1 .TH TURBOSTAT 8
2 .SH NAME
3 turbostat \- Report processor frequency and idle statistics
4 .SH SYNOPSIS
5 .ft B
6 .B turbostat
7 .RB [ Options ]
8 .RB command
9 .br
10 .B turbostat
11 .RB [ Options ]
12 .RB [ "\--interval seconds" ]
13 .SH DESCRIPTION
14 \fBturbostat \fP reports processor topology, frequency,
15 idle power-state statistics, temperature and power on X86 processors.
16 There are two ways to invoke turbostat.
17 The first method is to supply a
18 \fBcommand\fP, which is forked and statistics are printed
19 in one-shot upon its completion.
20 The second method is to omit the command,
21 and turbostat displays statistics every 5 seconds interval.
22 The 5-second interval can be changed using the --interval option.
23 .PP
24 Some information is not available on older processors.
25 .SS Options
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 others is necessary. Note that options are case-sensitive.
28 .PP
29 \fB--add attributes\fP add column with counter having specified 'attributes'. The 'location' attribute is required, all others are optional.
30 .nf
31 location: {\fBmsrDDD\fP | \fBmsr0xXXX\fP | \fB/sys/path...\fP | \fBperf/<device>/<event>\fP}
32 msrDDD is a decimal offset, eg. msr16
33 msr0xXXX is a hex offset, eg. msr0x10
34 /sys/path... is an absolute path to a sysfs attribute
35 <device> is a perf device from /sys/bus/event_source/devices/<device> eg. cstate_core
36 <event> is a perf event for given device from /sys/bus/event_source/devices/<device>/events/<event> eg. c1-residency
37 perf/cstate_core/c1-residency would then use /sys/bus/event_source/devices/cstate_core/events/c1-residency
38
39 scope: {\fBcpu\fP | \fBcore\fP | \fBpackage\fP}
40 sample and print the counter for every cpu, core, or package.
41 default: cpu
42
43 size: {\fBu32\fP | \fBu64\fP }
44 MSRs are read as 64-bits, u32 truncates the displayed value to 32-bits.
45 default: u64
46
47 format: {\fBraw\fP | \fBdelta\fP | \fBpercent\fP}
48 'raw' shows the MSR contents in hex.
49 'delta' shows the difference in values during the measurement interval.
50 'percent' shows the delta as a percentage of the cycles elapsed.
51 default: delta
52
53 name: "name_string"
54 Any string that does not match a key-word above is used
55 as the column header.
56 .fi
57 .PP
58 \fB--add pmt,[attr_name=attr_value, ...]\fP add column with a PMT (Intel Platform Monitoring Technology) counter in a similar way to --add option above, but require PMT metadata to be supplied to correctly read and display the counter. The metadata can be found in the Intel PMT XML files, hosted at https://github.com/intel/Intel-PMT. For a complete example see "ADD PMT COUNTER EXAMPLE".
59 .nf
60 name="name_string"
61 For column header.
62
63 type={\fBraw\fP}
64 'raw' shows the counter contents in hex.
65 default: raw
66
67 format={\fBraw\fP | \fBdelta\fP}
68 'raw' shows the counter contents in hex.
69 'delta' shows the difference in values during the measurement interval.
70 default: raw
71
72 domain={\fBcpu%u\fP | \fBcore%u\fP | \fBpackage%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 that the counter is associated with. For example core4 would mean that the counter is associated with core number 4.
77
78 offset=\fB%u\fP
79 '%u' offset within the PMT MMIO region.
80
81 lsb=\fB%u\fP
82 '%u' least significant bit within the 64 bit value read from 'offset'. Together with 'msb', used to form a read mask.
83
84 msb=\fB%u\fP
85 '%u' most significant bit within the 64 bit value read from 'offset'. Together with 'lsb', used to form a read mask.
86
87 guid=\fB%x\fP
88 '%x' hex identifier of the PMT MMIO region.
89 .fi
90 .PP
91 \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
93 \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.
94 .PP
95 \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 all disabled-by-default built-in counters.
97 .PP
98 \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.
99 .PP
100 \fB--show CATEGORY --hide CATEGORY\fP Show and hide also accept a single CATEGORY of columns: "all", "topology", "idle", "frequency", "power", "sysfs", "other".
101 .PP
102 \fB--Dump\fP displays the raw counter values.
103 .PP
104 \fB--quiet\fP Do not decode and print the system configuration header information.
105 .PP
106 \fB--no-msr\fP Disable all the uses of the MSR driver.
107 .PP
108 \fB--no-perf\fP Disable all the uses of the perf API.
109 .PP
110 \fB--interval seconds\fP overrides the default 5.0 second measurement interval.
111 .PP
112 \fB--num_iterations num\fP number of the measurement iterations.
113 .PP
114 \fB--out output_file\fP turbostat output is written to the specified output_file.
115 The file is truncated if it already exists, and it is created if it does not exist.
116 .PP
117 \fB--help\fP displays usage for the most common parameters.
118 .PP
119 \fB--Joules\fP displays energy in Joules, rather than dividing Joules by time to print power in Watts.
120 .PP
121 \fB--list\fP display column header names available for use by --show and --hide, then exit.
122 .PP
123 \fB--Summary\fP limits output to a 1-line System Summary for each interval.
124 .PP
125 \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
127 \fB--version\fP displays the version.
128 .PP
129 The \fBcommand\fP parameter forks \fBcommand\fP, and upon its exit,
130 displays the statistics gathered since it was forked.
131 .PP
132 .SH ROW DESCRIPTIONS
133 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
135 .PP
136 \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 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 multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology (HT).
141 .PP
142 \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 present on systems with a single processor package.
145 .PP
146 \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 interval that the CPU executes instructions, aka. % of time in "C0" state.
149 .PP
150 \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 during the entire interval.
153 .PP
154 \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 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 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 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 percentage residency in hardware core idle states. These numbers are from hardware residency counters.
163 .PP
164 \fBCoreTmp\fP Degrees Celsius reported by the per-core Digital Thermal Sensor.
165 .PP
166 \fBPkgTmp\fP Degrees Celsius reported by the per-package Package Thermal Monitor.
167 .PP
168 \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 or /sys/class/drm/card0/gt/gt0/rc6_residency_ms or /sys/class/drm/card0/device/tile0/gtN/gtidle/idle_residency_ms depending on the graphics driver being used.
169 .PP
170 \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 or /sys/class/drm/card0/gt_cur_freq_mhz or /sys/class/drm/card0/gt/gt0/rps_cur_freq_mhz or /sys/class/drm/card0/device/tile0/gtN/freq0/cur_freq depending on the graphics driver being used.
171 .PP
172 \fBGFXAMHz\fP Instantaneous snapshot of what sysfs presents at the end of the measurement interval. From /sys/class/graphics/fb0/device/drm/card0/gt_act_freq_mhz or /sys/class/drm/card0/gt_act_freq_mhz or /sys/class/drm/card0/gt/gt0/rps_act_freq_mhz or /sys/class/drm/card0/device/tile0/gtN/freq0/act_freq depending on the graphics driver being used.
173 .PP
174 \fBSAM%mc6\fP The percentage of time the SA Media is in the "module C6" state, mc6, during the measurement interval. From /sys/class/drm/card0/gt/gt1/rc6_residency_ms or /sys/class/drm/card0/device/tile0/gtN/gtidle/idle_residency_ms depending on the graphics driver being used.
175 .PP
176 \fBSAMMHz\fP Instantaneous snapshot of what sysfs presents at the end of the measurement interval. From /sys/class/drm/card0/gt/gt1/rps_cur_freq_mhz or /sys/class/drm/card0/device/tile0/gtN/freq0/cur_freq depending on the graphics driver being used.
177 .PP
178 \fBSAMAMHz\fP Instantaneous snapshot of what sysfs presents at the end of the measurement interval. From /sys/class/drm/card0/gt/gt1/rps_act_freq_mhz or /sys/class/drm/card0/device/tile0/gtN/freq0/act_freq depending on the graphics driver being used.
179 .PP
180 \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 package.
183 .PP
184 \fBCorWatt\fP Watts consumed by the core part of the package.
185 .PP
186 \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 -- available only on server processors.
189 .PP
190 \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 throttling was active on DRAM.
193 .PP
194 \fBUncMHz\fP per-package uncore MHz, instantaneous sample.
195 .PP
196 \fBUMHz1.0\fP per-package uncore MHz for domain=1 and fabric_cluster=0, instantaneous sample. System summary is the average of all packages.
197 .SH TOO MUCH INFORMATION EXAMPLE
198 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 "--out" option to save everything to a text file, and get that file to the expert helping you debug.
200 .PP
201 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
203 To find out what columns are available for --show and --hide, the "--list" option is available. Usually, the CATEGORY names above are used to refer to groups of counters. Also, for convenience, the special string "sysfs" can be used to refer to all of the sysfs C-state counters at once:
204 .PP
205 .nf
206 sudo ./turbostat --show sysfs --quiet sleep 10
207 10.003837 sec
208 C1 C1E C3 C6 C7s C1% C1E% C3% C6% C7s%
209 4 21 2 2 459 0.14 0.82 0.00 0.00 98.93
210 1 17 2 2 130 0.00 0.02 0.00 0.00 99.80
211 0 0 0 0 31 0.00 0.00 0.00 0.00 99.95
212 2 1 0 0 52 1.14 6.49 0.00 0.00 92.21
213 1 2 0 0 52 0.00 0.08 0.00 0.00 99.86
214 0 0 0 0 71 0.00 0.00 0.00 0.00 99.89
215 0 0 0 0 25 0.00 0.00 0.00 0.00 99.96
216 0 0 0 0 74 0.00 0.00 0.00 0.00 99.94
217 0 1 0 0 24 0.00 0.00 0.00 0.00 99.84
218 .fi
219 .PP
220 .SH ONE SHOT COMMAND EXAMPLE
221 If turbostat is invoked with a command, it will fork that command
222 and output the statistics gathered after the command exits.
223 In this case, turbostat output goes to stderr, by default.
224 Output can instead be saved to a file using the --out option.
225 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
227 .nf
228 [root@hsw]# ./turbostat -o ts.out sleep 10
229 [root@hsw]#
230 .fi
231
232 .SH PERIODIC INTERVAL EXAMPLE
233 Without a command to fork, turbostat displays statistics ever 5 seconds.
234 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 "-i sec" option.
236 .nf
237 sudo turbostat --quiet --show CPU,frequency
238 Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c7 UncMhz
239 - - 524 12.48 4198 3096 74.53 3800
240 0 0 4 0.09 4081 3096 98.88 3800
241 0 4 1 0.02 4063 3096
242 1 1 2 0.06 4063 3096 99.60
243 1 5 2 0.05 4070 3096
244 2 2 4178 99.52 4199 3096 0.00
245 2 6 3 0.08 4159 3096
246 3 3 1 0.04 4046 3096 99.66
247 3 7 0 0.01 3989 3096
248 Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c7 UncMhz
249 - - 525 12.52 4198 3096 74.54 3800
250 0 0 4 0.10 4051 3096 99.49 3800
251 0 4 2 0.04 3993 3096
252 1 1 3 0.07 4054 3096 99.56
253 1 5 4 0.10 4018 3096
254 2 2 4178 99.51 4199 3096 0.00
255 2 6 4 0.09 4143 3096
256 3 3 2 0.06 4026 3096 99.10
257 3 7 7 0.17 4074 3096
258 .fi
259 This example also shows the use of the --show option to show only the desired columns.
260
261 .SH SYSTEM CONFIGURATION INFORMATION EXAMPLE
262
263 By default, turbostat always dumps system configuration information
264 before taking measurements. In the example above, "--quiet" is used
265 to suppress that output. Here is an example of the configuration information:
266 .nf
267 turbostat version 2022.04.16 - Len Brown <lenb@kernel.org>
268 Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.18.0-rc6-00001-ge6891250e3b5 ...
269 CPUID(0): GenuineIntel 0x16 CPUID levels
270 CPUID(1): family:model:stepping 0x6:9e:9 (6:158:9) microcode 0xea
271 CPUID(0x80000000): max_extended_levels: 0x80000008
272 CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI-TM HT TM
273 CPUID(6): APERF, TURBO, DTS, PTM, HWP, HWPnotify, HWPwindow, HWPepp, No-HWPpkg, EPB
274 cpu7: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST MWAIT PREFETCH TURBO)
275 CPUID(7): SGX
276 cpu7: MSR_IA32_FEATURE_CONTROL: 0x00000005 (Locked )
277 CPUID(0x15): eax_crystal: 2 ebx_tsc: 258 ecx_crystal_hz: 0
278 TSC: 3096 MHz (24000000 Hz * 258 / 2 / 1000000)
279 CPUID(0x16): base_mhz: 3100 max_mhz: 4200 bus_mhz: 100
280 cpu7: MSR_MISC_PWR_MGMT: 0x00401cc0 (ENable-EIST_Coordination DISable-EPB DISable-OOB)
281 RAPL: 5825 sec. Joule Counter Range, at 45 Watts
282 cpu7: MSR_PLATFORM_INFO: 0x80839f1011f00
283 8 * 100.0 = 800.0 MHz max efficiency frequency
284 31 * 100.0 = 3100.0 MHz base frequency
285 cpu7: MSR_IA32_POWER_CTL: 0x002c005d (C1E auto-promotion: DISabled)
286 cpu7: MSR_TURBO_RATIO_LIMIT: 0x2728292a
287 39 * 100.0 = 3900.0 MHz max turbo 4 active cores
288 40 * 100.0 = 4000.0 MHz max turbo 3 active cores
289 41 * 100.0 = 4100.0 MHz max turbo 2 active cores
290 42 * 100.0 = 4200.0 MHz max turbo 1 active cores
291 cpu7: MSR_CONFIG_TDP_NOMINAL: 0x0000001f (base_ratio=31)
292 cpu7: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 ()
293 cpu7: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 ()
294 cpu7: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1)
295 cpu7: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 (MAX_NON_TURBO_RATIO=0 lock=0)
296 cpu7: MSR_PKG_CST_CONFIG_CONTROL: 0x1e008008 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, locked, pkg-cstate-limit=8 (unlimited))
297 Uncore Frequency pkg0 die0: 800 - 3900 MHz (800 - 3900 MHz)
298 /dev/cpu_dma_latency: 2000000000 usec (default)
299 current_driver: intel_idle
300 current_governor: menu
301 current_governor_ro: menu
302 cpu7: POLL: CPUIDLE CORE POLL IDLE
303 cpu7: C1: MWAIT 0x00
304 cpu7: C1E: MWAIT 0x01
305 cpu7: C3: MWAIT 0x10
306 cpu7: C6: MWAIT 0x20
307 cpu7: C7s: MWAIT 0x33
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-Prefetch L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch)
315 cpu0: MSR_PM_ENABLE: 0x00000001 (HWP)
316 cpu0: MSR_HWP_CAPABILITIES: 0x01101f53 (high 83 guar 31 eff 16 low 1)
317 cpu0: MSR_HWP_REQUEST: 0x00005353 (min 83 max 83 des 0 epp 0x0 window 0x0 pkg 0x0)
318 cpu0: MSR_HWP_INTERRUPT: 0x00000001 (EN_Guaranteed_Perf_Change, Dis_Excursion_Min)
319 cpu0: MSR_HWP_STATUS: 0x00000004 (No-Guaranteed_Perf_Change, No-Excursion_Min)
320 cpu0: EPB: 6 (balanced)
321 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, RAPL 0 - 0 W, 0.000000 sec.)
323 cpu0: MSR_PKG_POWER_LIMIT: 0x42820800218208 (UNlocked)
324 cpu0: PKG Limit #1: ENabled (65.000 Watts, 64.000000 sec, clamp ENabled)
325 cpu0: PKG Limit #2: ENabled (65.000 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 (UNlocked)
329 cpu0: DRAM Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled)
330 cpu0: MSR_PP0_POLICY: 0
331 cpu0: MSR_PP0_POWER_LIMIT: 0x00000000 (UNlocked)
332 cpu0: Cores Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled)
333 cpu0: MSR_PP1_POLICY: 0
334 cpu0: MSR_PP1_POWER_LIMIT: 0x00000000 (UNlocked)
335 cpu0: GFX Limit: DISabled (0.000 Watts, 0.000977 sec, clamp DISabled)
336 cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00640000 (100 C) (100 default - 0 offset)
337 cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x88200800 (68 C)
338 cpu0: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x00000003 (100 C, 100 C)
339 cpu7: MSR_PKGC3_IRTL: 0x0000884e (valid, 79872 ns)
340 cpu7: MSR_PKGC6_IRTL: 0x00008876 (valid, 120832 ns)
341 cpu7: MSR_PKGC7_IRTL: 0x00008894 (valid, 151552 ns)
342 cpu7: MSR_PKGC8_IRTL: 0x000088fa (valid, 256000 ns)
343 cpu7: MSR_PKGC9_IRTL: 0x0000894c (valid, 339968 ns)
344 cpu7: MSR_PKGC10_IRTL: 0x00008bf2 (valid, 1034240 ns)
345 .fi
346 .PP
347 The \fBmax efficiency\fP frequency, a.k.a. Low Frequency Mode, is the frequency
348 available at the minimum package voltage. The \fBTSC frequency\fP is the base
349 frequency of the processor -- this should match the brand string
350 in /proc/cpuinfo. This base frequency
351 should be sustainable on all CPUs indefinitely, given nominal power and cooling.
352 The remaining rows show what maximum turbo frequency is possible
353 depending on the number of idle cores. Note that not all information is
354 available on all processors.
355 .SH ADD COUNTER EXAMPLE
356 Here we limit turbostat to showing just the CPU number for cpu0 - cpu3.
357 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", and display it only once,
359 after the conclusion of a 0.1 second sleep.
360 .nf
361 sudo ./turbostat --quiet --cpu 0-3 --show CPU --add msr0x199,u32,raw,PRF_CTRL sleep .1
362 0.101604 sec
363 CPU PRF_CTRL
364 - 0x00000000
365 0 0x00000c00
366 1 0x00000800
367 2 0x00000a00
368 3 0x00000800
369
370 .fi
371
372 .SH ADD PERF COUNTER EXAMPLE
373 Here we limit turbostat to showing just the CPU number for cpu0 - cpu3.
374 We add a counter showing time spent in C1 core cstate,
375 labeling it with the column header, "pCPU%c1", and display it only once,
376 after the conclusion of 0.1 second sleep.
377 We also show CPU%c1 built-in counter that should show similar values.
378 .nf
379 sudo ./turbostat --quiet --cpu 0-3 --show CPU,CPU%c1 --add perf/cstate_core/c1-residency,cpu,delta,percent,pCPU%c1 sleep .1
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 CPU number 0.
392 We add two counters, showing crystal clock count and the DC6 residency.
393 All the parameters passed are based on the metadata found in the PMT XML files.
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 of clock ticks in hex,
398 we set the format to 'delta', to display the difference in ticks during the measurement interval,
399 we set the domain to 'package0', to collect it and associate it with the whole package number 0,
400 we set the offset to '0', which is a offset of the counter within the PMT MMIO region,
401 we set the lsb and msb to cover all 64 bits of the read 64 bit value,
402 and finally we set the guid to '0x1a067102', that identifies the PMT MMIO region to which the 'offset' is applied to read the counter value.
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 percent residency value
407 we set the format to 'delta', to display the difference in ticks during the measurement interval,
408 we set the domain to 'package0', to collect it and associate it with the whole package number 0,
409 we set the offset to '0', which is a offset of the counter within the PMT MMIO region,
410 we set the lsb and msb to cover all 64 bits of the read 64 bit value,
411 and finally we set the guid to '0x1a067102', that identifies the PMT MMIO region to which the 'offset' is applied to read the counter value.
412
413 .nf
414 sudo ./turbostat --quiet --cpu 0 --show CPU --add pmt,name=XTAL,type=raw,format=delta,domain=package0,offset=0,lsb=0,msb=63,guid=0x1a067102 --add pmt,name=Die%c6,type=txtal_time,format=delta,domain=package0,offset=120,lsb=0,msb=63,guid=0x1a067102
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
423
424 For interval-mode, turbostat will immediately end the current interval
425 when it sees a newline on standard input.
426 turbostat will then start the next interval.
427 Control-C will be send a SIGINT to turbostat,
428 which will immediately abort the program with no further processing.
429 .SH SIGNALS
430
431 SIGINT will interrupt interval-mode.
432 The end-of-interval data will be collected and displayed before turbostat exits.
433
434 SIGUSR1 will end current interval,
435 end-of-interval data will be collected and displayed before turbostat
436 starts a new interval.
437 .SH NOTES
438
439 .B "turbostat "
440 must be run as root.
441 Alternatively, non-root users can be enabled to run turbostat this way:
442
443 # setcap cap_sys_admin,cap_sys_rawio,cap_sys_nice=+ep path/to/turbostat
444
445 # chmod +r /dev/cpu/*/msr
446
447 # chmod +r /dev/cpu_dma_latency
448
449 .B "turbostat "
450 reads hardware counters, but doesn't write them.
451 So it will not interfere with the OS or other programs, including
452 multiple invocations of itself.
453
454 \fBturbostat \fP
455 may work poorly on Linux-2.6.20 through 2.6.29,
456 as \fBacpi-cpufreq \fPperiodically cleared the APERF and MPERF MSRs
457 in those kernels.
458
459 AVG_MHz = APERF_delta/measurement_interval. This is the actual
460 number of elapsed cycles divided by the entire sample interval --
461 including idle time. Note that this calculation is resilient
462 to systems lacking a non-stop TSC.
463
464 TSC_MHz = TSC_delta/measurement_interval.
465 On a system with an invariant TSC, this value will be constant
466 and will closely match the base frequency value shown
467 in the brand string in /proc/cpuinfo. On a system where
468 the TSC stops in idle, TSC_MHz will drop
469 below the processor's base frequency.
470
471 Busy% = MPERF_delta/TSC_delta
472
473 Bzy_MHz = TSC_delta*APERF_delta/MPERF_delta/measurement_interval
474
475 Note that these calculations depend on TSC_delta, so they
476 are not reliable during intervals when TSC_MHz is not running at the base frequency.
477
478 Turbostat data collection is not atomic.
479 Extremely short measurement intervals (much less than 1 second),
480 or system activity that prevents turbostat from being able
481 to run on all CPUS to quickly collect data, will result in
482 inconsistent results.
483
484 The APERF, MPERF MSRs are defined to count non-halted cycles.
485 Although it is not guaranteed by the architecture, turbostat assumes
486 that they count at TSC rate, which is true on all processors tested to date.
487
488 .SH REFERENCES
489 Volume 3B: System Programming Guide"
490 https://www.intel.com/products/processor/manuals/
491
492 .SH FILES
493 .ta
494 .nf
495 /dev/cpu/*/msr
496 .fi
497
498 .SH "SEE ALSO"
499 msr(4), vmstat(8)
500 .PP
501 .SH AUTHOR
502 .nf
503 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.