linux/tools/power/pm-graph/sleepgraph.8
<<
>>
Prefs
   1.TH SLEEPGRAPH 8
   2.SH NAME
   3sleepgraph \- Suspend/Resume timing analysis
   4.SH SYNOPSIS
   5.ft B
   6.B sleepgraph
   7.RB [ OPTIONS ]
   8.RB [ COMMAND ]
   9.SH DESCRIPTION
  10\fBsleepgraph \fP is designed to assist kernel and OS developers
  11in optimizing their linux stack's suspend/resume time. Using a kernel
  12image built with a few extra options enabled, the tool will execute a
  13suspend and capture dmesg and ftrace data until resume is complete.
  14This data is transformed into a device timeline and an optional
  15callgraph to give a detailed view of which devices/subsystems are
  16taking the most time in suspend/resume.
  17.PP
  18If no specific command is given, the default behavior is to initiate
  19a suspend/resume.
  20.PP
  21Generates output files in subdirectory: suspend-yymmdd-HHMMSS
  22   html timeline   :     <hostname>_<mode>.html
  23   raw dmesg file  :     <hostname>_<mode>_dmesg.txt
  24   raw ftrace file :     <hostname>_<mode>_ftrace.txt
  25.SH OPTIONS
  26.TP
  27\fB-h\fR
  28Print the help text.
  29.TP
  30\fB-v\fR
  31Print the current tool version.
  32.TP
  33\fB-verbose\fR
  34Print extra information during execution and analysis.
  35.TP
  36\fB-config \fIfile\fR
  37Pull arguments and config options from a file.
  38.TP
  39\fB-m \fImode\fR
  40Mode to initiate for suspend e.g. standby, freeze, mem (default: mem).
  41.TP
  42\fB-o \fIname\fR
  43Overrides the output subdirectory name when running a new test.
  44Use {date}, {time}, {hostname} for current values.
  45.sp
  46e.g. suspend-{hostname}-{date}-{time}
  47.TP
  48\fB-rtcwake \fIt\fR | off
  49Use rtcwake to autoresume after \fIt\fR seconds (default: 15). Set t to "off" to
  50disable rtcwake and require a user keypress to resume.
  51.TP
  52\fB-addlogs\fR
  53Add the dmesg and ftrace logs to the html output. They will be viewable by
  54clicking buttons in the timeline.
  55.TP
  56\fB-noturbostat\fR
  57By default, if turbostat is found and the requested mode is freeze, sleepgraph
  58will execute the suspend via turbostat and collect data in the timeline log.
  59This option disables the use of turbostat.
  60.TP
  61\fB-result \fIfile\fR
  62Export a results table to a text file for parsing.
  63.TP
  64\fB-sync\fR
  65Sync the filesystems before starting the test. This reduces the size of
  66the sys_sync call which happens in the suspend_prepare phase.
  67.TP
  68\fB-rs \fIenable/disable\fR
  69During test, enable/disable runtime suspend for all devices. The test is delayed
  70by 5 seconds to allow runtime suspend changes to occur. The settings are restored
  71after the test is complete.
  72.TP
  73\fB-display \fIon/off/standby/suspend\fR
  74Switch the display to the requested mode for the test using the xset command.
  75This helps maintain the consistency of test data for better comparison.
  76.TP
  77\fB-wifi\fR
  78If a wifi connection is available, check that it reconnects after resume. Include
  79the reconnect time in the total resume time calculation and treat wifi timeouts
  80as resume failures.
  81
  82.SS "advanced"
  83.TP
  84\fB-gzip\fR
  85Gzip the trace and dmesg logs to save space. The tool can also read in gzipped
  86logs for processing.
  87.TP
  88\fB-cmd \fIstr\fR
  89Run the timeline over a custom suspend command, e.g. pm-suspend. By default
  90the tool forces suspend via /sys/power/state so this allows testing over
  91an OS's official suspend method. The output file will change to
  92hostname_command.html and will autodetect which suspend mode was triggered.
  93.TP
  94\fB-filter \fI"d1,d2,..."\fR
  95Filter out all but these device callbacks. These strings can be device names
  96or module names. e.g. 0000:00:02.0, ata5, i915, usb, etc.
  97.TP
  98\fB-mindev \fIt\fR
  99Discard all device callbacks shorter than \fIt\fR milliseconds (default: 0.0).
 100This reduces the html file size as there can be many tiny callbacks which are barely
 101visible. The value is a float: e.g. 0.001 represents 1 us.
 102.TP
 103\fB-proc\fR
 104Add usermode process info into the timeline (default: disabled).
 105.TP
 106\fB-dev\fR
 107Add kernel source calls and threads to the timeline (default: disabled).
 108.TP
 109\fB-x2\fR
 110Run two suspend/resumes back to back (default: disabled).
 111.TP
 112\fB-x2delay \fIt\fR
 113Include \fIt\fR ms delay between multiple test runs (default: 0 ms).
 114.TP
 115\fB-predelay \fIt\fR
 116Include \fIt\fR ms delay before 1st suspend (default: 0 ms).
 117.TP
 118\fB-postdelay \fIt\fR
 119Include \fIt\fR ms delay after last resume (default: 0 ms).
 120.TP
 121\fB-multi \fIn d\fR
 122Used for endurance testing. If \fIn\fR is entirely numeric, it's treated as a count:
 123Execute \fIn\fR consecutive tests at \fId\fR second intervals.
 124If \fIn\fR is an integer followed by a "d", "h", or "m", it's treated as a duration:
 125Execute tests continuously over \fIn\fR days, hours, or minutes at \fId\fR second intervals.
 126The outputs will be created in a new subdirectory, for count: suspend-{date}-{time}-xN,
 127for duration: suspend-{date}-{time}-Nm. When the multitest run is done, the \fI-summary\fR
 128command is called automatically to create summary html files for all the data (unless you
 129use \fI-skiphtml\fR). \fI-skiphtml\fR will speed up the testing by not creating timelines
 130or summary html files. You can then run the tool again at a later time with \fI-summary\fR
 131and \fI-genhtml\fR to create the timelines.
 132.TP
 133\fB-maxfail \fIn\fR
 134Abort a -multi run after \fIn\fR consecutive fails. 0 means never abort (default = 0).
 135.TP
 136\fB-skiphtml\fR
 137Run the test and capture the trace logs, but skip the timeline generation.
 138You can generate the html timelines later with \fI-dmesg\fR & \fI-ftrace\fR, or
 139by running \fI-summary\fR and \fI-genhtml\fR.
 140
 141.SS "ftrace debug"
 142.TP
 143\fB-f\fR
 144Use ftrace to create device callgraphs (default: disabled). This can produce
 145very large outputs, i.e. 10MB - 100MB.
 146.TP
 147\fB-ftop\fR
 148Use ftrace on the top level call: "suspend_devices_and_enter" only (default: disabled).
 149This option implies -f and creates a single callgraph covering all of suspend/resume.
 150.TP
 151\fB-maxdepth \fIlevel\fR
 152limit the callgraph trace depth to \fIlevel\fR (default: 0=all). This is
 153the best way to limit the output size when using callgraphs via -f.
 154.TP
 155\fB-expandcg\fR
 156pre-expand the callgraph data in the html output (default: disabled)
 157.TP
 158\fB-fadd \fIfile\fR
 159Add functions to be graphed in the timeline from a list in a text file
 160.TP
 161\fB-mincg \fIt\fR
 162Discard all callgraphs shorter than \fIt\fR milliseconds (default: 0.0).
 163This reduces the html file size as there can be many tiny callgraphs
 164which are barely visible in the timeline.
 165The value is a float: e.g. 0.001 represents 1 us.
 166.TP
 167\fB-cgfilter \fI"func1,func2,..."\fR
 168Reduce callgraph output in the timeline by limiting it certain devices. The
 169argument can be a single device name or a comma delimited list.
 170(default: none)
 171.TP
 172\fB-cgskip \fIfile\fR
 173Reduce callgraph timeline size by skipping over uninteresting functions
 174in the trace, e.g. printk or console_unlock. The functions listed
 175in this file will show up as empty leaves in the callgraph with only the start/end
 176times displayed. cgskip.txt is used automatically if found in the path, so
 177use "off" to disable completely (default: cgskip.txt)
 178.TP
 179\fB-cgphase \fIp\fR
 180Only show callgraph data for phase \fIp\fR (e.g. suspend_late).
 181.TP
 182\fB-cgtest \fIn\fR
 183In an x2 run, only show callgraph data for test \fIn\fR (e.g. 0 or 1).
 184.TP
 185\fB-timeprec \fIn\fR
 186Number of significant digits in timestamps (0:S, [3:ms], 6:us).
 187.TP
 188\fB-bufsize \fIN\fR
 189Set trace buffer size to N kilo-bytes (default: all of free memory up to 3GB)
 190
 191.SH COMMANDS
 192.TP
 193\fB-summary \fIindir\fR
 194Create a set of summary pages for all tests in \fIindir\fR recursively.
 195Creates summary.html, summary-issues.html, and summary-devices.html in the current folder.
 196summary.html is a table of tests with relevant info sorted by kernel/host/mode,
 197and links to the test html files. It identifies the minimum, maximum, and median
 198suspend and resume times for you with highlights and links in the header.
 199summary-issues.html is a list of kernel issues found in dmesg from all the tests.
 200summary-devices.html is a list of devices and times from all the tests.
 201
 202Use \fI-genhtml\fR to regenerate any tests with missing html.
 203.TP
 204\fB-genhtml\fR
 205Used with \fI-summary\fR to regenerate any missing html timelines from their
 206dmesg and ftrace logs. This will require a significant amount of time if there
 207are thousands of tests.
 208.TP
 209\fB-modes\fR
 210List available suspend modes.
 211.TP
 212\fB-status\fR
 213Test to see if the system is able to run this tool. Use this along
 214with any options you intend to use to see if they will work.
 215.TP
 216\fB-fpdt\fR
 217Print out the contents of the ACPI Firmware Performance Data Table.
 218.TP
 219\fB-wificheck\fR
 220Print out wifi status and connection details.
 221.TP
 222\fB-xon/-xoff/-xstandby/-xsuspend\fR
 223Test xset by attempting to switch the display to the given mode. This
 224is the same command which will be issued by \fB-display \fImode\fR.
 225.TP
 226\fB-xstat\fR
 227Get the current DPMS display mode.
 228.TP
 229\fB-sysinfo\fR
 230Print out system info extracted from BIOS. Reads /dev/mem directly instead of going through dmidecode.
 231.TP
 232\fB-devinfo\fR
 233Print out the pm settings of all devices which support runtime suspend.
 234.TP
 235\fB-cmdinfo\fR
 236Print out all the platform data collected from the system that makes it into the logs.
 237.TP
 238\fB-flist\fR
 239Print the list of ftrace functions currently being captured. Functions
 240that are not available as symbols in the current kernel are shown in red.
 241By default, the tool traces a list of important suspend/resume functions
 242in order to better fill out the timeline. If the user has added their own
 243with -fadd they will also be checked.
 244.TP
 245\fB-flistall\fR
 246Print all ftrace functions capable of being captured. These are all the
 247possible values you can add to trace via the -fadd argument.
 248.SS "rebuild"
 249.TP
 250\fB-ftrace \fIfile\fR
 251Create HTML output from an existing ftrace file.
 252.TP
 253\fB-dmesg \fIfile\fR
 254Create HTML output from an existing dmesg file.
 255
 256.SH EXAMPLES
 257.SS "simple commands"
 258Check which suspend modes are currently supported.
 259.IP
 260\f(CW$ sleepgraph -modes\fR
 261.PP
 262Read the Firmware Performance Data Table (FPDT)
 263.IP
 264\f(CW$ sudo sleepgraph -fpdt\fR
 265.PP
 266Print out the current USB power topology
 267.IP
 268\f(CW$ sleepgraph -usbtopo
 269.PP
 270Verify that you can run a command with a set of arguments
 271.IP
 272\f(CW$ sudo sleepgraph -f -rtcwake 30 -status
 273.PP
 274Generate a summary of all timelines in a particular folder.
 275.IP
 276\f(CW$ sleepgraph -summary ~/workspace/myresults/\fR
 277.PP
 278
 279.SS "capturing basic timelines"
 280Execute a mem suspend with a 15 second wakeup. Include the logs in the html.
 281.IP
 282\f(CW$ sudo sleepgraph -rtcwake 15 -addlogs\fR
 283.PP
 284Execute a standby with a 15 second wakeup. Change the output folder name.
 285.IP
 286\f(CW$ sudo sleepgraph -m standby -rtcwake 15 -o "standby-{host}-{date}-{time}"\fR
 287.PP
 288Execute a freeze with no wakeup (require keypress). Change output folder name.
 289.IP
 290\f(CW$ sudo sleepgraph -m freeze -rtcwake off -o "freeze-{hostname}-{date}-{time}"\fR
 291.PP
 292
 293.SS "capturing advanced timelines"
 294Execute a suspend & include dev mode source calls, limit callbacks to 5ms or larger.
 295.IP
 296\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -dev -mindev 5\fR
 297.PP
 298Run two suspends back to back, include a 500ms delay before, after, and in between runs.
 299.IP
 300\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -x2 -predelay 500 -x2delay 500 -postdelay 500\fR
 301.PP
 302Execute a suspend using a custom command.
 303.IP
 304\f(CW$ sudo sleepgraph -cmd "echo mem > /sys/power/state" -rtcwake 15\fR
 305.PP
 306
 307.SS "endurance testing using -multi"
 308.PP
 309Do a batch run of 10 freezes with 30 seconds delay between runs.
 310.IP
 311\f(CW$ sudo sleepgraph -m freeze -rtcwake 15 -multi 10 30\fR
 312.PP
 313Do a batch run of freezes for 24 hours.
 314.IP
 315\f(CW$ sudo sleepgraph -m freeze -rtcwake 15 -multi 24h 0\fR
 316
 317.SS "adding callgraph data"
 318Add device callgraphs. Limit the trace depth and only show callgraphs 10ms or larger.
 319.IP
 320\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -f -maxdepth 5 -mincg 10\fR
 321.PP
 322Capture a full callgraph across all suspend, then filter the html by a single phase.
 323.IP
 324\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -f\fR
 325.IP
 326\f(CW$ sleepgraph -dmesg host_mem_dmesg.txt -ftrace host_mem_ftrace.txt -f -cgphase resume
 327.PP
 328
 329.SS "rebuild timeline from logs"
 330.PP
 331Rebuild the html from a previous run's logs, using the same options.
 332.IP
 333\f(CW$ sleepgraph -dmesg dmesg.txt -ftrace ftrace.txt -callgraph\fR
 334.PP
 335Rebuild the html with different options.
 336.IP
 337\f(CW$ sleepgraph -dmesg dmesg.txt -ftrace ftrace.txt -addlogs -srgap\fR
 338
 339.SH "SEE ALSO"
 340dmesg(1)
 341.PP
 342.SH AUTHOR
 343.nf
 344Written by Todd Brandt <todd.e.brandt@linux.intel.com>
 345