
424 For interval-mode, turbostat will immediately 274 For interval-mode, turbostat will immediately end the current interval
425 when it sees a newline on standard input. 275 when it sees a newline on standard input.
426 turbostat will then start the next interval. 276 turbostat will then start the next interval.
427 Control-C will be send a SIGINT to turbostat, 277 Control-C will be send a SIGINT to turbostat,
428 which will immediately abort the program with 278 which will immediately abort the program with no further processing.
429 .SH SIGNALS 279 .SH SIGNALS
430 280
431 SIGINT will interrupt interval-mode. 281 SIGINT will interrupt interval-mode.
432 The end-of-interval data will be collected and 282 The end-of-interval data will be collected and displayed before turbostat exits.
433 283
434 SIGUSR1 will end current interval, 284 SIGUSR1 will end current interval,
435 end-of-interval data will be collected and dis 285 end-of-interval data will be collected and displayed before turbostat
436 starts a new interval. 286 starts a new interval.
437 .SH NOTES 287 .SH NOTES
438 288
439 .B "turbostat " 289 .B "turbostat "
440 must be run as root. 290 must be run as root.
441 Alternatively, non-root users can be enabled t 291 Alternatively, non-root users can be enabled to run turbostat this way:
442 292
443 # setcap cap_sys_admin,cap_sys_rawio,cap_sys_n !! 293 # setcap cap_sys_rawio=ep ./turbostat
444 294
445 # chmod +r /dev/cpu/*/msr 295 # chmod +r /dev/cpu/*/msr
446 296
447 # chmod +r /dev/cpu_dma_latency <<
448 <<
449 .B "turbostat " 297 .B "turbostat "
450 reads hardware counters, but doesn't write the 298 reads hardware counters, but doesn't write them.
451 So it will not interfere with the OS or other 299 So it will not interfere with the OS or other programs, including
452 multiple invocations of itself. 300 multiple invocations of itself.
453 301
454 \fBturbostat \fP 302 \fBturbostat \fP
455 may work poorly on Linux-2.6.20 through 2.6.29 303 may work poorly on Linux-2.6.20 through 2.6.29,
456 as \fBacpi-cpufreq \fPperiodically cleared the 304 as \fBacpi-cpufreq \fPperiodically cleared the APERF and MPERF MSRs
457 in those kernels. 305 in those kernels.
458 306
459 AVG_MHz = APERF_delta/measurement_interval. T 307 AVG_MHz = APERF_delta/measurement_interval. This is the actual
460 number of elapsed cycles divided by the entire 308 number of elapsed cycles divided by the entire sample interval --
461 including idle time. Note that this calculati 309 including idle time. Note that this calculation is resilient
462 to systems lacking a non-stop TSC. 310 to systems lacking a non-stop TSC.
463 311
464 TSC_MHz = TSC_delta/measurement_interval. 312 TSC_MHz = TSC_delta/measurement_interval.
465 On a system with an invariant TSC, this value 313 On a system with an invariant TSC, this value will be constant
466 and will closely match the base frequency valu 314 and will closely match the base frequency value shown
467 in the brand string in /proc/cpuinfo. On a sy 315 in the brand string in /proc/cpuinfo. On a system where
468 the TSC stops in idle, TSC_MHz will drop 316 the TSC stops in idle, TSC_MHz will drop
469 below the processor's base frequency. 317 below the processor's base frequency.
470 318
471 Busy% = MPERF_delta/TSC_delta 319 Busy% = MPERF_delta/TSC_delta
472 320
473 Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/me !! 321 Bzy_MHz = TSC_delta/APERF_delta/MPERF_delta/measurement_interval
474 322
475 Note that these calculations depend on TSC_del 323 Note that these calculations depend on TSC_delta, so they
476 are not reliable during intervals when TSC_MHz 324 are not reliable during intervals when TSC_MHz is not running at the base frequency.
477 325
478 Turbostat data collection is not atomic. 326 Turbostat data collection is not atomic.
479 Extremely short measurement intervals (much le 327 Extremely short measurement intervals (much less than 1 second),
480 or system activity that prevents turbostat fro 328 or system activity that prevents turbostat from being able
481 to run on all CPUS to quickly collect data, wi 329 to run on all CPUS to quickly collect data, will result in
482 inconsistent results. 330 inconsistent results.
483 331
484 The APERF, MPERF MSRs are defined to count non 332 The APERF, MPERF MSRs are defined to count non-halted cycles.
485 Although it is not guaranteed by the architect 333 Although it is not guaranteed by the architecture, turbostat assumes
486 that they count at TSC rate, which is true on 334 that they count at TSC rate, which is true on all processors tested to date.
487 335
488 .SH REFERENCES 336 .SH REFERENCES
489 Volume 3B: System Programming Guide" 337 Volume 3B: System Programming Guide"
490 https://www.intel.com/products/processor/manua !! 338 http://www.intel.com/products/processor/manuals/
491 339
492 .SH FILES 340 .SH FILES
493 .ta 341 .ta
494 .nf 342 .nf
495 /dev/cpu/*/msr 343 /dev/cpu/*/msr
496 .fi 344 .fi
497 345
498 .SH "SEE ALSO" 346 .SH "SEE ALSO"
499 msr(4), vmstat(8) 347 msr(4), vmstat(8)
500 .PP 348 .PP
501 .SH AUTHOR 349 .SH AUTHOR
502 .nf 350 .nf
503 Written by Len Brown <len.brown@intel.com> 351 Written by Len Brown <len.brown@intel.com>
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.