diff options
Diffstat (limited to 'tools/power')
20 files changed, 8732 insertions, 613 deletions
diff --git a/tools/power/acpi/common/cmfsize.c b/tools/power/acpi/common/cmfsize.c index bc82596d7354..5b38dc2fec4f 100644 --- a/tools/power/acpi/common/cmfsize.c +++ b/tools/power/acpi/common/cmfsize.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/common/getopt.c b/tools/power/acpi/common/getopt.c index 3919970f5aea..6e78413bb2cb 100644 --- a/tools/power/acpi/common/getopt.c +++ b/tools/power/acpi/common/getopt.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c b/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c index 546cf4a503b7..82a2ff896a95 100644 --- a/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c +++ b/tools/power/acpi/os_specific/service_layers/oslinuxtbl.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/os_specific/service_layers/osunixdir.c b/tools/power/acpi/os_specific/service_layers/osunixdir.c index 66c4badf03e5..ea14eaeb268f 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixdir.c +++ b/tools/power/acpi/os_specific/service_layers/osunixdir.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/os_specific/service_layers/osunixmap.c b/tools/power/acpi/os_specific/service_layers/osunixmap.c index cbfbce18783d..cf9b5a54df92 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixmap.c +++ b/tools/power/acpi/os_specific/service_layers/osunixmap.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/os_specific/service_layers/osunixxf.c b/tools/power/acpi/os_specific/service_layers/osunixxf.c index 10648aaf6164..c04e8fea2c60 100644 --- a/tools/power/acpi/os_specific/service_layers/osunixxf.c +++ b/tools/power/acpi/os_specific/service_layers/osunixxf.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -318,6 +318,28 @@ acpi_os_physical_table_override(struct acpi_table_header *existing_table, /****************************************************************************** * + * FUNCTION: acpi_os_enter_sleep + * + * PARAMETERS: sleep_state - Which sleep state to enter + * rega_value - Register A value + * regb_value - Register B value + * + * RETURN: Status + * + * DESCRIPTION: A hook before writing sleep registers to enter the sleep + * state. Return AE_CTRL_TERMINATE to skip further sleep register + * writes. + * + *****************************************************************************/ + +acpi_status acpi_os_enter_sleep(u8 sleep_state, u32 rega_value, u32 regb_value) +{ + + return (AE_OK); +} + +/****************************************************************************** + * * FUNCTION: acpi_os_redirect_output * * PARAMETERS: destination - An open file handle/pointer diff --git a/tools/power/acpi/tools/acpidump/acpidump.h b/tools/power/acpi/tools/acpidump/acpidump.h index 00423fc45e7c..d6aa40fce2b1 100644 --- a/tools/power/acpi/tools/acpidump/acpidump.h +++ b/tools/power/acpi/tools/acpidump/acpidump.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/tools/acpidump/apdump.c b/tools/power/acpi/tools/acpidump/apdump.c index 9031be1afe63..60df1fbd4a77 100644 --- a/tools/power/acpi/tools/acpidump/apdump.c +++ b/tools/power/acpi/tools/acpidump/apdump.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/tools/acpidump/apfiles.c b/tools/power/acpi/tools/acpidump/apfiles.c index dd5b861dc4a8..31b5a7f74015 100644 --- a/tools/power/acpi/tools/acpidump/apfiles.c +++ b/tools/power/acpi/tools/acpidump/apfiles.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/acpi/tools/acpidump/apmain.c b/tools/power/acpi/tools/acpidump/apmain.c index 7ff46be908f0..dd82afa897bd 100644 --- a/tools/power/acpi/tools/acpidump/apmain.c +++ b/tools/power/acpi/tools/acpidump/apmain.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2016, Intel Corp. + * Copyright (C) 2000 - 2017, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/tools/power/cpupower/utils/cpufreq-info.c b/tools/power/cpupower/utils/cpufreq-info.c index 590d12a25f6e..3e701f0e9c14 100644 --- a/tools/power/cpupower/utils/cpufreq-info.c +++ b/tools/power/cpupower/utils/cpufreq-info.c @@ -285,20 +285,24 @@ static int get_freq_hardware(unsigned int cpu, unsigned int human) /* --hwlimits / -l */ -static int get_hardware_limits(unsigned int cpu) +static int get_hardware_limits(unsigned int cpu, unsigned int human) { unsigned long min, max; - printf(_(" hardware limits: ")); if (cpufreq_get_hardware_limits(cpu, &min, &max)) { printf(_("Not Available\n")); return -EINVAL; } - print_speed(min); - printf(" - "); - print_speed(max); - printf("\n"); + if (human) { + printf(_(" hardware limits: ")); + print_speed(min); + printf(" - "); + print_speed(max); + printf("\n"); + } else { + printf("%lu %lu\n", min, max); + } return 0; } @@ -456,7 +460,7 @@ static void debug_output_one(unsigned int cpu) get_related_cpus(cpu); get_affected_cpus(cpu); get_latency(cpu, 1); - get_hardware_limits(cpu); + get_hardware_limits(cpu, 1); freqs = cpufreq_get_available_frequencies(cpu); if (freqs) { @@ -622,7 +626,7 @@ int cmd_freq_info(int argc, char **argv) ret = get_driver(cpu); break; case 'l': - ret = get_hardware_limits(cpu); + ret = get_hardware_limits(cpu, human); break; case 'w': ret = get_freq_hardware(cpu, human); @@ -639,7 +643,6 @@ int cmd_freq_info(int argc, char **argv) } if (ret) return ret; - printf("\n"); } return ret; } diff --git a/tools/power/cpupower/utils/helpers/cpuid.c b/tools/power/cpupower/utils/helpers/cpuid.c index 93b0aa74ca03..39c2c7d067bb 100644 --- a/tools/power/cpupower/utils/helpers/cpuid.c +++ b/tools/power/cpupower/utils/helpers/cpuid.c @@ -156,6 +156,7 @@ out: */ case 0x2C: /* Westmere EP - Gulftown */ cpu_info->caps |= CPUPOWER_CAP_HAS_TURBO_RATIO; + break; case 0x2A: /* SNB */ case 0x2D: /* SNB Xeon */ case 0x3A: /* IVB */ diff --git a/tools/power/pm-graph/Makefile b/tools/power/pm-graph/Makefile new file mode 100644 index 000000000000..4d0ccc89e6c6 --- /dev/null +++ b/tools/power/pm-graph/Makefile @@ -0,0 +1,28 @@ +PREFIX ?= /usr +DESTDIR ?= + +all: + @echo "Nothing to build" + +install : + install -d $(DESTDIR)$(PREFIX)/lib/pm-graph + install analyze_suspend.py $(DESTDIR)$(PREFIX)/lib/pm-graph + install analyze_boot.py $(DESTDIR)$(PREFIX)/lib/pm-graph + + ln -s $(DESTDIR)$(PREFIX)/lib/pm-graph/analyze_boot.py $(DESTDIR)$(PREFIX)/bin/bootgraph + ln -s $(DESTDIR)$(PREFIX)/lib/pm-graph/analyze_suspend.py $(DESTDIR)$(PREFIX)/bin/sleepgraph + + install -d $(DESTDIR)$(PREFIX)/share/man/man8 + install bootgraph.8 $(DESTDIR)$(PREFIX)/share/man/man8 + install sleepgraph.8 $(DESTDIR)$(PREFIX)/share/man/man8 + +uninstall : + rm $(DESTDIR)$(PREFIX)/share/man/man8/bootgraph.8 + rm $(DESTDIR)$(PREFIX)/share/man/man8/sleepgraph.8 + + rm $(DESTDIR)$(PREFIX)/bin/bootgraph + rm $(DESTDIR)$(PREFIX)/bin/sleepgraph + + rm $(DESTDIR)$(PREFIX)/lib/pm-graph/analyze_boot.py + rm $(DESTDIR)$(PREFIX)/lib/pm-graph/analyze_suspend.py + rmdir $(DESTDIR)$(PREFIX)/lib/pm-graph diff --git a/tools/power/pm-graph/analyze_boot.py b/tools/power/pm-graph/analyze_boot.py new file mode 100755 index 000000000000..3e1dcbbf1adc --- /dev/null +++ b/tools/power/pm-graph/analyze_boot.py @@ -0,0 +1,824 @@ +#!/usr/bin/python +# +# Tool for analyzing boot timing +# Copyright (c) 2013, Intel Corporation. +# +# This program is free software; you can redistribute it and/or modify it +# under the terms and conditions of the GNU General Public License, +# version 2, as published by the Free Software Foundation. +# +# This program is distributed in the hope 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. +# +# Authors: +# Todd Brandt <todd.e.brandt@linux.intel.com> +# +# Description: +# This tool is designed to assist kernel and OS developers in optimizing +# their linux stack's boot time. It creates an html representation of +# the kernel boot timeline up to the start of the init process. +# + +# ----------------- LIBRARIES -------------------- + +import sys +import time +import os +import string +import re +import platform +import shutil +from datetime import datetime, timedelta +from subprocess import call, Popen, PIPE +import analyze_suspend as aslib + +# ----------------- CLASSES -------------------- + +# Class: SystemValues +# Description: +# A global, single-instance container used to +# store system values and test parameters +class SystemValues(aslib.SystemValues): + title = 'BootGraph' + version = 2.0 + hostname = 'localhost' + testtime = '' + kernel = '' + dmesgfile = '' + ftracefile = '' + htmlfile = 'bootgraph.html' + outfile = '' + phoronix = False + addlogs = False + useftrace = False + usedevsrc = True + suspendmode = 'boot' + max_graph_depth = 2 + graph_filter = 'do_one_initcall' + reboot = False + manual = False + iscronjob = False + timeformat = '%.6f' + def __init__(self): + if('LOG_FILE' in os.environ and 'TEST_RESULTS_IDENTIFIER' in os.environ): + self.phoronix = True + self.addlogs = True + self.outfile = os.environ['LOG_FILE'] + self.htmlfile = os.environ['LOG_FILE'] + self.hostname = platform.node() + self.testtime = datetime.now().strftime('%Y-%m-%d_%H:%M:%S') + if os.path.exists('/proc/version'): + fp = open('/proc/version', 'r') + val = fp.read().strip() + fp.close() + self.kernel = self.kernelVersion(val) + else: + self.kernel = 'unknown' + def kernelVersion(self, msg): + return msg.split()[2] + def kernelParams(self): + cmdline = 'initcall_debug log_buf_len=32M' + if self.useftrace: + cmdline += ' trace_buf_size=128M trace_clock=global '\ + 'trace_options=nooverwrite,funcgraph-abstime,funcgraph-cpu,'\ + 'funcgraph-duration,funcgraph-proc,funcgraph-tail,'\ + 'nofuncgraph-overhead,context-info,graph-time '\ + 'ftrace=function_graph '\ + 'ftrace_graph_max_depth=%d '\ + 'ftrace_graph_filter=%s' % \ + (self.max_graph_depth, self.graph_filter) + return cmdline + def setGraphFilter(self, val): + fp = open(self.tpath+'available_filter_functions') + master = fp.read().split('\n') + fp.close() + for i in val.split(','): + func = i.strip() + if func not in master: + doError('function "%s" not available for ftrace' % func) + self.graph_filter = val + def cronjobCmdString(self): + cmdline = '%s -cronjob' % os.path.abspath(sys.argv[0]) + args = iter(sys.argv[1:]) + for arg in args: + if arg in ['-h', '-v', '-cronjob', '-reboot']: + continue + elif arg in ['-o', '-dmesg', '-ftrace', '-filter']: + args.next() + continue + cmdline += ' '+arg + if self.graph_filter != 'do_one_initcall': + cmdline += ' -filter "%s"' % self.graph_filter + cmdline += ' -o "%s"' % os.path.abspath(self.htmlfile) + return cmdline + def manualRebootRequired(self): + cmdline = self.kernelParams() + print 'To generate a new timeline manually, follow these steps:\n' + print '1. Add the CMDLINE string to your kernel command line.' + print '2. Reboot the system.' + print '3. After reboot, re-run this tool with the same arguments but no command (w/o -reboot or -manual).\n' + print 'CMDLINE="%s"' % cmdline + sys.exit() + +sysvals = SystemValues() + +# Class: Data +# Description: +# The primary container for test data. +class Data(aslib.Data): + dmesg = {} # root data structure + start = 0.0 # test start + end = 0.0 # test end + dmesgtext = [] # dmesg text file in memory + testnumber = 0 + idstr = '' + html_device_id = 0 + valid = False + initstart = 0.0 + boottime = '' + phases = ['boot'] + do_one_initcall = False + def __init__(self, num): + self.testnumber = num + self.idstr = 'a' + self.dmesgtext = [] + self.dmesg = { + 'boot': {'list': dict(), 'start': -1.0, 'end': -1.0, 'row': 0, 'color': '#dddddd'} + } + def deviceTopology(self): + return '' + def newAction(self, phase, name, start, end, ret, ulen): + # new device callback for a specific phase + self.html_device_id += 1 + devid = '%s%d' % (self.idstr, self.html_device_id) + list = self.dmesg[phase]['list'] + length = -1.0 + if(start >= 0 and end >= 0): + length = end - start + i = 2 + origname = name + while(name in list): + name = '%s[%d]' % (origname, i) + i += 1 + list[name] = {'name': name, 'start': start, 'end': end, + 'pid': 0, 'length': length, 'row': 0, 'id': devid, + 'ret': ret, 'ulen': ulen } + return name + def deviceMatch(self, cg): + if cg.end - cg.start == 0: + return True + list = self.dmesg['boot']['list'] + for devname in list: + dev = list[devname] + if cg.name == 'do_one_initcall': + if(cg.start <= dev['start'] and cg.end >= dev['end'] and dev['length'] > 0): + dev['ftrace'] = cg + self.do_one_initcall = True + return True + else: + if(cg.start > dev['start'] and cg.end < dev['end']): + if 'ftraces' not in dev: + dev['ftraces'] = [] + dev['ftraces'].append(cg) + return True + return False + +# ----------------- FUNCTIONS -------------------- + +# Function: loadKernelLog +# Description: +# Load a raw kernel log from dmesg +def loadKernelLog(): + data = Data(0) + data.dmesg['boot']['start'] = data.start = ktime = 0.0 + sysvals.stamp = { + 'time': datetime.now().strftime('%B %d %Y, %I:%M:%S %p'), + 'host': sysvals.hostname, + 'mode': 'boot', 'kernel': ''} + + devtemp = dict() + if(sysvals.dmesgfile): + lf = open(sysvals.dmesgfile, 'r') + else: + lf = Popen('dmesg', stdout=PIPE).stdout + for line in lf: + line = line.replace('\r\n', '') + idx = line.find('[') + if idx > 1: + line = line[idx:] + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) + if(not m): + continue + ktime = float(m.group('ktime')) + if(ktime > 120): + break + msg = m.group('msg') + data.end = data.initstart = ktime + data.dmesgtext.append(line) + if(ktime == 0.0 and re.match('^Linux version .*', msg)): + if(not sysvals.stamp['kernel']): + sysvals.stamp['kernel'] = sysvals.kernelVersion(msg) + continue + m = re.match('.* setting system clock to (?P<t>.*) UTC.*', msg) + if(m): + bt = datetime.strptime(m.group('t'), '%Y-%m-%d %H:%M:%S') + bt = bt - timedelta(seconds=int(ktime)) + data.boottime = bt.strftime('%Y-%m-%d_%H:%M:%S') + sysvals.stamp['time'] = bt.strftime('%B %d %Y, %I:%M:%S %p') + continue + m = re.match('^calling *(?P<f>.*)\+.*', msg) + if(m): + devtemp[m.group('f')] = ktime + continue + m = re.match('^initcall *(?P<f>.*)\+.* returned (?P<r>.*) after (?P<t>.*) usecs', msg) + if(m): + data.valid = True + f, r, t = m.group('f', 'r', 't') + if(f in devtemp): + data.newAction('boot', f, devtemp[f], ktime, int(r), int(t)) + data.end = ktime + del devtemp[f] + continue + if(re.match('^Freeing unused kernel memory.*', msg)): + break + + data.dmesg['boot']['end'] = data.end + lf.close() + return data + +# Function: loadTraceLog +# Description: +# Check if trace is available and copy to a temp file +def loadTraceLog(data): + # load the data to a temp file if none given + if not sysvals.ftracefile: + lib = aslib.sysvals + aslib.rootCheck(True) + if not lib.verifyFtrace(): + doError('ftrace not available') + if lib.fgetVal('current_tracer').strip() != 'function_graph': + doError('ftrace not configured for a boot callgraph') + sysvals.ftracefile = '/tmp/boot_ftrace.%s.txt' % os.getpid() + call('cat '+lib.tpath+'trace > '+sysvals.ftracefile, shell=True) + if not sysvals.ftracefile: + doError('No trace data available') + + # parse the trace log + ftemp = dict() + tp = aslib.TestProps() + tp.setTracerType('function_graph') + tf = open(sysvals.ftracefile, 'r') + for line in tf: + if line[0] == '#': + continue + m = re.match(tp.ftrace_line_fmt, line.strip()) + if(not m): + continue + m_time, m_proc, m_pid, m_msg, m_dur = \ + m.group('time', 'proc', 'pid', 'msg', 'dur') + if float(m_time) > data.end: + break + if(m_time and m_pid and m_msg): + t = aslib.FTraceLine(m_time, m_msg, m_dur) + pid = int(m_pid) + else: + continue + if t.fevent or t.fkprobe: + continue + key = (m_proc, pid) + if(key not in ftemp): + ftemp[key] = [] + ftemp[key].append(aslib.FTraceCallGraph(pid)) + cg = ftemp[key][-1] + if(cg.addLine(t)): + ftemp[key].append(aslib.FTraceCallGraph(pid)) + tf.close() + + # add the callgraph data to the device hierarchy + for key in ftemp: + proc, pid = key + for cg in ftemp[key]: + if len(cg.list) < 1 or cg.invalid: + continue + if(not cg.postProcess()): + print('Sanity check failed for %s-%d' % (proc, pid)) + continue + # match cg data to devices + if not data.deviceMatch(cg): + print ' BAD: %s %s-%d [%f - %f]' % (cg.name, proc, pid, cg.start, cg.end) + +# Function: colorForName +# Description: +# Generate a repeatable color from a list for a given name +def colorForName(name): + list = [ + ('c1', '#ec9999'), + ('c2', '#ffc1a6'), + ('c3', '#fff0a6'), + ('c4', '#adf199'), + ('c5', '#9fadea'), + ('c6', '#a699c1'), + ('c7', '#ad99b4'), + ('c8', '#eaffea'), + ('c9', '#dcecfb'), + ('c10', '#ffffea') + ] + i = 0 + total = 0 + count = len(list) + while i < len(name): + total += ord(name[i]) + i += 1 + return list[total % count] + +def cgOverview(cg, minlen): + stats = dict() + large = [] + for l in cg.list: + if l.fcall and l.depth == 1: + if l.length >= minlen: + large.append(l) + if l.name not in stats: + stats[l.name] = [0, 0.0] + stats[l.name][0] += (l.length * 1000.0) + stats[l.name][1] += 1 + return (large, stats) + +# Function: createBootGraph +# Description: +# Create the output html file from the resident test data +# Arguments: +# testruns: array of Data objects from parseKernelLog or parseTraceLog +# Output: +# True if the html file was created, false if it failed +def createBootGraph(data, embedded): + # html function templates + html_srccall = '<div id={6} title="{5}" class="srccall" style="left:{1}%;top:{2}px;height:{3}px;width:{4}%;line-height:{3}px;">{0}</div>\n' + html_timetotal = '<table class="time1">\n<tr>'\ + '<td class="blue">Time from Kernel Boot to start of User Mode: <b>{0} ms</b></td>'\ + '</tr>\n</table>\n' + + # device timeline + devtl = aslib.Timeline(100, 20) + + # write the test title and general info header + devtl.createHeader(sysvals, 'noftrace') + + # Generate the header for this timeline + t0 = data.start + tMax = data.end + tTotal = tMax - t0 + if(tTotal == 0): + print('ERROR: No timeline data') + return False + boot_time = '%.0f'%(tTotal*1000) + devtl.html += html_timetotal.format(boot_time) + + # determine the maximum number of rows we need to draw + phase = 'boot' + list = data.dmesg[phase]['list'] + devlist = [] + for devname in list: + d = aslib.DevItem(0, phase, list[devname]) + devlist.append(d) + devtl.getPhaseRows(devlist) + devtl.calcTotalRows() + + # draw the timeline background + devtl.createZoomBox() + boot = data.dmesg[phase] + length = boot['end']-boot['start'] + left = '%.3f' % (((boot['start']-t0)*100.0)/tTotal) + width = '%.3f' % ((length*100.0)/tTotal) + devtl.html += devtl.html_tblock.format(phase, left, width, devtl.scaleH) + devtl.html += devtl.html_phase.format('0', '100', \ + '%.3f'%devtl.scaleH, '%.3f'%devtl.bodyH, \ + 'white', '') + + # draw the device timeline + num = 0 + devstats = dict() + for devname in sorted(list): + cls, color = colorForName(devname) + dev = list[devname] + info = '@|%.3f|%.3f|%.3f|%d' % (dev['start']*1000.0, dev['end']*1000.0, + dev['ulen']/1000.0, dev['ret']) + devstats[dev['id']] = {'info':info} + dev['color'] = color + height = devtl.phaseRowHeight(0, phase, dev['row']) + top = '%.6f' % ((dev['row']*height) + devtl.scaleH) + left = '%.6f' % (((dev['start']-t0)*100)/tTotal) + width = '%.6f' % (((dev['end']-dev['start'])*100)/tTotal) + length = ' (%0.3f ms) ' % ((dev['end']-dev['start'])*1000) + devtl.html += devtl.html_device.format(dev['id'], + devname+length+'kernel_mode', left, top, '%.3f'%height, + width, devname, ' '+cls, '') + rowtop = devtl.phaseRowTop(0, phase, dev['row']) + height = '%.6f' % (devtl.rowH / 2) + top = '%.6f' % (rowtop + devtl.scaleH + (devtl.rowH / 2)) + if data.do_one_initcall: + if('ftrace' not in dev): + continue + cg = dev['ftrace'] + large, stats = cgOverview(cg, 0.001) + devstats[dev['id']]['fstat'] = stats + for l in large: + left = '%f' % (((l.time-t0)*100)/tTotal) + width = '%f' % (l.length*100/tTotal) + title = '%s (%0.3fms)' % (l.name, l.length * 1000.0) + devtl.html += html_srccall.format(l.name, left, + top, height, width, title, 'x%d'%num) + num += 1 + continue + if('ftraces' not in dev): + continue + for cg in dev['ftraces']: + left = '%f' % (((cg.start-t0)*100)/tTotal) + width = '%f' % ((cg.end-cg.start)*100/tTotal) + cglen = (cg.end - cg.start) * 1000.0 + title = '%s (%0.3fms)' % (cg.name, cglen) + cg.id = 'x%d' % num + devtl.html += html_srccall.format(cg.name, left, + top, height, width, title, dev['id']+cg.id) + num += 1 + + # draw the time scale, try to make the number of labels readable + devtl.createTimeScale(t0, tMax, tTotal, phase) + devtl.html += '</div>\n' + + # timeline is finished + devtl.html += '</div>\n</div>\n' + + if(sysvals.outfile == sysvals.htmlfile): + hf = open(sysvals.htmlfile, 'a') + else: + hf = open(sysvals.htmlfile, 'w') + + # add the css if this is not an embedded run + extra = '\ + .c1 {background:rgba(209,0,0,0.4);}\n\ + .c2 {background:rgba(255,102,34,0.4);}\n\ + .c3 {background:rgba(255,218,33,0.4);}\n\ + .c4 {background:rgba(51,221,0,0.4);}\n\ + .c5 {background:rgba(17,51,204,0.4);}\n\ + .c6 {background:rgba(34,0,102,0.4);}\n\ + .c7 {background:rgba(51,0,68,0.4);}\n\ + .c8 {background:rgba(204,255,204,0.4);}\n\ + .c9 {background:rgba(169,208,245,0.4);}\n\ + .c10 {background:rgba(255,255,204,0.4);}\n\ + .vt {transform:rotate(-60deg);transform-origin:0 0;}\n\ + table.fstat {table-layout:fixed;padding:150px 15px 0 0;font-size:10px;column-width:30px;}\n\ + .fstat th {width:55px;}\n\ + .fstat td {text-align:left;width:35px;}\n\ + .srccall {position:absolute;font-size:10px;z-index:7;overflow:hidden;color:black;text-align:center;white-space:nowrap;border-radius:5px;border:1px solid black;background:linear-gradient(to bottom right,#CCC,#969696);}\n\ + .srccall:hover {color:white;font-weight:bold;border:1px solid white;}\n' + if(not embedded): + aslib.addCSS(hf, sysvals, 1, False, extra) + + # write the device timeline + hf.write(devtl.html) + + # add boot specific html + statinfo = 'var devstats = {\n' + for n in sorted(devstats): + statinfo += '\t"%s": [\n\t\t"%s",\n' % (n, devstats[n]['info']) + if 'fstat' in devstats[n]: + funcs = devstats[n]['fstat'] + for f in sorted(funcs, key=funcs.get, reverse=True): + if funcs[f][0] < 0.01 and len(funcs) > 10: + break + statinfo += '\t\t"%f|%s|%d",\n' % (funcs[f][0], f, funcs[f][1]) + statinfo += '\t],\n' + statinfo += '};\n' + html = \ + '<div id="devicedetailtitle"></div>\n'\ + '<div id="devicedetail" style="display:none;">\n'\ + '<div id="devicedetail0">\n'\ + '<div id="kernel_mode" class="phaselet" style="left:0%;width:100%;background:#DDDDDD"></div>\n'\ + '</div>\n</div>\n'\ + '<script type="text/javascript">\n'+statinfo+\ + '</script>\n' + hf.write(html) + + # add the callgraph html + if(sysvals.usecallgraph): + aslib.addCallgraphs(sysvals, hf, data) + + # add the dmesg log as a hidden div + if sysvals.addlogs: + hf.write('<div id="dmesglog" style="display:none;">\n') + for line in data.dmesgtext: + line = line.replace('<', '<').replace('>', '>') + hf.write(line) + hf.write('</div>\n') + + if(not embedded): + # write the footer and close + aslib.addScriptCode(hf, [data]) + hf.write('</body>\n</html>\n') + else: + # embedded out will be loaded in a page, skip the js + hf.write('<div id=bounds style=display:none>%f,%f</div>' % \ + (data.start*1000, data.initstart*1000)) + hf.close() + return True + +# Function: updateCron +# Description: +# (restore=False) Set the tool to run automatically on reboot +# (restore=True) Restore the original crontab +def updateCron(restore=False): + if not restore: + sysvals.rootUser(True) + crondir = '/var/spool/cron/crontabs/' + cronfile = crondir+'root' + backfile = crondir+'root-analyze_boot-backup' + if not os.path.exists(crondir): + doError('%s not found' % crondir) + out = Popen(['which', 'crontab'], stdout=PIPE).stdout.read() + if not out: + doError('crontab not found') + # on restore: move the backup cron back into place + if restore: + if os.path.exists(backfile): + shutil.move(backfile, cronfile) + return + # backup current cron and install new one with reboot + if os.path.exists(cronfile): + shutil.move(cronfile, backfile) + else: + fp = open(backfile, 'w') + fp.close() + res = -1 + try: + fp = open(backfile, 'r') + op = open(cronfile, 'w') + for line in fp: + if '@reboot' not in line: + op.write(line) + continue + fp.close() + op.write('@reboot python %s\n' % sysvals.cronjobCmdString()) + op.close() + res = call('crontab %s' % cronfile, shell=True) + except Exception, e: + print 'Exception: %s' % str(e) + shutil.move(backfile, cronfile) + res = -1 + if res != 0: + doError('crontab failed') + +# Function: updateGrub +# Description: +# update grub.cfg for all kernels with our parameters +def updateGrub(restore=False): + # call update-grub on restore + if restore: + try: + call(['update-grub'], stderr=PIPE, stdout=PIPE, + env={'PATH': '.:/sbin:/usr/sbin:/usr/bin:/sbin:/bin'}) + except Exception, e: + print 'Exception: %s\n' % str(e) + return + # verify we can do this + sysvals.rootUser(True) + grubfile = '/etc/default/grub' + if not os.path.exists(grubfile): + print 'ERROR: Unable to set the kernel parameters via grub.\n' + sysvals.manualRebootRequired() + out = Popen(['which', 'update-grub'], stdout=PIPE).stdout.read() + if not out: + print 'ERROR: Unable to set the kernel parameters via grub.\n' + sysvals.manualRebootRequired() + + # extract the option and create a grub config without it + tgtopt = 'GRUB_CMDLINE_LINUX_DEFAULT' + cmdline = '' + tempfile = '/etc/default/grub.analyze_boot' + shutil.move(grubfile, tempfile) + res = -1 + try: + fp = open(tempfile, 'r') + op = open(grubfile, 'w') + cont = False + for line in fp: + line = line.strip() + if len(line) == 0 or line[0] == '#': + continue + opt = line.split('=')[0].strip() + if opt == tgtopt: + cmdline = line.split('=', 1)[1].strip('\\') + if line[-1] == '\\': + cont = True + elif cont: + cmdline += line.strip('\\') + if line[-1] != '\\': + cont = False + else: + op.write('%s\n' % line) + fp.close() + # if the target option value is in quotes, strip them + sp = '"' + val = cmdline.strip() + if val[0] == '\'' or val[0] == '"': + sp = val[0] + val = val.strip(sp) + cmdline = val + # append our cmd line options + if len(cmdline) > 0: + cmdline += ' ' + cmdline += sysvals.kernelParams() + # write out the updated target option + op.write('\n%s=%s%s%s\n' % (tgtopt, sp, cmdline, sp)) + op.close() + res = call('update-grub') + os.remove(grubfile) + except Exception, e: + print 'Exception: %s' % str(e) + res = -1 + # cleanup + shutil.move(tempfile, grubfile) + if res != 0: + doError('update-grub failed') + +# Function: doError +# Description: +# generic error function for catastrphic failures +# Arguments: +# msg: the error message to print +# help: True if printHelp should be called after, False otherwise +def doError(msg, help=False): + if help == True: + printHelp() + print 'ERROR: %s\n' % msg + sys.exit() + +# Function: printHelp +# Description: +# print out the help text +def printHelp(): + print('') + print('%s v%.1f' % (sysvals.title, sysvals.version)) + print('Usage: bootgraph <options> <command>') + print('') + print('Description:') + print(' This tool reads in a dmesg log of linux kernel boot and') + print(' creates an html representation of the boot timeline up to') + print(' the start of the init process.') + print('') + print(' If no specific command is given the tool reads the current dmesg') + print(' and/or ftrace log and outputs bootgraph.html') + print('') + print('Options:') + print(' -h Print this help text') + print(' -v Print the current tool version') + print(' -addlogs Add the dmesg log to the html output') + print(' -o file Html timeline name (default: bootgraph.html)') + print(' [advanced]') + print(' -f Use ftrace to add function detail (default: disabled)') + print(' -callgraph Add callgraph detail, can be very large (default: disabled)') + print(' -maxdepth N limit the callgraph data to N call levels (default: 2)') + print(' -mincg ms Discard all callgraphs shorter than ms milliseconds (e.g. 0.001 for us)') + print(' -timeprec N Number of significant digits in timestamps (0:S, 3:ms, [6:us])') + print(' -expandcg pre-expand the callgraph data in the html output (default: disabled)') + print(' -filter list Limit ftrace to comma-delimited list of functions (default: do_one_initcall)') + print(' [commands]') + print(' -reboot Reboot the machine automatically and generate a new timeline') + print(' -manual Show the requirements to generate a new timeline manually') + print(' -dmesg file Load a stored dmesg file (used with -ftrace)') + print(' -ftrace file Load a stored ftrace file (used with -dmesg)') + print(' -flistall Print all functions capable of being captured in ftrace') + print('') + return True + +# ----------------- MAIN -------------------- +# exec start (skipped if script is loaded as library) +if __name__ == '__main__': + # loop through the command line arguments + cmd = '' + simplecmds = ['-updategrub', '-flistall'] + args = iter(sys.argv[1:]) + for arg in args: + if(arg == '-h'): + printHelp() + sys.exit() + elif(arg == '-v'): + print("Version %.1f" % sysvals.version) + sys.exit() + elif(arg in simplecmds): + cmd = arg[1:] + elif(arg == '-f'): + sysvals.useftrace = True + elif(arg == '-callgraph'): + sysvals.useftrace = True + sysvals.usecallgraph = True + elif(arg == '-mincg'): + sysvals.mincglen = aslib.getArgFloat('-mincg', args, 0.0, 10000.0) + elif(arg == '-timeprec'): + sysvals.setPrecision(aslib.getArgInt('-timeprec', args, 0, 6)) + elif(arg == '-maxdepth'): + sysvals.max_graph_depth = aslib.getArgInt('-maxdepth', args, 0, 1000) + elif(arg == '-filter'): + try: + val = args.next() + except: + doError('No filter functions supplied', True) + aslib.rootCheck(True) + sysvals.setGraphFilter(val) + elif(arg == '-ftrace'): + try: + val = args.next() + except: + doError('No ftrace file supplied', True) + if(os.path.exists(val) == False): + doError('%s does not exist' % val) + sysvals.ftracefile = val + elif(arg == '-addlogs'): + sysvals.addlogs = True + elif(arg == '-expandcg'): + sysvals.cgexp = True + elif(arg == '-dmesg'): + try: + val = args.next() + except: + doError('No dmesg file supplied', True) + if(os.path.exists(val) == False): + doError('%s does not exist' % val) + if(sysvals.htmlfile == val or sysvals.outfile == val): + doError('Output filename collision') + sysvals.dmesgfile = val + elif(arg == '-o'): + try: + val = args.next() + except: + doError('No HTML filename supplied', True) + if(sysvals.dmesgfile == val or sysvals.ftracefile == val): + doError('Output filename collision') + sysvals.htmlfile = val + elif(arg == '-reboot'): + if sysvals.iscronjob: + doError('-reboot and -cronjob are incompatible') + sysvals.reboot = True + elif(arg == '-manual'): + sysvals.reboot = True + sysvals.manual = True + # remaining options are only for cron job use + elif(arg == '-cronjob'): + sysvals.iscronjob = True + if sysvals.reboot: + doError('-reboot and -cronjob are incompatible') + else: + doError('Invalid argument: '+arg, True) + + if cmd != '': + if cmd == 'updategrub': + updateGrub() + elif cmd == 'flistall': + sysvals.getFtraceFilterFunctions(False) + sys.exit() + + # update grub, setup a cronjob, and reboot + if sysvals.reboot: + if not sysvals.manual: + updateGrub() + updateCron() + call('reboot') + else: + sysvals.manualRebootRequired() + sys.exit() + + # disable the cronjob + if sysvals.iscronjob: + updateCron(True) + updateGrub(True) + + data = loadKernelLog() + if sysvals.useftrace: + loadTraceLog(data) + if sysvals.iscronjob: + try: + sysvals.fsetVal('0', 'tracing_on') + except: + pass + + if(sysvals.outfile and sysvals.phoronix): + fp = open(sysvals.outfile, 'w') + fp.write('pass %s initstart %.3f end %.3f boot %s\n' % + (data.valid, data.initstart*1000, data.end*1000, data.boottime)) + fp.close() + if(not data.valid): + if sysvals.dmesgfile: + doError('No initcall data found in %s' % sysvals.dmesgfile) + else: + doError('No initcall data found, is initcall_debug enabled?') + + print(' Host: %s' % sysvals.hostname) + print(' Test time: %s' % sysvals.testtime) + print(' Boot time: %s' % data.boottime) + print('Kernel Version: %s' % sysvals.kernel) + print(' Kernel start: %.3f' % (data.start * 1000)) + print(' init start: %.3f' % (data.initstart * 1000)) + + createBootGraph(data, sysvals.phoronix) diff --git a/tools/power/pm-graph/analyze_suspend.py b/tools/power/pm-graph/analyze_suspend.py new file mode 100755 index 000000000000..a9206e67fc1f --- /dev/null +++ b/tools/power/pm-graph/analyze_suspend.py @@ -0,0 +1,5309 @@ +#!/usr/bin/python +# +# Tool for analyzing suspend/resume timing +# Copyright (c) 2013, Intel Corporation. +# +# This program is free software; you can redistribute it and/or modify it +# under the terms and conditions of the GNU General Public License, +# version 2, as published by the Free Software Foundation. +# +# This program is distributed in the hope 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. +# +# Authors: +# Todd Brandt <todd.e.brandt@linux.intel.com> +# +# Links: +# Home Page +# https://01.org/suspendresume +# Source repo +# https://github.com/01org/pm-graph +# +# Description: +# This tool is designed to assist kernel and OS developers in optimizing +# their linux stack's suspend/resume time. Using a kernel image built +# with a few extra options enabled, the tool will execute a suspend and +# will capture dmesg and ftrace data until resume is complete. This data +# is transformed into a device timeline and a callgraph to give a quick +# and detailed view of which devices and callbacks are taking the most +# time in suspend/resume. The output is a single html file which can be +# viewed in firefox or chrome. +# +# The following kernel build options are required: +# CONFIG_PM_DEBUG=y +# CONFIG_PM_SLEEP_DEBUG=y +# CONFIG_FTRACE=y +# CONFIG_FUNCTION_TRACER=y +# CONFIG_FUNCTION_GRAPH_TRACER=y +# CONFIG_KPROBES=y +# CONFIG_KPROBES_ON_FTRACE=y +# +# For kernel versions older than 3.15: +# The following additional kernel parameters are required: +# (e.g. in file /etc/default/grub) +# GRUB_CMDLINE_LINUX_DEFAULT="... initcall_debug log_buf_len=16M ..." +# + +# ----------------- LIBRARIES -------------------- + +import sys +import time +import os +import string +import re +import platform +from datetime import datetime +import struct +import ConfigParser +from threading import Thread +from subprocess import call, Popen, PIPE + +# ----------------- CLASSES -------------------- + +# Class: SystemValues +# Description: +# A global, single-instance container used to +# store system values and test parameters +class SystemValues: + title = 'SleepGraph' + version = '4.6' + ansi = False + verbose = False + addlogs = False + mindevlen = 0.0 + mincglen = 0.0 + cgphase = '' + cgtest = -1 + max_graph_depth = 0 + callloopmaxgap = 0.0001 + callloopmaxlen = 0.005 + srgap = 0 + cgexp = False + outdir = '' + testdir = '.' + tpath = '/sys/kernel/debug/tracing/' + fpdtpath = '/sys/firmware/acpi/tables/FPDT' + epath = '/sys/kernel/debug/tracing/events/power/' + traceevents = [ + 'suspend_resume', + 'device_pm_callback_end', + 'device_pm_callback_start' + ] + logmsg = '' + testcommand = '' + mempath = '/dev/mem' + powerfile = '/sys/power/state' + suspendmode = 'mem' + hostname = 'localhost' + prefix = 'test' + teststamp = '' + dmesgstart = 0.0 + dmesgfile = '' + ftracefile = '' + htmlfile = '' + embedded = False + rtcwake = True + rtcwaketime = 15 + rtcpath = '' + devicefilter = [] + stamp = 0 + execcount = 1 + x2delay = 0 + usecallgraph = False + usetraceevents = False + usetraceeventsonly = False + usetracemarkers = True + usekprobes = True + usedevsrc = False + useprocmon = False + notestrun = False + mixedphaseheight = True + devprops = dict() + predelay = 0 + postdelay = 0 + procexecfmt = 'ps - (?P<ps>.*)$' + devpropfmt = '# Device Properties: .*' + tracertypefmt = '# tracer: (?P<t>.*)' + firmwarefmt = '# fwsuspend (?P<s>[0-9]*) fwresume (?P<r>[0-9]*)$' + stampfmt = '# suspend-(?P<m>[0-9]{2})(?P<d>[0-9]{2})(?P<y>[0-9]{2})-'+\ + '(?P<H>[0-9]{2})(?P<M>[0-9]{2})(?P<S>[0-9]{2})'+\ + ' (?P<host>.*) (?P<mode>.*) (?P<kernel>.*)$' + tracefuncs = { + 'sys_sync': dict(), + 'pm_prepare_console': dict(), + 'pm_notifier_call_chain': dict(), + 'freeze_processes': dict(), + 'freeze_kernel_threads': dict(), + 'pm_restrict_gfp_mask': dict(), + 'acpi_suspend_begin': dict(), + 'suspend_console': dict(), + 'acpi_pm_prepare': dict(), + 'syscore_suspend': dict(), + 'arch_enable_nonboot_cpus_end': dict(), + 'syscore_resume': dict(), + 'acpi_pm_finish': dict(), + 'resume_console': dict(), + 'acpi_pm_end': dict(), + 'pm_restore_gfp_mask': dict(), + 'thaw_processes': dict(), + 'pm_restore_console': dict(), + 'CPU_OFF': { + 'func':'_cpu_down', + 'args_x86_64': {'cpu':'%di:s32'}, + 'format': 'CPU_OFF[{cpu}]' + }, + 'CPU_ON': { + 'func':'_cpu_up', + 'args_x86_64': {'cpu':'%di:s32'}, + 'format': 'CPU_ON[{cpu}]' + }, + } + dev_tracefuncs = { + # general wait/delay/sleep + 'msleep': { 'args_x86_64': {'time':'%di:s32'}, 'ub': 1 }, + 'schedule_timeout_uninterruptible': { 'args_x86_64': {'timeout':'%di:s32'}, 'ub': 1 }, + 'schedule_timeout': { 'args_x86_64': {'timeout':'%di:s32'}, 'ub': 1 }, + 'udelay': { 'func':'__const_udelay', 'args_x86_64': {'loops':'%di:s32'}, 'ub': 1 }, + 'usleep_range': { 'args_x86_64': {'min':'%di:s32', 'max':'%si:s32'}, 'ub': 1 }, + 'mutex_lock_slowpath': { 'func':'__mutex_lock_slowpath', 'ub': 1 }, + 'acpi_os_stall': {'ub': 1}, + # ACPI + 'acpi_resume_power_resources': dict(), + 'acpi_ps_parse_aml': dict(), + # filesystem + 'ext4_sync_fs': dict(), + # 80211 + 'iwlagn_mac_start': dict(), + 'iwlagn_alloc_bcast_station': dict(), + 'iwl_trans_pcie_start_hw': dict(), + 'iwl_trans_pcie_start_fw': dict(), + 'iwl_run_init_ucode': dict(), + 'iwl_load_ucode_wait_alive': dict(), + 'iwl_alive_start': dict(), + 'iwlagn_mac_stop': dict(), + 'iwlagn_mac_suspend': dict(), + 'iwlagn_mac_resume': dict(), + 'iwlagn_mac_add_interface': dict(), + 'iwlagn_mac_remove_interface': dict(), + 'iwlagn_mac_change_interface': dict(), + 'iwlagn_mac_config': dict(), + 'iwlagn_configure_filter': dict(), + 'iwlagn_mac_hw_scan': dict(), + 'iwlagn_bss_info_changed': dict(), + 'iwlagn_mac_channel_switch': dict(), + 'iwlagn_mac_flush': dict(), + # ATA + 'ata_eh_recover': { 'args_x86_64': {'port':'+36(%di):s32'} }, + # i915 + 'i915_gem_resume': dict(), + 'i915_restore_state': dict(), + 'intel_opregion_setup': dict(), + 'g4x_pre_enable_dp': dict(), + 'vlv_pre_enable_dp': dict(), + 'chv_pre_enable_dp': dict(), + 'g4x_enable_dp': dict(), + 'vlv_enable_dp': dict(), + 'intel_hpd_init': dict(), + 'intel_opregion_register': dict(), + 'intel_dp_detect': dict(), + 'intel_hdmi_detect': dict(), + 'intel_opregion_init': dict(), + 'intel_fbdev_set_suspend': dict(), + } + kprobes = dict() + timeformat = '%.3f' + def __init__(self): + # if this is a phoronix test run, set some default options + if('LOG_FILE' in os.environ and 'TEST_RESULTS_IDENTIFIER' in os.environ): + self.embedded = True + self.addlogs = True + self.htmlfile = os.environ['LOG_FILE'] + self.archargs = 'args_'+platform.machine() + self.hostname = platform.node() + if(self.hostname == ''): + self.hostname = 'localhost' + rtc = "rtc0" + if os.path.exists('/dev/rtc'): + rtc = os.readlink('/dev/rtc') + rtc = '/sys/class/rtc/'+rtc + if os.path.exists(rtc) and os.path.exists(rtc+'/date') and \ + os.path.exists(rtc+'/time') and os.path.exists(rtc+'/wakealarm'): + self.rtcpath = rtc + if (hasattr(sys.stdout, 'isatty') and sys.stdout.isatty()): + self.ansi = True + def rootUser(self, fatal=False): + if 'USER' in os.environ and os.environ['USER'] == 'root': + return True + if fatal: + doError('This command must be run as root') + return False + def setPrecision(self, num): + if num < 0 or num > 6: + return + self.timeformat = '%.{0}f'.format(num) + def setOutputFolder(self, value): + args = dict() + n = datetime.now() + args['date'] = n.strftime('%y%m%d') + args['time'] = n.strftime('%H%M%S') + args['hostname'] = self.hostname + self.outdir = value.format(**args) + def setOutputFile(self): + if((self.htmlfile == '') and (self.dmesgfile != '')): + m = re.match('(?P<name>.*)_dmesg\.txt$', self.dmesgfile) + if(m): + self.htmlfile = m.group('name')+'.html' + if((self.htmlfile == '') and (self.ftracefile != '')): + m = re.match('(?P<name>.*)_ftrace\.txt$', self.ftracefile) + if(m): + self.htmlfile = m.group('name')+'.html' + if(self.htmlfile == ''): + self.htmlfile = 'output.html' + def initTestOutput(self, subdir, testpath=''): + self.prefix = self.hostname + v = open('/proc/version', 'r').read().strip() + kver = string.split(v)[2] + n = datetime.now() + testtime = n.strftime('suspend-%m%d%y-%H%M%S') + if not testpath: + testpath = n.strftime('suspend-%y%m%d-%H%M%S') + if(subdir != "."): + self.testdir = subdir+"/"+testpath + else: + self.testdir = testpath + self.teststamp = \ + '# '+testtime+' '+self.prefix+' '+self.suspendmode+' '+kver + if(self.embedded): + self.dmesgfile = \ + '/tmp/'+testtime+'_'+self.suspendmode+'_dmesg.txt' + self.ftracefile = \ + '/tmp/'+testtime+'_'+self.suspendmode+'_ftrace.txt' + return + self.dmesgfile = \ + self.testdir+'/'+self.prefix+'_'+self.suspendmode+'_dmesg.txt' + self.ftracefile = \ + self.testdir+'/'+self.prefix+'_'+self.suspendmode+'_ftrace.txt' + self.htmlfile = \ + self.testdir+'/'+self.prefix+'_'+self.suspendmode+'.html' + if not os.path.isdir(self.testdir): + os.mkdir(self.testdir) + def setDeviceFilter(self, value): + self.devicefilter = [] + if value: + value = value.split(',') + for i in value: + self.devicefilter.append(i.strip()) + def rtcWakeAlarmOn(self): + call('echo 0 > '+self.rtcpath+'/wakealarm', shell=True) + outD = open(self.rtcpath+'/date', 'r').read().strip() + outT = open(self.rtcpath+'/time', 'r').read().strip() + mD = re.match('^(?P<y>[0-9]*)-(?P<m>[0-9]*)-(?P<d>[0-9]*)', outD) + mT = re.match('^(?P<h>[0-9]*):(?P<m>[0-9]*):(?P<s>[0-9]*)', outT) + if(mD and mT): + # get the current time from hardware + utcoffset = int((datetime.now() - datetime.utcnow()).total_seconds()) + dt = datetime(\ + int(mD.group('y')), int(mD.group('m')), int(mD.group('d')), + int(mT.group('h')), int(mT.group('m')), int(mT.group('s'))) + nowtime = int(dt.strftime('%s')) + utcoffset + else: + # if hardware time fails, use the software time + nowtime = int(datetime.now().strftime('%s')) + alarm = nowtime + self.rtcwaketime + call('echo %d > %s/wakealarm' % (alarm, self.rtcpath), shell=True) + def rtcWakeAlarmOff(self): + call('echo 0 > %s/wakealarm' % self.rtcpath, shell=True) + def initdmesg(self): + # get the latest time stamp from the dmesg log + fp = Popen('dmesg', stdout=PIPE).stdout + ktime = '0' + for line in fp: + line = line.replace('\r\n', '') + idx = line.find('[') + if idx > 1: + line = line[idx:] + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) + if(m): + ktime = m.group('ktime') + fp.close() + self.dmesgstart = float(ktime) + def getdmesg(self): + # store all new dmesg lines since initdmesg was called + fp = Popen('dmesg', stdout=PIPE).stdout + op = open(self.dmesgfile, 'a') + for line in fp: + line = line.replace('\r\n', '') + idx = line.find('[') + if idx > 1: + line = line[idx:] + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) + if(not m): + continue + ktime = float(m.group('ktime')) + if ktime > self.dmesgstart: + op.write(line) + fp.close() + op.close() + def addFtraceFilterFunctions(self, file): + fp = open(file) + list = fp.read().split('\n') + fp.close() + for i in list: + if len(i) < 2: + continue + self.tracefuncs[i] = dict() + def getFtraceFilterFunctions(self, current): + rootCheck(True) + if not current: + call('cat '+self.tpath+'available_filter_functions', shell=True) + return + fp = open(self.tpath+'available_filter_functions') + master = fp.read().split('\n') + fp.close() + for i in self.tracefuncs: + if 'func' in self.tracefuncs[i]: + i = self.tracefuncs[i]['func'] + if i in master: + print i + else: + print self.colorText(i) + def setFtraceFilterFunctions(self, list): + fp = open(self.tpath+'available_filter_functions') + master = fp.read().split('\n') + fp.close() + flist = '' + for i in list: + if i not in master: + continue + if ' [' in i: + flist += i.split(' ')[0]+'\n' + else: + flist += i+'\n' + fp = open(self.tpath+'set_graph_function', 'w') + fp.write(flist) + fp.close() + def basicKprobe(self, name): + self.kprobes[name] = {'name': name,'func': name,'args': dict(),'format': name} + def defaultKprobe(self, name, kdata): + k = kdata + for field in ['name', 'format', 'func']: + if field not in k: + k[field] = name + if self.archargs in k: + k['args'] = k[self.archargs] + else: + k['args'] = dict() + k['format'] = name + self.kprobes[name] = k + def kprobeColor(self, name): + if name not in self.kprobes or 'color' not in self.kprobes[name]: + return '' + return self.kprobes[name]['color'] + def kprobeDisplayName(self, name, dataraw): + if name not in self.kprobes: + self.basicKprobe(name) + data = '' + quote=0 + # first remvoe any spaces inside quotes, and the quotes + for c in dataraw: + if c == '"': + quote = (quote + 1) % 2 + if quote and c == ' ': + data += '_' + elif c != '"': + data += c + fmt, args = self.kprobes[name]['format'], self.kprobes[name]['args'] + arglist = dict() + # now process the args + for arg in sorted(args): + arglist[arg] = '' + m = re.match('.* '+arg+'=(?P<arg>.*) ', data); + if m: + arglist[arg] = m.group('arg') + else: + m = re.match('.* '+arg+'=(?P<arg>.*)', data); + if m: + arglist[arg] = m.group('arg') + out = fmt.format(**arglist) + out = out.replace(' ', '_').replace('"', '') + return out + def kprobeText(self, kname, kprobe): + name = fmt = func = kname + args = dict() + if 'name' in kprobe: + name = kprobe['name'] + if 'format' in kprobe: + fmt = kprobe['format'] + if 'func' in kprobe: + func = kprobe['func'] + if self.archargs in kprobe: + args = kprobe[self.archargs] + if 'args' in kprobe: + args = kprobe['args'] + if re.findall('{(?P<n>[a-z,A-Z,0-9]*)}', func): + doError('Kprobe "%s" has format info in the function name "%s"' % (name, func)) + for arg in re.findall('{(?P<n>[a-z,A-Z,0-9]*)}', fmt): + if arg not in args: + doError('Kprobe "%s" is missing argument "%s"' % (name, arg)) + val = 'p:%s_cal %s' % (name, func) + for i in sorted(args): + val += ' %s=%s' % (i, args[i]) + val += '\nr:%s_ret %s $retval\n' % (name, func) + return val + def addKprobes(self, output=False): + if len(sysvals.kprobes) < 1: + return + if output: + print(' kprobe functions in this kernel:') + # first test each kprobe + rejects = [] + # sort kprobes: trace, ub-dev, custom, dev + kpl = [[], [], [], []] + for name in sorted(self.kprobes): + res = self.colorText('YES', 32) + if not self.testKprobe(name, self.kprobes[name]): + res = self.colorText('NO') + rejects.append(name) + else: + if name in self.tracefuncs: + kpl[0].append(name) + elif name in self.dev_tracefuncs: + if 'ub' in self.dev_tracefuncs[name]: + kpl[1].append(name) + else: + kpl[3].append(name) + else: + kpl[2].append(name) + if output: + print(' %s: %s' % (name, res)) + kplist = kpl[0] + kpl[1] + kpl[2] + kpl[3] + # remove all failed ones from the list + for name in rejects: + self.kprobes.pop(name) + # set the kprobes all at once + self.fsetVal('', 'kprobe_events') + kprobeevents = '' + for kp in kplist: + kprobeevents += self.kprobeText(kp, self.kprobes[kp]) + self.fsetVal(kprobeevents, 'kprobe_events') + # verify that the kprobes were set as ordered + check = self.fgetVal('kprobe_events') + linesout = len(kprobeevents.split('\n')) - 1 + linesack = len(check.split('\n')) - 1 + if output: + res = '%d/%d' % (linesack, linesout) + if linesack < linesout: + res = self.colorText(res, 31) + else: + res = self.colorText(res, 32) + print(' working kprobe functions enabled: %s' % res) + self.fsetVal('1', 'events/kprobes/enable') + def testKprobe(self, kname, kprobe): + self.fsetVal('0', 'events/kprobes/enable') + kprobeevents = self.kprobeText(kname, kprobe) + if not kprobeevents: + return False + try: + self.fsetVal(kprobeevents, 'kprobe_events') + check = self.fgetVal('kprobe_events') + except: + return False + linesout = len(kprobeevents.split('\n')) + linesack = len(check.split('\n')) + if linesack < linesout: + return False + return True + def fsetVal(self, val, path, mode='w'): + file = self.tpath+path + if not os.path.exists(file): + return False + try: + fp = open(file, mode, 0) + fp.write(val) + fp.flush() + fp.close() + except: + pass + return True + def fgetVal(self, path): + file = self.tpath+path + res = '' + if not os.path.exists(file): + return res + try: + fp = open(file, 'r') + res = fp.read() + fp.close() + except: + pass + return res + def cleanupFtrace(self): + if(self.usecallgraph or self.usetraceevents): + self.fsetVal('0', 'events/kprobes/enable') + self.fsetVal('', 'kprobe_events') + def setupAllKprobes(self): + for name in self.tracefuncs: + self.defaultKprobe(name, self.tracefuncs[name]) + for name in self.dev_tracefuncs: + self.defaultKprobe(name, self.dev_tracefuncs[name]) + def isCallgraphFunc(self, name): + if len(self.tracefuncs) < 1 and self.suspendmode == 'command': + return True + for i in self.tracefuncs: + if 'func' in self.tracefuncs[i]: + f = self.tracefuncs[i]['func'] + else: + f = i + if name == f: + return True + return False + def initFtrace(self, testing=False): + print('INITIALIZING FTRACE...') + # turn trace off + self.fsetVal('0', 'tracing_on') + self.cleanupFtrace() + # set the trace clock to global + self.fsetVal('global', 'trace_clock') + # set trace buffer to a huge value + self.fsetVal('nop', 'current_tracer') + self.fsetVal('131073', 'buffer_size_kb') + # go no further if this is just a status check + if testing: + return + # initialize the callgraph trace + if(self.usecallgraph): + # set trace type + self.fsetVal('function_graph', 'current_tracer') + self.fsetVal('', 'set_ftrace_filter') + # set trace format options + self.fsetVal('print-parent', 'trace_options') + self.fsetVal('funcgraph-abstime', 'trace_options') + self.fsetVal('funcgraph-cpu', 'trace_options') + self.fsetVal('funcgraph-duration', 'trace_options') + self.fsetVal('funcgraph-proc', 'trace_options') + self.fsetVal('funcgraph-tail', 'trace_options') + self.fsetVal('nofuncgraph-overhead', 'trace_options') + self.fsetVal('context-info', 'trace_options') + self.fsetVal('graph-time', 'trace_options') + self.fsetVal('%d' % self.max_graph_depth, 'max_graph_depth') + cf = ['dpm_run_callback'] + if(self.usetraceeventsonly): + cf += ['dpm_prepare', 'dpm_complete'] + for fn in self.tracefuncs: + if 'func' in self.tracefuncs[fn]: + cf.append(self.tracefuncs[fn]['func']) + else: + cf.append(fn) + self.setFtraceFilterFunctions(cf) + # initialize the kprobe trace + elif self.usekprobes: + for name in self.tracefuncs: + self.defaultKprobe(name, self.tracefuncs[name]) + if self.usedevsrc: + for name in self.dev_tracefuncs: + self.defaultKprobe(name, self.dev_tracefuncs[name]) + print('INITIALIZING KPROBES...') + self.addKprobes(self.verbose) + if(self.usetraceevents): + # turn trace events on + events = iter(self.traceevents) + for e in events: + self.fsetVal('1', 'events/power/'+e+'/enable') + # clear the trace buffer + self.fsetVal('', 'trace') + def verifyFtrace(self): + # files needed for any trace data + files = ['buffer_size_kb', 'current_tracer', 'trace', 'trace_clock', + 'trace_marker', 'trace_options', 'tracing_on'] + # files needed for callgraph trace data + tp = self.tpath + if(self.usecallgraph): + files += [ + 'available_filter_functions', + 'set_ftrace_filter', + 'set_graph_function' + ] + for f in files: + if(os.path.exists(tp+f) == False): + return False + return True + def verifyKprobes(self): + # files needed for kprobes to work + files = ['kprobe_events', 'events'] + tp = self.tpath + for f in files: + if(os.path.exists(tp+f) == False): + return False + return True + def colorText(self, str, color=31): + if not self.ansi: + return str + return '\x1B[%d;40m%s\x1B[m' % (color, str) + +sysvals = SystemValues() +suspendmodename = { + 'freeze': 'Freeze (S0)', + 'standby': 'Standby (S1)', + 'mem': 'Suspend (S3)', + 'disk': 'Hibernate (S4)' +} + +# Class: DevProps +# Description: +# Simple class which holds property values collected +# for all the devices used in the timeline. +class DevProps: + syspath = '' + altname = '' + async = True + xtraclass = '' + xtrainfo = '' + def out(self, dev): + return '%s,%s,%d;' % (dev, self.altname, self.async) + def debug(self, dev): + print '%s:\n\taltname = %s\n\t async = %s' % (dev, self.altname, self.async) + def altName(self, dev): + if not self.altname or self.altname == dev: + return dev + return '%s [%s]' % (self.altname, dev) + def xtraClass(self): + if self.xtraclass: + return ' '+self.xtraclass + if not self.async: + return ' sync' + return '' + def xtraInfo(self): + if self.xtraclass: + return ' '+self.xtraclass + if self.async: + return ' async_device' + return ' sync_device' + +# Class: DeviceNode +# Description: +# A container used to create a device hierachy, with a single root node +# and a tree of child nodes. Used by Data.deviceTopology() +class DeviceNode: + name = '' + children = 0 + depth = 0 + def __init__(self, nodename, nodedepth): + self.name = nodename + self.children = [] + self.depth = nodedepth + +# Class: Data +# Description: +# The primary container for suspend/resume test data. There is one for +# each test run. The data is organized into a cronological hierarchy: +# Data.dmesg { +# phases { +# 10 sequential, non-overlapping phases of S/R +# contents: times for phase start/end, order/color data for html +# devlist { +# device callback or action list for this phase +# device { +# a single device callback or generic action +# contents: start/stop times, pid/cpu/driver info +# parents/children, html id for timeline/callgraph +# optionally includes an ftrace callgraph +# optionally includes dev/ps data +# } +# } +# } +# } +# +class Data: + dmesg = {} # root data structure + phases = [] # ordered list of phases + start = 0.0 # test start + end = 0.0 # test end + tSuspended = 0.0 # low-level suspend start + tResumed = 0.0 # low-level resume start + tKernSus = 0.0 # kernel level suspend start + tKernRes = 0.0 # kernel level resume end + tLow = 0.0 # time spent in low-level suspend (standby/freeze) + fwValid = False # is firmware data available + fwSuspend = 0 # time spent in firmware suspend + fwResume = 0 # time spent in firmware resume + dmesgtext = [] # dmesg text file in memory + pstl = 0 # process timeline + testnumber = 0 + idstr = '' + html_device_id = 0 + stamp = 0 + outfile = '' + devpids = [] + kerror = False + def __init__(self, num): + idchar = 'abcdefghij' + self.pstl = dict() + self.testnumber = num + self.idstr = idchar[num] + self.dmesgtext = [] + self.phases = [] + self.dmesg = { # fixed list of 10 phases + 'suspend_prepare': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#CCFFCC', 'order': 0}, + 'suspend': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#88FF88', 'order': 1}, + 'suspend_late': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#00AA00', 'order': 2}, + 'suspend_noirq': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#008888', 'order': 3}, + 'suspend_machine': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#0000FF', 'order': 4}, + 'resume_machine': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#FF0000', 'order': 5}, + 'resume_noirq': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#FF9900', 'order': 6}, + 'resume_early': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#FFCC00', 'order': 7}, + 'resume': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#FFFF88', 'order': 8}, + 'resume_complete': {'list': dict(), 'start': -1.0, 'end': -1.0, + 'row': 0, 'color': '#FFFFCC', 'order': 9} + } + self.phases = self.sortedPhases() + self.devicegroups = [] + for phase in self.phases: + self.devicegroups.append([phase]) + self.errorinfo = {'suspend':[],'resume':[]} + def extractErrorInfo(self, dmesg): + error = '' + tm = 0.0 + for i in range(len(dmesg)): + if 'Call Trace:' in dmesg[i]: + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) .*', dmesg[i]) + if not m: + continue + tm = float(m.group('ktime')) + if tm < self.start or tm > self.end: + continue + for j in range(i-10, i+1): + error += dmesg[j] + continue + if error: + m = re.match('[ \t]*\[ *[0-9\.]*\] \[\<[0-9a-fA-F]*\>\] .*', dmesg[i]) + if m: + error += dmesg[i] + else: + if tm < self.tSuspended: + dir = 'suspend' + else: + dir = 'resume' + error = error.replace('<', '<').replace('>', '>') + vprint('kernel error found in %s at %f' % (dir, tm)) + self.errorinfo[dir].append((tm, error)) + self.kerror = True + error = '' + def setStart(self, time): + self.start = time + def setEnd(self, time): + self.end = time + def isTraceEventOutsideDeviceCalls(self, pid, time): + for phase in self.phases: + list = self.dmesg[phase]['list'] + for dev in list: + d = list[dev] + if(d['pid'] == pid and time >= d['start'] and + time < d['end']): + return False + return True + def sourcePhase(self, start): + for phase in self.phases: + pend = self.dmesg[phase]['end'] + if start <= pend: + return phase + return 'resume_complete' + def sourceDevice(self, phaselist, start, end, pid, type): + tgtdev = '' + for phase in phaselist: + list = self.dmesg[phase]['list'] + for devname in list: + dev = list[devname] + # pid must match + if dev['pid'] != pid: + continue + devS = dev['start'] + devE = dev['end'] + if type == 'device': + # device target event is entirely inside the source boundary + if(start < devS or start >= devE or end <= devS or end > devE): + continue + elif type == 'thread': + # thread target event will expand the source boundary + if start < devS: + dev['start'] = start + if end > devE: + dev['end'] = end + tgtdev = dev + break + return tgtdev + def addDeviceFunctionCall(self, displayname, kprobename, proc, pid, start, end, cdata, rdata): + # try to place the call in a device + tgtdev = self.sourceDevice(self.phases, start, end, pid, 'device') + # calls with device pids that occur outside device bounds are dropped + # TODO: include these somehow + if not tgtdev and pid in self.devpids: + return False + # try to place the call in a thread + if not tgtdev: + tgtdev = self.sourceDevice(self.phases, start, end, pid, 'thread') + # create new thread blocks, expand as new calls are found + if not tgtdev: + if proc == '<...>': + threadname = 'kthread-%d' % (pid) + else: + threadname = '%s-%d' % (proc, pid) + tgtphase = self.sourcePhase(start) + self.newAction(tgtphase, threadname, pid, '', start, end, '', ' kth', '') + return self.addDeviceFunctionCall(displayname, kprobename, proc, pid, start, end, cdata, rdata) + # this should not happen + if not tgtdev: + vprint('[%f - %f] %s-%d %s %s %s' % \ + (start, end, proc, pid, kprobename, cdata, rdata)) + return False + # place the call data inside the src element of the tgtdev + if('src' not in tgtdev): + tgtdev['src'] = [] + dtf = sysvals.dev_tracefuncs + ubiquitous = False + if kprobename in dtf and 'ub' in dtf[kprobename]: + ubiquitous = True + title = cdata+' '+rdata + mstr = '\(.*\) *(?P<args>.*) *\((?P<caller>.*)\+.* arg1=(?P<ret>.*)' + m = re.match(mstr, title) + if m: + c = m.group('caller') + a = m.group('args').strip() + r = m.group('ret') + if len(r) > 6: + r = '' + else: + r = 'ret=%s ' % r + if ubiquitous and c in dtf and 'ub' in dtf[c]: + return False + color = sysvals.kprobeColor(kprobename) + e = DevFunction(displayname, a, c, r, start, end, ubiquitous, proc, pid, color) + tgtdev['src'].append(e) + return True + def overflowDevices(self): + # get a list of devices that extend beyond the end of this test run + devlist = [] + for phase in self.phases: + list = self.dmesg[phase]['list'] + for devname in list: + dev = list[devname] + if dev['end'] > self.end: + devlist.append(dev) + return devlist + def mergeOverlapDevices(self, devlist): + # merge any devices that overlap devlist + for dev in devlist: + devname = dev['name'] + for phase in self.phases: + list = self.dmesg[phase]['list'] + if devname not in list: + continue + tdev = list[devname] + o = min(dev['end'], tdev['end']) - max(dev['start'], tdev['start']) + if o <= 0: + continue + dev['end'] = tdev['end'] + if 'src' not in dev or 'src' not in tdev: + continue + dev['src'] += tdev['src'] + del list[devname] + def usurpTouchingThread(self, name, dev): + # the caller test has priority of this thread, give it to him + for phase in self.phases: + list = self.dmesg[phase]['list'] + if name in list: + tdev = list[name] + if tdev['start'] - dev['end'] < 0.1: + dev['end'] = tdev['end'] + if 'src' not in dev: + dev['src'] = [] + if 'src' in tdev: + dev['src'] += tdev['src'] + del list[name] + break + def stitchTouchingThreads(self, testlist): + # merge any threads between tests that touch + for phase in self.phases: + list = self.dmesg[phase]['list'] + for devname in list: + dev = list[devname] + if 'htmlclass' not in dev or 'kth' not in dev['htmlclass']: + continue + for data in testlist: + data.usurpTouchingThread(devname, dev) + def optimizeDevSrc(self): + # merge any src call loops to reduce timeline size + for phase in self.phases: + list = self.dmesg[phase]['list'] + for dev in list: + if 'src' not in list[dev]: + continue + src = list[dev]['src'] + p = 0 + for e in sorted(src, key=lambda event: event.time): + if not p or not e.repeat(p): + p = e + continue + # e is another iteration of p, move it into p + p.end = e.end + p.length = p.end - p.time + p.count += 1 + src.remove(e) + def trimTimeVal(self, t, t0, dT, left): + if left: + if(t > t0): + if(t - dT < t0): + return t0 + return t - dT + else: + return t + else: + if(t < t0 + dT): + if(t > t0): + return t0 + dT + return t + dT + else: + return t + def trimTime(self, t0, dT, left): + self.tSuspended = self.trimTimeVal(self.tSuspended, t0, dT, left) + self.tResumed = self.trimTimeVal(self.tResumed, t0, dT, left) + self.start = self.trimTimeVal(self.start, t0, dT, left) + self.tKernSus = self.trimTimeVal(self.tKernSus, t0, dT, left) + self.tKernRes = self.trimTimeVal(self.tKernRes, t0, dT, left) + self.end = self.trimTimeVal(self.end, t0, dT, left) + for phase in self.phases: + p = self.dmesg[phase] + p['start'] = self.trimTimeVal(p['start'], t0, dT, left) + p['end'] = self.trimTimeVal(p['end'], t0, dT, left) + list = p['list'] + for name in list: + d = list[name] + d['start'] = self.trimTimeVal(d['start'], t0, dT, left) + d['end'] = self.trimTimeVal(d['end'], t0, dT, left) + if('ftrace' in d): + cg = d['ftrace'] + cg.start = self.trimTimeVal(cg.start, t0, dT, left) + cg.end = self.trimTimeVal(cg.end, t0, dT, left) + for line in cg.list: + line.time = self.trimTimeVal(line.time, t0, dT, left) + if('src' in d): + for e in d['src']: + e.time = self.trimTimeVal(e.time, t0, dT, left) + def normalizeTime(self, tZero): + # trim out any standby or freeze clock time + if(self.tSuspended != self.tResumed): + if(self.tResumed > tZero): + self.trimTime(self.tSuspended, \ + self.tResumed-self.tSuspended, True) + else: + self.trimTime(self.tSuspended, \ + self.tResumed-self.tSuspended, False) + def setPhase(self, phase, ktime, isbegin): + if(isbegin): + self.dmesg[phase]['start'] = ktime + else: + self.dmesg[phase]['end'] = ktime + def dmesgSortVal(self, phase): + return self.dmesg[phase]['order'] + def sortedPhases(self): + return sorted(self.dmesg, key=self.dmesgSortVal) + def sortedDevices(self, phase): + list = self.dmesg[phase]['list'] + slist = [] + tmp = dict() + for devname in list: + dev = list[devname] + if dev['length'] == 0: + continue + tmp[dev['start']] = devname + for t in sorted(tmp): + slist.append(tmp[t]) + return slist + def fixupInitcalls(self, phase): + # if any calls never returned, clip them at system resume end + phaselist = self.dmesg[phase]['list'] + for devname in phaselist: + dev = phaselist[devname] + if(dev['end'] < 0): + for p in self.phases: + if self.dmesg[p]['end'] > dev['start']: + dev['end'] = self.dmesg[p]['end'] + break + vprint('%s (%s): callback didnt return' % (devname, phase)) + def deviceFilter(self, devicefilter): + for phase in self.phases: + list = self.dmesg[phase]['list'] + rmlist = [] + for name in list: + keep = False + for filter in devicefilter: + if filter in name or \ + ('drv' in list[name] and filter in list[name]['drv']): + keep = True + if not keep: + rmlist.append(name) + for name in rmlist: + del list[name] + def fixupInitcallsThatDidntReturn(self): + # if any calls never returned, clip them at system resume end + for phase in self.phases: + self.fixupInitcalls(phase) + def phaseOverlap(self, phases): + rmgroups = [] + newgroup = [] + for group in self.devicegroups: + for phase in phases: + if phase not in group: + continue + for p in group: + if p not in newgroup: + newgroup.append(p) + if group not in rmgroups: + rmgroups.append(group) + for group in rmgroups: + self.devicegroups.remove(group) + self.devicegroups.append(newgroup) + def newActionGlobal(self, name, start, end, pid=-1, color=''): + # which phase is this device callback or action in + targetphase = 'none' + htmlclass = '' + overlap = 0.0 + phases = [] + for phase in self.phases: + pstart = self.dmesg[phase]['start'] + pend = self.dmesg[phase]['end'] + # see if the action overlaps this phase + o = max(0, min(end, pend) - max(start, pstart)) + if o > 0: + phases.append(phase) + # set the target phase to the one that overlaps most + if o > overlap: + if overlap > 0 and phase == 'post_resume': + continue + targetphase = phase + overlap = o + # if no target phase was found, pin it to the edge + if targetphase == 'none': + p0start = self.dmesg[self.phases[0]]['start'] + if start <= p0start: + targetphase = self.phases[0] + else: + targetphase = self.phases[-1] + if pid == -2: + htmlclass = ' bg' + elif pid == -3: + htmlclass = ' ps' + if len(phases) > 1: + htmlclass = ' bg' + self.phaseOverlap(phases) + if targetphase in self.phases: + newname = self.newAction(targetphase, name, pid, '', start, end, '', htmlclass, color) + return (targetphase, newname) + return False + def newAction(self, phase, name, pid, parent, start, end, drv, htmlclass='', color=''): + # new device callback for a specific phase + self.html_device_id += 1 + devid = '%s%d' % (self.idstr, self.html_device_id) + list = self.dmesg[phase]['list'] + length = -1.0 + if(start >= 0 and end >= 0): + length = end - start + if pid == -2: + i = 2 + origname = name + while(name in list): + name = '%s[%d]' % (origname, i) + i += 1 + list[name] = {'name': name, 'start': start, 'end': end, 'pid': pid, + 'par': parent, 'length': length, 'row': 0, 'id': devid, 'drv': drv } + if htmlclass: + list[name]['htmlclass'] = htmlclass + if color: + list[name]['color'] = color + return name + def deviceChildren(self, devname, phase): + devlist = [] + list = self.dmesg[phase]['list'] + for child in list: + if(list[child]['par'] == devname): + devlist.append(child) + return devlist + def printDetails(self): + vprint('Timeline Details:') + vprint(' test start: %f' % self.start) + vprint('kernel suspend start: %f' % self.tKernSus) + for phase in self.phases: + dc = len(self.dmesg[phase]['list']) + vprint(' %16s: %f - %f (%d devices)' % (phase, \ + self.dmesg[phase]['start'], self.dmesg[phase]['end'], dc)) + vprint(' kernel resume end: %f' % self.tKernRes) + vprint(' test end: %f' % self.end) + def deviceChildrenAllPhases(self, devname): + devlist = [] + for phase in self.phases: + list = self.deviceChildren(devname, phase) + for dev in list: + if dev not in devlist: + devlist.append(dev) + return devlist + def masterTopology(self, name, list, depth): + node = DeviceNode(name, depth) + for cname in list: + # avoid recursions + if name == cname: + continue + clist = self.deviceChildrenAllPhases(cname) + cnode = self.masterTopology(cname, clist, depth+1) + node.children.append(cnode) + return node + def printTopology(self, node): + html = '' + if node.name: + info = '' + drv = '' + for phase in self.phases: + list = self.dmesg[phase]['list'] + if node.name in list: + s = list[node.name]['start'] + e = list[node.name]['end'] + if list[node.name]['drv']: + drv = ' {'+list[node.name]['drv']+'}' + info += ('<li>%s: %.3fms</li>' % (phase, (e-s)*1000)) + html += '<li><b>'+node.name+drv+'</b>' + if info: + html += '<ul>'+info+'</ul>' + html += '</li>' + if len(node.children) > 0: + html += '<ul>' + for cnode in node.children: + html += self.printTopology(cnode) + html += '</ul>' + return html + def rootDeviceList(self): + # list of devices graphed + real = [] + for phase in self.dmesg: + list = self.dmesg[phase]['list'] + for dev in list: + if list[dev]['pid'] >= 0 and dev not in real: + real.append(dev) + # list of top-most root devices + rootlist = [] + for phase in self.dmesg: + list = self.dmesg[phase]['list'] + for dev in list: + pdev = list[dev]['par'] + pid = list[dev]['pid'] + if(pid < 0 or re.match('[0-9]*-[0-9]*\.[0-9]*[\.0-9]*\:[\.0-9]*$', pdev)): + continue + if pdev and pdev not in real and pdev not in rootlist: + rootlist.append(pdev) + return rootlist + def deviceTopology(self): + rootlist = self.rootDeviceList() + master = self.masterTopology('', rootlist, 0) + return self.printTopology(master) + def selectTimelineDevices(self, widfmt, tTotal, mindevlen): + # only select devices that will actually show up in html + self.tdevlist = dict() + for phase in self.dmesg: + devlist = [] + list = self.dmesg[phase]['list'] + for dev in list: + length = (list[dev]['end'] - list[dev]['start']) * 1000 + width = widfmt % (((list[dev]['end']-list[dev]['start'])*100)/tTotal) + if width != '0.000000' and length >= mindevlen: + devlist.append(dev) + self.tdevlist[phase] = devlist + def addHorizontalDivider(self, devname, devend): + phase = 'suspend_prepare' + self.newAction(phase, devname, -2, '', \ + self.start, devend, '', ' sec', '') + if phase not in self.tdevlist: + self.tdevlist[phase] = [] + self.tdevlist[phase].append(devname) + d = DevItem(0, phase, self.dmesg[phase]['list'][devname]) + return d + def addProcessUsageEvent(self, name, times): + # get the start and end times for this process + maxC = 0 + tlast = 0 + start = -1 + end = -1 + for t in sorted(times): + if tlast == 0: + tlast = t + continue + if name in self.pstl[t]: + if start == -1 or tlast < start: + start = tlast + if end == -1 or t > end: + end = t + tlast = t + if start == -1 or end == -1: + return 0 + # add a new action for this process and get the object + out = self.newActionGlobal(name, start, end, -3) + if not out: + return 0 + phase, devname = out + dev = self.dmesg[phase]['list'][devname] + # get the cpu exec data + tlast = 0 + clast = 0 + cpuexec = dict() + for t in sorted(times): + if tlast == 0 or t <= start or t > end: + tlast = t + continue + list = self.pstl[t] + c = 0 + if name in list: + c = list[name] + if c > maxC: + maxC = c + if c != clast: + key = (tlast, t) + cpuexec[key] = c + tlast = t + clast = c + dev['cpuexec'] = cpuexec + return maxC + def createProcessUsageEvents(self): + # get an array of process names + proclist = [] + for t in self.pstl: + pslist = self.pstl[t] + for ps in pslist: + if ps not in proclist: + proclist.append(ps) + # get a list of data points for suspend and resume + tsus = [] + tres = [] + for t in sorted(self.pstl): + if t < self.tSuspended: + tsus.append(t) + else: + tres.append(t) + # process the events for suspend and resume + if len(proclist) > 0: + vprint('Process Execution:') + for ps in proclist: + c = self.addProcessUsageEvent(ps, tsus) + if c > 0: + vprint('%25s (sus): %d' % (ps, c)) + c = self.addProcessUsageEvent(ps, tres) + if c > 0: + vprint('%25s (res): %d' % (ps, c)) + +# Class: DevFunction +# Description: +# A container for kprobe function data we want in the dev timeline +class DevFunction: + row = 0 + count = 1 + def __init__(self, name, args, caller, ret, start, end, u, proc, pid, color): + self.name = name + self.args = args + self.caller = caller + self.ret = ret + self.time = start + self.length = end - start + self.end = end + self.ubiquitous = u + self.proc = proc + self.pid = pid + self.color = color + def title(self): + cnt = '' + if self.count > 1: + cnt = '(x%d)' % self.count + l = '%0.3fms' % (self.length * 1000) + if self.ubiquitous: + title = '%s(%s)%s <- %s, %s(%s)' % \ + (self.name, self.args, cnt, self.caller, self.ret, l) + else: + title = '%s(%s) %s%s(%s)' % (self.name, self.args, self.ret, cnt, l) + return title.replace('"', '') + def text(self): + if self.count > 1: + text = '%s(x%d)' % (self.name, self.count) + else: + text = self.name + return text + def repeat(self, tgt): + # is the tgt call just a repeat of this call (e.g. are we in a loop) + dt = self.time - tgt.end + # only combine calls if -all- attributes are identical + if tgt.caller == self.caller and \ + tgt.name == self.name and tgt.args == self.args and \ + tgt.proc == self.proc and tgt.pid == self.pid and \ + tgt.ret == self.ret and dt >= 0 and \ + dt <= sysvals.callloopmaxgap and \ + self.length < sysvals.callloopmaxlen: + return True + return False + +# Class: FTraceLine +# Description: +# A container for a single line of ftrace data. There are six basic types: +# callgraph line: +# call: " dpm_run_callback() {" +# return: " }" +# leaf: " dpm_run_callback();" +# trace event: +# tracing_mark_write: SUSPEND START or RESUME COMPLETE +# suspend_resume: phase or custom exec block data +# device_pm_callback: device callback info +class FTraceLine: + time = 0.0 + length = 0.0 + fcall = False + freturn = False + fevent = False + fkprobe = False + depth = 0 + name = '' + type = '' + def __init__(self, t, m='', d=''): + self.time = float(t) + if not m and not d: + return + # is this a trace event + if(d == 'traceevent' or re.match('^ *\/\* *(?P<msg>.*) \*\/ *$', m)): + if(d == 'traceevent'): + # nop format trace event + msg = m + else: + # function_graph format trace event + em = re.match('^ *\/\* *(?P<msg>.*) \*\/ *$', m) + msg = em.group('msg') + + emm = re.match('^(?P<call>.*?): (?P<msg>.*)', msg) + if(emm): + self.name = emm.group('msg') + self.type = emm.group('call') + else: + self.name = msg + km = re.match('^(?P<n>.*)_cal$', self.type) + if km: + self.fcall = True + self.fkprobe = True + self.type = km.group('n') + return + km = re.match('^(?P<n>.*)_ret$', self.type) + if km: + self.freturn = True + self.fkprobe = True + self.type = km.group('n') + return + self.fevent = True + return + # convert the duration to seconds + if(d): + self.length = float(d)/1000000 + # the indentation determines the depth + match = re.match('^(?P<d> *)(?P<o>.*)$', m) + if(not match): + return + self.depth = self.getDepth(match.group('d')) + m = match.group('o') + # function return + if(m[0] == '}'): + self.freturn = True + if(len(m) > 1): + # includes comment with function name + match = re.match('^} *\/\* *(?P<n>.*) *\*\/$', m) + if(match): + self.name = match.group('n').strip() + # function call + else: + self.fcall = True + # function call with children + if(m[-1] == '{'): + match = re.match('^(?P<n>.*) *\(.*', m) + if(match): + self.name = match.group('n').strip() + # function call with no children (leaf) + elif(m[-1] == ';'): + self.freturn = True + match = re.match('^(?P<n>.*) *\(.*', m) + if(match): + self.name = match.group('n').strip() + # something else (possibly a trace marker) + else: + self.name = m + def getDepth(self, str): + return len(str)/2 + def debugPrint(self, dev=''): + if(self.freturn and self.fcall): + print('%s -- %f (%02d): %s(); (%.3f us)' % (dev, self.time, \ + self.depth, self.name, self.length*1000000)) + elif(self.freturn): + print('%s -- %f (%02d): %s} (%.3f us)' % (dev, self.time, \ + self.depth, self.name, self.length*1000000)) + else: + print('%s -- %f (%02d): %s() { (%.3f us)' % (dev, self.time, \ + self.depth, self.name, self.length*1000000)) + def startMarker(self): + # Is this the starting line of a suspend? + if not self.fevent: + return False + if sysvals.usetracemarkers: + if(self.name == 'SUSPEND START'): + return True + return False + else: + if(self.type == 'suspend_resume' and + re.match('suspend_enter\[.*\] begin', self.name)): + return True + return False + def endMarker(self): + # Is this the ending line of a resume? + if not self.fevent: + return False + if sysvals.usetracemarkers: + if(self.name == 'RESUME COMPLETE'): + return True + return False + else: + if(self.type == 'suspend_resume' and + re.match('thaw_processes\[.*\] end', self.name)): + return True + return False + +# Class: FTraceCallGraph +# Description: +# A container for the ftrace callgraph of a single recursive function. +# This can be a dpm_run_callback, dpm_prepare, or dpm_complete callgraph +# Each instance is tied to a single device in a single phase, and is +# comprised of an ordered list of FTraceLine objects +class FTraceCallGraph: + id = '' + start = -1.0 + end = -1.0 + list = [] + invalid = False + depth = 0 + pid = 0 + name = '' + def __init__(self, pid): + self.start = -1.0 + self.end = -1.0 + self.list = [] + self.depth = 0 + self.pid = pid + def addLine(self, line, debug=False): + # if this is already invalid, just leave + if(self.invalid): + return False + # invalidate on too much data or bad depth + if(len(self.list) >= 1000000 or self.depth < 0): + self.invalidate(line) + return False + # compare current depth with this lines pre-call depth + prelinedep = line.depth + if(line.freturn and not line.fcall): + prelinedep += 1 + last = 0 + lasttime = line.time + virtualfname = 'execution_misalignment' + if len(self.list) > 0: + last = self.list[-1] + lasttime = last.time + # handle low misalignments by inserting returns + if prelinedep < self.depth: + if debug and last: + print '-------- task %d --------' % self.pid + last.debugPrint() + idx = 0 + # add return calls to get the depth down + while prelinedep < self.depth: + if debug: + print 'MISALIGN LOW (add returns): C%d - eC%d' % (self.depth, prelinedep) + self.depth -= 1 + if idx == 0 and last and last.fcall and not last.freturn: + # special case, turn last call into a leaf + last.depth = self.depth + last.freturn = True + last.length = line.time - last.time + if debug: + last.debugPrint() + else: + vline = FTraceLine(lasttime) + vline.depth = self.depth + vline.name = virtualfname + vline.freturn = True + self.list.append(vline) + if debug: + vline.debugPrint() + idx += 1 + if debug: + line.debugPrint() + print '' + # handle high misalignments by inserting calls + elif prelinedep > self.depth: + if debug and last: + print '-------- task %d --------' % self.pid + last.debugPrint() + idx = 0 + # add calls to get the depth up + while prelinedep > self.depth: + if debug: + print 'MISALIGN HIGH (add calls): C%d - eC%d' % (self.depth, prelinedep) + if idx == 0 and line.freturn and not line.fcall: + # special case, turn this return into a leaf + line.fcall = True + prelinedep -= 1 + else: + vline = FTraceLine(lasttime) + vline.depth = self.depth + vline.name = virtualfname + vline.fcall = True + if debug: + vline.debugPrint() + self.list.append(vline) + self.depth += 1 + if not last: + self.start = vline.time + idx += 1 + if debug: + line.debugPrint() + print '' + # process the call and set the new depth + if(line.fcall and not line.freturn): + self.depth += 1 + elif(line.freturn and not line.fcall): + self.depth -= 1 + if len(self.list) < 1: + self.start = line.time + self.list.append(line) + if(line.depth == 0 and line.freturn): + if(self.start < 0): + self.start = line.time + self.end = line.time + if line.fcall: + self.end += line.length + if self.list[0].name == virtualfname: + self.invalid = True + return True + return False + def invalidate(self, line): + if(len(self.list) > 0): + first = self.list[0] + self.list = [] + self.list.append(first) + self.invalid = True + id = 'task %s' % (self.pid) + window = '(%f - %f)' % (self.start, line.time) + if(self.depth < 0): + vprint('Too much data for '+id+\ + ' (buffer overflow), ignoring this callback') + else: + vprint('Too much data for '+id+\ + ' '+window+', ignoring this callback') + def slice(self, t0, tN): + minicg = FTraceCallGraph(0) + count = -1 + firstdepth = 0 + for l in self.list: + if(l.time < t0 or l.time > tN): + continue + if(count < 0): + if(not l.fcall or l.name == 'dev_driver_string'): + continue + firstdepth = l.depth + count = 0 + l.depth -= firstdepth + minicg.addLine(l) + if((count == 0 and l.freturn and l.fcall) or + (count > 0 and l.depth <= 0)): + break + count += 1 + return minicg + def repair(self, enddepth): + # bring the depth back to 0 with additional returns + fixed = False + last = self.list[-1] + for i in reversed(range(enddepth)): + t = FTraceLine(last.time) + t.depth = i + t.freturn = True + fixed = self.addLine(t) + if fixed: + self.end = last.time + return True + return False + def postProcess(self, debug=False): + if len(self.list) > 0: + self.name = self.list[0].name + stack = dict() + cnt = 0 + last = 0 + for l in self.list: + # ftrace bug: reported duration is not reliable + # check each leaf and clip it at max possible length + if(last and last.freturn and last.fcall): + if last.length > l.time - last.time: + last.length = l.time - last.time + if(l.fcall and not l.freturn): + stack[l.depth] = l + cnt += 1 + elif(l.freturn and not l.fcall): + if(l.depth not in stack): + if debug: + print 'Post Process Error: Depth missing' + l.debugPrint() + return False + # calculate call length from call/return lines + stack[l.depth].length = l.time - stack[l.depth].time + stack.pop(l.depth) + l.length = 0 + cnt -= 1 + last = l + if(cnt == 0): + # trace caught the whole call tree + return True + elif(cnt < 0): + if debug: + print 'Post Process Error: Depth is less than 0' + return False + # trace ended before call tree finished + return self.repair(cnt) + def deviceMatch(self, pid, data): + found = False + # add the callgraph data to the device hierarchy + borderphase = { + 'dpm_prepare': 'suspend_prepare', + 'dpm_complete': 'resume_complete' + } + if(self.name in borderphase): + p = borderphase[self.name] + list = data.dmesg[p]['list'] + for devname in list: + dev = list[devname] + if(pid == dev['pid'] and + self.start <= dev['start'] and + self.end >= dev['end']): + dev['ftrace'] = self.slice(dev['start'], dev['end']) + found = True + return found + for p in data.phases: + if(data.dmesg[p]['start'] <= self.start and + self.start <= data.dmesg[p]['end']): + list = data.dmesg[p]['list'] + for devname in list: + dev = list[devname] + if(pid == dev['pid'] and + self.start <= dev['start'] and + self.end >= dev['end']): + dev['ftrace'] = self + found = True + break + break + return found + def newActionFromFunction(self, data): + name = self.name + if name in ['dpm_run_callback', 'dpm_prepare', 'dpm_complete']: + return + fs = self.start + fe = self.end + if fs < data.start or fe > data.end: + return + phase = '' + for p in data.phases: + if(data.dmesg[p]['start'] <= self.start and + self.start < data.dmesg[p]['end']): + phase = p + break + if not phase: + return + out = data.newActionGlobal(name, fs, fe, -2) + if out: + phase, myname = out + data.dmesg[phase]['list'][myname]['ftrace'] = self + def debugPrint(self): + print('[%f - %f] %s (%d)') % (self.start, self.end, self.name, self.pid) + for l in self.list: + if(l.freturn and l.fcall): + print('%f (%02d): %s(); (%.3f us)' % (l.time, \ + l.depth, l.name, l.length*1000000)) + elif(l.freturn): + print('%f (%02d): %s} (%.3f us)' % (l.time, \ + l.depth, l.name, l.length*1000000)) + else: + print('%f (%02d): %s() { (%.3f us)' % (l.time, \ + l.depth, l.name, l.length*1000000)) + print(' ') + +class DevItem: + def __init__(self, test, phase, dev): + self.test = test + self.phase = phase + self.dev = dev + def isa(self, cls): + if 'htmlclass' in self.dev and cls in self.dev['htmlclass']: + return True + return False + +# Class: Timeline +# Description: +# A container for a device timeline which calculates +# all the html properties to display it correctly +class Timeline: + html = '' + height = 0 # total timeline height + scaleH = 20 # timescale (top) row height + rowH = 30 # device row height + bodyH = 0 # body height + rows = 0 # total timeline rows + rowlines = dict() + rowheight = dict() + html_tblock = '<div id="block{0}" class="tblock" style="left:{1}%;width:{2}%;"><div class="tback" style="height:{3}px"></div>\n' + html_device = '<div id="{0}" title="{1}" class="thread{7}" style="left:{2}%;top:{3}px;height:{4}px;width:{5}%;{8}">{6}</div>\n' + html_phase = '<div class="phase" style="left:{0}%;width:{1}%;top:{2}px;height:{3}px;background:{4}">{5}</div>\n' + html_phaselet = '<div id="{0}" class="phaselet" style="left:{1}%;width:{2}%;background:{3}"></div>\n' + def __init__(self, rowheight, scaleheight): + self.rowH = rowheight + self.scaleH = scaleheight + self.html = '' + def createHeader(self, sv, suppress=''): + if(not sv.stamp['time']): + return + self.html += '<div class="version"><a href="https://01.org/suspendresume">%s v%s</a></div>' \ + % (sv.title, sv.version) + if sv.logmsg and 'log' not in suppress: + self.html += '<button id="showtest" class="logbtn">log</button>' + if sv.addlogs and 'dmesg' not in suppress: + self.html += '<button id="showdmesg" class="logbtn">dmesg</button>' + if sv.addlogs and sv.ftracefile and 'ftrace' not in suppress: + self.html += '<button id="showftrace" class="logbtn">ftrace</button>' + headline_stamp = '<div class="stamp">{0} {1} {2} {3}</div>\n' + self.html += headline_stamp.format(sv.stamp['host'], sv.stamp['kernel'], + sv.stamp['mode'], sv.stamp['time']) + # Function: getDeviceRows + # Description: + # determine how may rows the device funcs will take + # Arguments: + # rawlist: the list of devices/actions for a single phase + # Output: + # The total number of rows needed to display this phase of the timeline + def getDeviceRows(self, rawlist): + # clear all rows and set them to undefined + sortdict = dict() + for item in rawlist: + item.row = -1 + sortdict[item] = item.length + sortlist = sorted(sortdict, key=sortdict.get, reverse=True) + remaining = len(sortlist) + rowdata = dict() + row = 1 + # try to pack each row with as many ranges as possible + while(remaining > 0): + if(row not in rowdata): + rowdata[row] = [] + for i in sortlist: + if(i.row >= 0): + continue + s = i.time + e = i.time + i.length + valid = True + for ritem in rowdata[row]: + rs = ritem.time + re = ritem.time + ritem.length + if(not (((s <= rs) and (e <= rs)) or + ((s >= re) and (e >= re)))): + valid = False + break + if(valid): + rowdata[row].append(i) + i.row = row + remaining -= 1 + row += 1 + return row + # Function: getPhaseRows + # Description: + # Organize the timeline entries into the smallest + # number of rows possible, with no entry overlapping + # Arguments: + # devlist: the list of devices/actions in a group of contiguous phases + # Output: + # The total number of rows needed to display this phase of the timeline + def getPhaseRows(self, devlist, row=0): + # clear all rows and set them to undefined + remaining = len(devlist) + rowdata = dict() + sortdict = dict() + myphases = [] + # initialize all device rows to -1 and calculate devrows + for item in devlist: + dev = item.dev + tp = (item.test, item.phase) + if tp not in myphases: + myphases.append(tp) + dev['row'] = -1 + # sort by length 1st, then name 2nd + sortdict[item] = (float(dev['end']) - float(dev['start']), item.dev['name']) + if 'src' in dev: + dev['devrows'] = self.getDeviceRows(dev['src']) + # sort the devlist by length so that large items graph on top + sortlist = sorted(sortdict, key=sortdict.get, reverse=True) + orderedlist = [] + for item in sortlist: + if item.dev['pid'] == -2: + orderedlist.append(item) + for item in sortlist: + if item not in orderedlist: + orderedlist.append(item) + # try to pack each row with as many devices as possible + while(remaining > 0): + rowheight = 1 + if(row not in rowdata): + rowdata[row] = [] + for item in orderedlist: + dev = item.dev + if(dev['row'] < 0): + s = dev['start'] + e = dev['end'] + valid = True + for ritem in rowdata[row]: + rs = ritem.dev['start'] + re = ritem.dev['end'] + if(not (((s <= rs) and (e <= rs)) or + ((s >= re) and (e >= re)))): + valid = False + break + if(valid): + rowdata[row].append(item) + dev['row'] = row + remaining -= 1 + if 'devrows' in dev and dev['devrows'] > rowheight: + rowheight = dev['devrows'] + for t, p in myphases: + if t not in self.rowlines or t not in self.rowheight: + self.rowlines[t] = dict() + self.rowheight[t] = dict() + if p not in self.rowlines[t] or p not in self.rowheight[t]: + self.rowlines[t][p] = dict() + self.rowheight[t][p] = dict() + rh = self.rowH + # section headers should use a different row height + if len(rowdata[row]) == 1 and \ + 'htmlclass' in rowdata[row][0].dev and \ + 'sec' in rowdata[row][0].dev['htmlclass']: + rh = 15 + self.rowlines[t][p][row] = rowheight + self.rowheight[t][p][row] = rowheight * rh + row += 1 + if(row > self.rows): + self.rows = int(row) + return row + def phaseRowHeight(self, test, phase, row): + return self.rowheight[test][phase][row] + def phaseRowTop(self, test, phase, row): + top = 0 + for i in sorted(self.rowheight[test][phase]): + if i >= row: + break + top += self.rowheight[test][phase][i] + return top + def calcTotalRows(self): + # Calculate the heights and offsets for the header and rows + maxrows = 0 + standardphases = [] + for t in self.rowlines: + for p in self.rowlines[t]: + total = 0 + for i in sorted(self.rowlines[t][p]): + total += self.rowlines[t][p][i] + if total > maxrows: + maxrows = total + if total == len(self.rowlines[t][p]): + standardphases.append((t, p)) + self.height = self.scaleH + (maxrows*self.rowH) + self.bodyH = self.height - self.scaleH + # if there is 1 line per row, draw them the standard way + for t, p in standardphases: + for i in sorted(self.rowheight[t][p]): + self.rowheight[t][p][i] = self.bodyH/len(self.rowlines[t][p]) + def createZoomBox(self, mode='command', testcount=1): + # Create bounding box, add buttons + html_zoombox = '<center><button id="zoomin">ZOOM IN +</button><button id="zoomout">ZOOM OUT -</button><button id="zoomdef">ZOOM 1:1</button></center>\n' + html_timeline = '<div id="dmesgzoombox" class="zoombox">\n<div id="{0}" class="timeline" style="height:{1}px">\n' + html_devlist1 = '<button id="devlist1" class="devlist" style="float:left;">Device Detail{0}</button>' + html_devlist2 = '<button id="devlist2" class="devlist" style="float:right;">Device Detail2</button>\n' + if mode != 'command': + if testcount > 1: + self.html += html_devlist2 + self.html += html_devlist1.format('1') + else: + self.html += html_devlist1.format('') + self.html += html_zoombox + self.html += html_timeline.format('dmesg', self.height) + # Function: createTimeScale + # Description: + # Create the timescale for a timeline block + # Arguments: + # m0: start time (mode begin) + # mMax: end time (mode end) + # tTotal: total timeline time + # mode: suspend or resume + # Output: + # The html code needed to display the time scale + def createTimeScale(self, m0, mMax, tTotal, mode): + timescale = '<div class="t" style="right:{0}%">{1}</div>\n' + rline = '<div class="t" style="left:0;border-left:1px solid black;border-right:0;">{0}</div>\n' + output = '<div class="timescale">\n' + # set scale for timeline + mTotal = mMax - m0 + tS = 0.1 + if(tTotal <= 0): + return output+'</div>\n' + if(tTotal > 4): + tS = 1 + divTotal = int(mTotal/tS) + 1 + divEdge = (mTotal - tS*(divTotal-1))*100/mTotal + for i in range(divTotal): + htmlline = '' + if(mode == 'suspend'): + pos = '%0.3f' % (100 - ((float(i)*tS*100)/mTotal) - divEdge) + val = '%0.fms' % (float(i-divTotal+1)*tS*1000) + if(i == divTotal - 1): + val = mode + htmlline = timescale.format(pos, val) + else: + pos = '%0.3f' % (100 - ((float(i)*tS*100)/mTotal)) + val = '%0.fms' % (float(i)*tS*1000) + htmlline = timescale.format(pos, val) + if(i == 0): + htmlline = rline.format(mode) + output += htmlline + self.html += output+'</div>\n' + +# Class: TestProps +# Description: +# A list of values describing the properties of these test runs +class TestProps: + stamp = '' + S0i3 = False + fwdata = [] + ftrace_line_fmt_fg = \ + '^ *(?P<time>[0-9\.]*) *\| *(?P<cpu>[0-9]*)\)'+\ + ' *(?P<proc>.*)-(?P<pid>[0-9]*) *\|'+\ + '[ +!#\*@$]*(?P<dur>[0-9\.]*) .*\| (?P<msg>.*)' + ftrace_line_fmt_nop = \ + ' *(?P<proc>.*)-(?P<pid>[0-9]*) *\[(?P<cpu>[0-9]*)\] *'+\ + '(?P<flags>.{4}) *(?P<time>[0-9\.]*): *'+\ + '(?P<msg>.*)' + ftrace_line_fmt = ftrace_line_fmt_nop + cgformat = False + data = 0 + ktemp = dict() + def __init__(self): + self.ktemp = dict() + def setTracerType(self, tracer): + if(tracer == 'function_graph'): + self.cgformat = True + self.ftrace_line_fmt = self.ftrace_line_fmt_fg + elif(tracer == 'nop'): + self.ftrace_line_fmt = self.ftrace_line_fmt_nop + else: + doError('Invalid tracer format: [%s]' % tracer) + +# Class: TestRun +# Description: +# A container for a suspend/resume test run. This is necessary as +# there could be more than one, and they need to be separate. +class TestRun: + ftemp = dict() + ttemp = dict() + data = 0 + def __init__(self, dataobj): + self.data = dataobj + self.ftemp = dict() + self.ttemp = dict() + +class ProcessMonitor: + proclist = dict() + running = False + def procstat(self): + c = ['cat /proc/[1-9]*/stat 2>/dev/null'] + process = Popen(c, shell=True, stdout=PIPE) + running = dict() + for line in process.stdout: + data = line.split() + pid = data[0] + name = re.sub('[()]', '', data[1]) + user = int(data[13]) + kern = int(data[14]) + kjiff = ujiff = 0 + if pid not in self.proclist: + self.proclist[pid] = {'name' : name, 'user' : user, 'kern' : kern} + else: + val = self.proclist[pid] + ujiff = user - val['user'] + kjiff = kern - val['kern'] + val['user'] = user + val['kern'] = kern + if ujiff > 0 or kjiff > 0: + running[pid] = ujiff + kjiff + process.wait() + out = '' + for pid in running: + jiffies = running[pid] + val = self.proclist[pid] + if out: + out += ',' + out += '%s-%s %d' % (val['name'], pid, jiffies) + return 'ps - '+out + def processMonitor(self, tid): + while self.running: + out = self.procstat() + if out: + sysvals.fsetVal(out, 'trace_marker') + def start(self): + self.thread = Thread(target=self.processMonitor, args=(0,)) + self.running = True + self.thread.start() + def stop(self): + self.running = False + +# ----------------- FUNCTIONS -------------------- + +# Function: vprint +# Description: +# verbose print (prints only with -verbose option) +# Arguments: +# msg: the debug/log message to print +def vprint(msg): + sysvals.logmsg += msg+'\n' + if(sysvals.verbose): + print(msg) + +# Function: parseStamp +# Description: +# Pull in the stamp comment line from the data file(s), +# create the stamp, and add it to the global sysvals object +# Arguments: +# m: the valid re.match output for the stamp line +def parseStamp(line, data): + m = re.match(sysvals.stampfmt, line) + data.stamp = {'time': '', 'host': '', 'mode': ''} + dt = datetime(int(m.group('y'))+2000, int(m.group('m')), + int(m.group('d')), int(m.group('H')), int(m.group('M')), + int(m.group('S'))) + data.stamp['time'] = dt.strftime('%B %d %Y, %I:%M:%S %p') + data.stamp['host'] = m.group('host') + data.stamp['mode'] = m.group('mode') + data.stamp['kernel'] = m.group('kernel') + sysvals.hostname = data.stamp['host'] + sysvals.suspendmode = data.stamp['mode'] + if sysvals.suspendmode == 'command' and sysvals.ftracefile != '': + modes = ['on', 'freeze', 'standby', 'mem'] + out = Popen(['grep', 'suspend_enter', sysvals.ftracefile], + stderr=PIPE, stdout=PIPE).stdout.read() + m = re.match('.* suspend_enter\[(?P<mode>.*)\]', out) + if m and m.group('mode') in ['1', '2', '3']: + sysvals.suspendmode = modes[int(m.group('mode'))] + data.stamp['mode'] = sysvals.suspendmode + if not sysvals.stamp: + sysvals.stamp = data.stamp + +# Function: doesTraceLogHaveTraceEvents +# Description: +# Quickly determine if the ftrace log has some or all of the trace events +# required for primary parsing. Set the usetraceevents and/or +# usetraceeventsonly flags in the global sysvals object +def doesTraceLogHaveTraceEvents(): + # check for kprobes + sysvals.usekprobes = False + out = call('grep -q "_cal: (" '+sysvals.ftracefile, shell=True) + if(out == 0): + sysvals.usekprobes = True + # check for callgraph data on trace event blocks + out = call('grep -q "_cpu_down()" '+sysvals.ftracefile, shell=True) + if(out == 0): + sysvals.usekprobes = True + out = Popen(['head', '-1', sysvals.ftracefile], + stderr=PIPE, stdout=PIPE).stdout.read().replace('\n', '') + m = re.match(sysvals.stampfmt, out) + if m and m.group('mode') == 'command': + sysvals.usetraceeventsonly = True + sysvals.usetraceevents = True + return + # figure out what level of trace events are supported + sysvals.usetraceeventsonly = True + sysvals.usetraceevents = False + for e in sysvals.traceevents: + out = call('grep -q "'+e+': " '+sysvals.ftracefile, shell=True) + if(out != 0): + sysvals.usetraceeventsonly = False + if(e == 'suspend_resume' and out == 0): + sysvals.usetraceevents = True + # determine is this log is properly formatted + for e in ['SUSPEND START', 'RESUME COMPLETE']: + out = call('grep -q "'+e+'" '+sysvals.ftracefile, shell=True) + if(out != 0): + sysvals.usetracemarkers = False + +# Function: appendIncompleteTraceLog +# Description: +# [deprecated for kernel 3.15 or newer] +# Legacy support of ftrace outputs that lack the device_pm_callback +# and/or suspend_resume trace events. The primary data should be +# taken from dmesg, and this ftrace is used only for callgraph data +# or custom actions in the timeline. The data is appended to the Data +# objects provided. +# Arguments: +# testruns: the array of Data objects obtained from parseKernelLog +def appendIncompleteTraceLog(testruns): + # create TestRun vessels for ftrace parsing + testcnt = len(testruns) + testidx = 0 + testrun = [] + for data in testruns: + testrun.append(TestRun(data)) + + # extract the callgraph and traceevent data + vprint('Analyzing the ftrace data...') + tp = TestProps() + tf = open(sysvals.ftracefile, 'r') + data = 0 + for line in tf: + # remove any latent carriage returns + line = line.replace('\r\n', '') + # grab the time stamp + m = re.match(sysvals.stampfmt, line) + if(m): + tp.stamp = line + continue + # determine the trace data type (required for further parsing) + m = re.match(sysvals.tracertypefmt, line) + if(m): + tp.setTracerType(m.group('t')) + continue + # device properties line + if(re.match(sysvals.devpropfmt, line)): + devProps(line) + continue + # parse only valid lines, if this is not one move on + m = re.match(tp.ftrace_line_fmt, line) + if(not m): + continue + # gather the basic message data from the line + m_time = m.group('time') + m_pid = m.group('pid') + m_msg = m.group('msg') + if(tp.cgformat): + m_param3 = m.group('dur') + else: + m_param3 = 'traceevent' + if(m_time and m_pid and m_msg): + t = FTraceLine(m_time, m_msg, m_param3) + pid = int(m_pid) + else: + continue + # the line should be a call, return, or event + if(not t.fcall and not t.freturn and not t.fevent): + continue + # look for the suspend start marker + if(t.startMarker()): + data = testrun[testidx].data + parseStamp(tp.stamp, data) + data.setStart(t.time) + continue + if(not data): + continue + # find the end of resume + if(t.endMarker()): + data.setEnd(t.time) + testidx += 1 + if(testidx >= testcnt): + break + continue + # trace event processing + if(t.fevent): + # general trace events have two types, begin and end + if(re.match('(?P<name>.*) begin$', t.name)): + isbegin = True + elif(re.match('(?P<name>.*) end$', t.name)): + isbegin = False + else: + continue + m = re.match('(?P<name>.*)\[(?P<val>[0-9]*)\] .*', t.name) + if(m): + val = m.group('val') + if val == '0': + name = m.group('name') + else: + name = m.group('name')+'['+val+']' + else: + m = re.match('(?P<name>.*) .*', t.name) + name = m.group('name') + # special processing for trace events + if re.match('dpm_prepare\[.*', name): + continue + elif re.match('machine_suspend.*', name): + continue + elif re.match('suspend_enter\[.*', name): + if(not isbegin): + data.dmesg['suspend_prepare']['end'] = t.time + continue + elif re.match('dpm_suspend\[.*', name): + if(not isbegin): + data.dmesg['suspend']['end'] = t.time + continue + elif re.match('dpm_suspend_late\[.*', name): + if(isbegin): + data.dmesg['suspend_late']['start'] = t.time + else: + data.dmesg['suspend_late']['end'] = t.time + continue + elif re.match('dpm_suspend_noirq\[.*', name): + if(isbegin): + data.dmesg['suspend_noirq']['start'] = t.time + else: + data.dmesg['suspend_noirq']['end'] = t.time + continue + elif re.match('dpm_resume_noirq\[.*', name): + if(isbegin): + data.dmesg['resume_machine']['end'] = t.time + data.dmesg['resume_noirq']['start'] = t.time + else: + data.dmesg['resume_noirq']['end'] = t.time + continue + elif re.match('dpm_resume_early\[.*', name): + if(isbegin): + data.dmesg['resume_early']['start'] = t.time + else: + data.dmesg['resume_early']['end'] = t.time + continue + elif re.match('dpm_resume\[.*', name): + if(isbegin): + data.dmesg['resume']['start'] = t.time + else: + data.dmesg['resume']['end'] = t.time + continue + elif re.match('dpm_complete\[.*', name): + if(isbegin): + data.dmesg['resume_complete']['start'] = t.time + else: + data.dmesg['resume_complete']['end'] = t.time + continue + # skip trace events inside devices calls + if(not data.isTraceEventOutsideDeviceCalls(pid, t.time)): + continue + # global events (outside device calls) are simply graphed + if(isbegin): + # store each trace event in ttemp + if(name not in testrun[testidx].ttemp): + testrun[testidx].ttemp[name] = [] + testrun[testidx].ttemp[name].append(\ + {'begin': t.time, 'end': t.time}) + else: + # finish off matching trace event in ttemp + if(name in testrun[testidx].ttemp): + testrun[testidx].ttemp[name][-1]['end'] = t.time + # call/return processing + elif sysvals.usecallgraph: + # create a callgraph object for the data + if(pid not in testrun[testidx].ftemp): + testrun[testidx].ftemp[pid] = [] + testrun[testidx].ftemp[pid].append(FTraceCallGraph(pid)) + # when the call is finished, see which device matches it + cg = testrun[testidx].ftemp[pid][-1] + if(cg.addLine(t)): + testrun[testidx].ftemp[pid].append(FTraceCallGraph(pid)) + tf.close() + + for test in testrun: + # add the traceevent data to the device hierarchy + if(sysvals.usetraceevents): + for name in test.ttemp: + for event in test.ttemp[name]: + test.data.newActionGlobal(name, event['begin'], event['end']) + + # add the callgraph data to the device hierarchy + for pid in test.ftemp: + for cg in test.ftemp[pid]: + if len(cg.list) < 1 or cg.invalid: + continue + if(not cg.postProcess()): + id = 'task %s cpu %s' % (pid, m.group('cpu')) + vprint('Sanity check failed for '+\ + id+', ignoring this callback') + continue + callstart = cg.start + callend = cg.end + for p in test.data.phases: + if(test.data.dmesg[p]['start'] <= callstart and + callstart <= test.data.dmesg[p]['end']): + list = test.data.dmesg[p]['list'] + for devname in list: + dev = list[devname] + if(pid == dev['pid'] and + callstart <= dev['start'] and + callend >= dev['end']): + dev['ftrace'] = cg + break + + test.data.printDetails() + +# Function: parseTraceLog +# Description: +# Analyze an ftrace log output file generated from this app during +# the execution phase. Used when the ftrace log is the primary data source +# and includes the suspend_resume and device_pm_callback trace events +# The ftrace filename is taken from sysvals +# Output: +# An array of Data objects +def parseTraceLog(): + vprint('Analyzing the ftrace data...') + if(os.path.exists(sysvals.ftracefile) == False): + doError('%s does not exist' % sysvals.ftracefile) + + sysvals.setupAllKprobes() + tracewatch = [] + if sysvals.usekprobes: + tracewatch += ['sync_filesystems', 'freeze_processes', 'syscore_suspend', + 'syscore_resume', 'resume_console', 'thaw_processes', 'CPU_ON', 'CPU_OFF'] + + # extract the callgraph and traceevent data + tp = TestProps() + testruns = [] + testdata = [] + testrun = 0 + data = 0 + tf = open(sysvals.ftracefile, 'r') + phase = 'suspend_prepare' + for line in tf: + # remove any latent carriage returns + line = line.replace('\r\n', '') + # stamp line: each stamp means a new test run + m = re.match(sysvals.stampfmt, line) + if(m): + tp.stamp = line + continue + # firmware line: pull out any firmware data + m = re.match(sysvals.firmwarefmt, line) + if(m): + tp.fwdata.append((int(m.group('s')), int(m.group('r')))) + continue + # tracer type line: determine the trace data type + m = re.match(sysvals.tracertypefmt, line) + if(m): + tp.setTracerType(m.group('t')) + continue + # device properties line + if(re.match(sysvals.devpropfmt, line)): + devProps(line) + continue + # ignore all other commented lines + if line[0] == '#': + continue + # ftrace line: parse only valid lines + m = re.match(tp.ftrace_line_fmt, line) + if(not m): + continue + # gather the basic message data from the line + m_time = m.group('time') + m_proc = m.group('proc') + m_pid = m.group('pid') + m_msg = m.group('msg') + if(tp.cgformat): + m_param3 = m.group('dur') + else: + m_param3 = 'traceevent' + if(m_time and m_pid and m_msg): + t = FTraceLine(m_time, m_msg, m_param3) + pid = int(m_pid) + else: + continue + # the line should be a call, return, or event + if(not t.fcall and not t.freturn and not t.fevent): + continue + # find the start of suspend + if(t.startMarker()): + phase = 'suspend_prepare' + data = Data(len(testdata)) + testdata.append(data) + testrun = TestRun(data) + testruns.append(testrun) + parseStamp(tp.stamp, data) + data.setStart(t.time) + data.tKernSus = t.time + continue + if(not data): + continue + # process cpu exec line + if t.type == 'tracing_mark_write': + m = re.match(sysvals.procexecfmt, t.name) + if(m): + proclist = dict() + for ps in m.group('ps').split(','): + val = ps.split() + if not val: + continue + name = val[0].replace('--', '-') + proclist[name] = int(val[1]) + data.pstl[t.time] = proclist + continue + # find the end of resume + if(t.endMarker()): + data.setEnd(t.time) + if data.tKernRes == 0.0: + data.tKernRes = t.time + if data.dmesg['resume_complete']['end'] < 0: + data.dmesg['resume_complete']['end'] = t.time + if sysvals.suspendmode == 'mem' and len(tp.fwdata) > data.testnumber: + data.fwSuspend, data.fwResume = tp.fwdata[data.testnumber] + if(data.tSuspended != 0 and data.tResumed != 0 and \ + (data.fwSuspend > 0 or data.fwResume > 0)): + data.fwValid = True + if(not sysvals.usetracemarkers): + # no trace markers? then quit and be sure to finish recording + # the event we used to trigger resume end + if(len(testrun.ttemp['thaw_processes']) > 0): + # if an entry exists, assume this is its end + testrun.ttemp['thaw_processes'][-1]['end'] = t.time + break + continue + # trace event processing + if(t.fevent): + if(phase == 'post_resume'): + data.setEnd(t.time) + if(t.type == 'suspend_resume'): + # suspend_resume trace events have two types, begin and end + if(re.match('(?P<name>.*) begin$', t.name)): + isbegin = True + elif(re.match('(?P<name>.*) end$', t.name)): + isbegin = False + else: + continue + m = re.match('(?P<name>.*)\[(?P<val>[0-9]*)\] .*', t.name) + if(m): + val = m.group('val') + if val == '0': + name = m.group('name') + else: + name = m.group('name')+'['+val+']' + else: + m = re.match('(?P<name>.*) .*', t.name) + name = m.group('name') + # ignore these events + if(name.split('[')[0] in tracewatch): + continue + # -- phase changes -- + # start of kernel suspend + if(re.match('suspend_enter\[.*', t.name)): + if(isbegin): + data.dmesg[phase]['start'] = t.time + data.tKernSus = t.time + continue + # suspend_prepare start + elif(re.match('dpm_prepare\[.*', t.name)): + phase = 'suspend_prepare' + if(not isbegin): + data.dmesg[phase]['end'] = t.time + continue + # suspend start + elif(re.match('dpm_suspend\[.*', t.name)): + phase = 'suspend' + data.setPhase(phase, t.time, isbegin) + continue + # suspend_late start + elif(re.match('dpm_suspend_late\[.*', t.name)): + phase = 'suspend_late' + data.setPhase(phase, t.time, isbegin) + continue + # suspend_noirq start + elif(re.match('dpm_suspend_noirq\[.*', t.name)): + phase = 'suspend_noirq' + data.setPhase(phase, t.time, isbegin) + if(not isbegin): + phase = 'suspend_machine' + data.dmesg[phase]['start'] = t.time + continue + # suspend_machine/resume_machine + elif(re.match('machine_suspend\[.*', t.name)): + if(isbegin): + phase = 'suspend_machine' + data.dmesg[phase]['end'] = t.time + data.tSuspended = t.time + else: + if(sysvals.suspendmode in ['mem', 'disk'] and not tp.S0i3): + data.dmesg['suspend_machine']['end'] = t.time + data.tSuspended = t.time + phase = 'resume_machine' + data.dmesg[phase]['start'] = t.time + data.tResumed = t.time + data.tLow = data.tResumed - data.tSuspended + continue + # acpi_suspend + elif(re.match('acpi_suspend\[.*', t.name)): + # acpi_suspend[0] S0i3 + if(re.match('acpi_suspend\[0\] begin', t.name)): + if(sysvals.suspendmode == 'mem'): + tp.S0i3 = True + data.dmesg['suspend_machine']['end'] = t.time + data.tSuspended = t.time + continue + # resume_noirq start + elif(re.match('dpm_resume_noirq\[.*', t.name)): + phase = 'resume_noirq' + data.setPhase(phase, t.time, isbegin) + if(isbegin): + data.dmesg['resume_machine']['end'] = t.time + continue + # resume_early start + elif(re.match('dpm_resume_early\[.*', t.name)): + phase = 'resume_early' + data.setPhase(phase, t.time, isbegin) + continue + # resume start + elif(re.match('dpm_resume\[.*', t.name)): + phase = 'resume' + data.setPhase(phase, t.time, isbegin) + continue + # resume complete start + elif(re.match('dpm_complete\[.*', t.name)): + phase = 'resume_complete' + if(isbegin): + data.dmesg[phase]['start'] = t.time + continue + # skip trace events inside devices calls + if(not data.isTraceEventOutsideDeviceCalls(pid, t.time)): + continue + # global events (outside device calls) are graphed + if(name not in testrun.ttemp): + testrun.ttemp[name] = [] + if(isbegin): + # create a new list entry + testrun.ttemp[name].append(\ + {'begin': t.time, 'end': t.time, 'pid': pid}) + else: + if(len(testrun.ttemp[name]) > 0): + # if an entry exists, assume this is its end + testrun.ttemp[name][-1]['end'] = t.time + elif(phase == 'post_resume'): + # post resume events can just have ends + testrun.ttemp[name].append({ + 'begin': data.dmesg[phase]['start'], + 'end': t.time}) + # device callback start + elif(t.type == 'device_pm_callback_start'): + m = re.match('(?P<drv>.*) (?P<d>.*), parent: *(?P<p>.*), .*',\ + t.name); + if(not m): + continue + drv = m.group('drv') + n = m.group('d') + p = m.group('p') + if(n and p): + data.newAction(phase, n, pid, p, t.time, -1, drv) + if pid not in data.devpids: + data.devpids.append(pid) + # device callback finish + elif(t.type == 'device_pm_callback_end'): + m = re.match('(?P<drv>.*) (?P<d>.*), err.*', t.name); + if(not m): + continue + n = m.group('d') + list = data.dmesg[phase]['list'] + if(n in list): + dev = list[n] + dev['length'] = t.time - dev['start'] + dev['end'] = t.time + # kprobe event processing + elif(t.fkprobe): + kprobename = t.type + kprobedata = t.name + key = (kprobename, pid) + # displayname is generated from kprobe data + displayname = '' + if(t.fcall): + displayname = sysvals.kprobeDisplayName(kprobename, kprobedata) + if not displayname: + continue + if(key not in tp.ktemp): + tp.ktemp[key] = [] + tp.ktemp[key].append({ + 'pid': pid, + 'begin': t.time, + 'end': t.time, + 'name': displayname, + 'cdata': kprobedata, + 'proc': m_proc, + }) + elif(t.freturn): + if(key not in tp.ktemp) or len(tp.ktemp[key]) < 1: + continue + e = tp.ktemp[key][-1] + if e['begin'] < 0.0 or t.time - e['begin'] < 0.000001: + tp.ktemp[key].pop() + else: + e['end'] = t.time + e['rdata'] = kprobedata + # end of kernel resume + if(kprobename == 'pm_notifier_call_chain' or \ + kprobename == 'pm_restore_console'): + data.dmesg[phase]['end'] = t.time + data.tKernRes = t.time + + # callgraph processing + elif sysvals.usecallgraph: + # create a callgraph object for the data + key = (m_proc, pid) + if(key not in testrun.ftemp): + testrun.ftemp[key] = [] + testrun.ftemp[key].append(FTraceCallGraph(pid)) + # when the call is finished, see which device matches it + cg = testrun.ftemp[key][-1] + if(cg.addLine(t)): + testrun.ftemp[key].append(FTraceCallGraph(pid)) + tf.close() + + if sysvals.suspendmode == 'command': + for test in testruns: + for p in test.data.phases: + if p == 'suspend_prepare': + test.data.dmesg[p]['start'] = test.data.start + test.data.dmesg[p]['end'] = test.data.end + else: + test.data.dmesg[p]['start'] = test.data.end + test.data.dmesg[p]['end'] = test.data.end + test.data.tSuspended = test.data.end + test.data.tResumed = test.data.end + test.data.tLow = 0 + test.data.fwValid = False + + # dev source and procmon events can be unreadable with mixed phase height + if sysvals.usedevsrc or sysvals.useprocmon: + sysvals.mixedphaseheight = False + + for i in range(len(testruns)): + test = testruns[i] + data = test.data + # find the total time range for this test (begin, end) + tlb, tle = data.start, data.end + if i < len(testruns) - 1: + tle = testruns[i+1].data.start + # add the process usage data to the timeline + if sysvals.useprocmon: + data.createProcessUsageEvents() + # add the traceevent data to the device hierarchy + if(sysvals.usetraceevents): + # add actual trace funcs + for name in test.ttemp: + for event in test.ttemp[name]: + data.newActionGlobal(name, event['begin'], event['end'], event['pid']) + # add the kprobe based virtual tracefuncs as actual devices + for key in tp.ktemp: + name, pid = key + if name not in sysvals.tracefuncs: + continue + for e in tp.ktemp[key]: + kb, ke = e['begin'], e['end'] + if kb == ke or tlb > kb or tle <= kb: + continue + color = sysvals.kprobeColor(name) + data.newActionGlobal(e['name'], kb, ke, pid, color) + # add config base kprobes and dev kprobes + if sysvals.usedevsrc: + for key in tp.ktemp: + name, pid = key + if name in sysvals.tracefuncs or name not in sysvals.dev_tracefuncs: + continue + for e in tp.ktemp[key]: + kb, ke = e['begin'], e['end'] + if kb == ke or tlb > kb or tle <= kb: + continue + data.addDeviceFunctionCall(e['name'], name, e['proc'], pid, kb, + ke, e['cdata'], e['rdata']) + if sysvals.usecallgraph: + # add the callgraph data to the device hierarchy + sortlist = dict() + for key in test.ftemp: + proc, pid = key + for cg in test.ftemp[key]: + if len(cg.list) < 1 or cg.invalid: + continue + if(not cg.postProcess()): + id = 'task %s' % (pid) + vprint('Sanity check failed for '+\ + id+', ignoring this callback') + continue + # match cg data to devices + if sysvals.suspendmode == 'command' or not cg.deviceMatch(pid, data): + sortkey = '%f%f%d' % (cg.start, cg.end, pid) + sortlist[sortkey] = cg + # create blocks for orphan cg data + for sortkey in sorted(sortlist): + cg = sortlist[sortkey] + name = cg.name + if sysvals.isCallgraphFunc(name): + vprint('Callgraph found for task %d: %.3fms, %s' % (cg.pid, (cg.end - cg.start)*1000, name)) + cg.newActionFromFunction(data) + + if sysvals.suspendmode == 'command': + for data in testdata: + data.printDetails() + return testdata + + # fill in any missing phases + for data in testdata: + lp = data.phases[0] + for p in data.phases: + if(data.dmesg[p]['start'] < 0 and data.dmesg[p]['end'] < 0): + vprint('WARNING: phase "%s" is missing!' % p) + if(data.dmesg[p]['start'] < 0): + data.dmesg[p]['start'] = data.dmesg[lp]['end'] + if(p == 'resume_machine'): + data.tSuspended = data.dmesg[lp]['end'] + data.tResumed = data.dmesg[lp]['end'] + data.tLow = 0 + if(data.dmesg[p]['end'] < 0): + data.dmesg[p]['end'] = data.dmesg[p]['start'] + if(p != lp and not ('machine' in p and 'machine' in lp)): + data.dmesg[lp]['end'] = data.dmesg[p]['start'] + lp = p + + if(len(sysvals.devicefilter) > 0): + data.deviceFilter(sysvals.devicefilter) + data.fixupInitcallsThatDidntReturn() + if sysvals.usedevsrc: + data.optimizeDevSrc() + data.printDetails() + + # x2: merge any overlapping devices between test runs + if sysvals.usedevsrc and len(testdata) > 1: + tc = len(testdata) + for i in range(tc - 1): + devlist = testdata[i].overflowDevices() + for j in range(i + 1, tc): + testdata[j].mergeOverlapDevices(devlist) + testdata[0].stitchTouchingThreads(testdata[1:]) + return testdata + +# Function: loadKernelLog +# Description: +# [deprecated for kernel 3.15.0 or newer] +# load the dmesg file into memory and fix up any ordering issues +# The dmesg filename is taken from sysvals +# Output: +# An array of empty Data objects with only their dmesgtext attributes set +def loadKernelLog(justtext=False): + vprint('Analyzing the dmesg data...') + if(os.path.exists(sysvals.dmesgfile) == False): + doError('%s does not exist' % sysvals.dmesgfile) + + if justtext: + dmesgtext = [] + # there can be multiple test runs in a single file + tp = TestProps() + tp.stamp = datetime.now().strftime('# suspend-%m%d%y-%H%M%S localhost mem unknown') + testruns = [] + data = 0 + lf = open(sysvals.dmesgfile, 'r') + for line in lf: + line = line.replace('\r\n', '') + idx = line.find('[') + if idx > 1: + line = line[idx:] + m = re.match(sysvals.stampfmt, line) + if(m): + tp.stamp = line + continue + m = re.match(sysvals.firmwarefmt, line) + if(m): + tp.fwdata.append((int(m.group('s')), int(m.group('r')))) + continue + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) + if(not m): + continue + msg = m.group("msg") + if justtext: + dmesgtext.append(line) + continue + if(re.match('PM: Syncing filesystems.*', msg)): + if(data): + testruns.append(data) + data = Data(len(testruns)) + parseStamp(tp.stamp, data) + if len(tp.fwdata) > data.testnumber: + data.fwSuspend, data.fwResume = tp.fwdata[data.testnumber] + if(data.fwSuspend > 0 or data.fwResume > 0): + data.fwValid = True + if(not data): + continue + m = re.match('.* *(?P<k>[0-9]\.[0-9]{2}\.[0-9]-.*) .*', msg) + if(m): + sysvals.stamp['kernel'] = m.group('k') + m = re.match('PM: Preparing system for (?P<m>.*) sleep', msg) + if(m): + sysvals.stamp['mode'] = sysvals.suspendmode = m.group('m') + data.dmesgtext.append(line) + lf.close() + + if justtext: + return dmesgtext + if data: + testruns.append(data) + if len(testruns) < 1: + doError(' dmesg log has no suspend/resume data: %s' \ + % sysvals.dmesgfile) + + # fix lines with same timestamp/function with the call and return swapped + for data in testruns: + last = '' + for line in data.dmesgtext: + mc = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) calling '+\ + '(?P<f>.*)\+ @ .*, parent: .*', line) + mr = re.match('.*(\[ *)(?P<t>[0-9\.]*)(\]) call '+\ + '(?P<f>.*)\+ returned .* after (?P<dt>.*) usecs', last) + if(mc and mr and (mc.group('t') == mr.group('t')) and + (mc.group('f') == mr.group('f'))): + i = data.dmesgtext.index(last) + j = data.dmesgtext.index(line) + data.dmesgtext[i] = line + data.dmesgtext[j] = last + last = line + return testruns + +# Function: parseKernelLog +# Description: +# [deprecated for kernel 3.15.0 or newer] +# Analyse a dmesg log output file generated from this app during +# the execution phase. Create a set of device structures in memory +# for subsequent formatting in the html output file +# This call is only for legacy support on kernels where the ftrace +# data lacks the suspend_resume or device_pm_callbacks trace events. +# Arguments: +# data: an empty Data object (with dmesgtext) obtained from loadKernelLog +# Output: +# The filled Data object +def parseKernelLog(data): + phase = 'suspend_runtime' + + if(data.fwValid): + vprint('Firmware Suspend = %u ns, Firmware Resume = %u ns' % \ + (data.fwSuspend, data.fwResume)) + + # dmesg phase match table + dm = { + 'suspend_prepare': 'PM: Syncing filesystems.*', + 'suspend': 'PM: Entering [a-z]* sleep.*', + 'suspend_late': 'PM: suspend of devices complete after.*', + 'suspend_noirq': 'PM: late suspend of devices complete after.*', + 'suspend_machine': 'PM: noirq suspend of devices complete after.*', + 'resume_machine': 'ACPI: Low-level resume complete.*', + 'resume_noirq': 'ACPI: Waking up from system sleep state.*', + 'resume_early': 'PM: noirq resume of devices complete after.*', + 'resume': 'PM: early resume of devices complete after.*', + 'resume_complete': 'PM: resume of devices complete after.*', + 'post_resume': '.*Restarting tasks \.\.\..*', + } + if(sysvals.suspendmode == 'standby'): + dm['resume_machine'] = 'PM: Restoring platform NVS memory' + elif(sysvals.suspendmode == 'disk'): + dm['suspend_late'] = 'PM: freeze of devices complete after.*' + dm['suspend_noirq'] = 'PM: late freeze of devices complete after.*' + dm['suspend_machine'] = 'PM: noirq freeze of devices complete after.*' + dm['resume_machine'] = 'PM: Restoring platform NVS memory' + dm['resume_early'] = 'PM: noirq restore of devices complete after.*' + dm['resume'] = 'PM: early restore of devices complete after.*' + dm['resume_complete'] = 'PM: restore of devices complete after.*' + elif(sysvals.suspendmode == 'freeze'): + dm['resume_machine'] = 'ACPI: resume from mwait' + + # action table (expected events that occur and show up in dmesg) + at = { + 'sync_filesystems': { + 'smsg': 'PM: Syncing filesystems.*', + 'emsg': 'PM: Preparing system for mem sleep.*' }, + 'freeze_user_processes': { + 'smsg': 'Freezing user space processes .*', + 'emsg': 'Freezing remaining freezable tasks.*' }, + 'freeze_tasks': { + 'smsg': 'Freezing remaining freezable tasks.*', + 'emsg': 'PM: Entering (?P<mode>[a-z,A-Z]*) sleep.*' }, + 'ACPI prepare': { + 'smsg': 'ACPI: Preparing to enter system sleep state.*', + 'emsg': 'PM: Saving platform NVS memory.*' }, + 'PM vns': { + 'smsg': 'PM: Saving platform NVS memory.*', + 'emsg': 'Disabling non-boot CPUs .*' }, + } + + t0 = -1.0 + cpu_start = -1.0 + prevktime = -1.0 + actions = dict() + for line in data.dmesgtext: + # parse each dmesg line into the time and message + m = re.match('[ \t]*(\[ *)(?P<ktime>[0-9\.]*)(\]) (?P<msg>.*)', line) + if(m): + val = m.group('ktime') + try: + ktime = float(val) + except: + continue + msg = m.group('msg') + # initialize data start to first line time + if t0 < 0: + data.setStart(ktime) + t0 = ktime + else: + continue + + # hack for determining resume_machine end for freeze + if(not sysvals.usetraceevents and sysvals.suspendmode == 'freeze' \ + and phase == 'resume_machine' and \ + re.match('calling (?P<f>.*)\+ @ .*, parent: .*', msg)): + data.dmesg['resume_machine']['end'] = ktime + phase = 'resume_noirq' + data.dmesg[phase]['start'] = ktime + + # suspend start + if(re.match(dm['suspend_prepare'], msg)): + phase = 'suspend_prepare' + data.dmesg[phase]['start'] = ktime + data.setStart(ktime) + data.tKernSus = ktime + # suspend start + elif(re.match(dm['suspend'], msg)): + data.dmesg['suspend_prepare']['end'] = ktime + phase = 'suspend' + data.dmesg[phase]['start'] = ktime + # suspend_late start + elif(re.match(dm['suspend_late'], msg)): + data.dmesg['suspend']['end'] = ktime + phase = 'suspend_late' + data.dmesg[phase]['start'] = ktime + # suspend_noirq start + elif(re.match(dm['suspend_noirq'], msg)): + data.dmesg['suspend_late']['end'] = ktime + phase = 'suspend_noirq' + data.dmesg[phase]['start'] = ktime + # suspend_machine start + elif(re.match(dm['suspend_machine'], msg)): + data.dmesg['suspend_noirq']['end'] = ktime + phase = 'suspend_machine' + data.dmesg[phase]['start'] = ktime + # resume_machine start + elif(re.match(dm['resume_machine'], msg)): + if(sysvals.suspendmode in ['freeze', 'standby']): + data.tSuspended = prevktime + data.dmesg['suspend_machine']['end'] = prevktime + else: + data.tSuspended = ktime + data.dmesg['suspend_machine']['end'] = ktime + phase = 'resume_machine' + data.tResumed = ktime + data.tLow = data.tResumed - data.tSuspended + data.dmesg[phase]['start'] = ktime + # resume_noirq start + elif(re.match(dm['resume_noirq'], msg)): + data.dmesg['resume_machine']['end'] = ktime + phase = 'resume_noirq' + data.dmesg[phase]['start'] = ktime + # resume_early start + elif(re.match(dm['resume_early'], msg)): + data.dmesg['resume_noirq']['end'] = ktime + phase = 'resume_early' + data.dmesg[phase]['start'] = ktime + # resume start + elif(re.match(dm['resume'], msg)): + data.dmesg['resume_early']['end'] = ktime + phase = 'resume' + data.dmesg[phase]['start'] = ktime + # resume complete start + elif(re.match(dm['resume_complete'], msg)): + data.dmesg['resume']['end'] = ktime + phase = 'resume_complete' + data.dmesg[phase]['start'] = ktime + # post resume start + elif(re.match(dm['post_resume'], msg)): + data.dmesg['resume_complete']['end'] = ktime + data.setEnd(ktime) + data.tKernRes = ktime + break + + # -- device callbacks -- + if(phase in data.phases): + # device init call + if(re.match('calling (?P<f>.*)\+ @ .*, parent: .*', msg)): + sm = re.match('calling (?P<f>.*)\+ @ '+\ + '(?P<n>.*), parent: (?P<p>.*)', msg); + f = sm.group('f') + n = sm.group('n') + p = sm.group('p') + if(f and n and p): + data.newAction(phase, f, int(n), p, ktime, -1, '') + # device init return + elif(re.match('call (?P<f>.*)\+ returned .* after '+\ + '(?P<t>.*) usecs', msg)): + sm = re.match('call (?P<f>.*)\+ returned .* after '+\ + '(?P<t>.*) usecs(?P<a>.*)', msg); + f = sm.group('f') + t = sm.group('t') + list = data.dmesg[phase]['list'] + if(f in list): + dev = list[f] + dev['length'] = int(t) + dev['end'] = ktime + + # if trace events are not available, these are better than nothing + if(not sysvals.usetraceevents): + # look for known actions + for a in at: + if(re.match(at[a]['smsg'], msg)): + if(a not in actions): + actions[a] = [] + actions[a].append({'begin': ktime, 'end': ktime}) + if(re.match(at[a]['emsg'], msg)): + if(a in actions): + actions[a][-1]['end'] = ktime + # now look for CPU on/off events + if(re.match('Disabling non-boot CPUs .*', msg)): + # start of first cpu suspend + cpu_start = ktime + elif(re.match('Enabling non-boot CPUs .*', msg)): + # start of first cpu resume + cpu_start = ktime + elif(re.match('smpboot: CPU (?P<cpu>[0-9]*) is now offline', msg)): + # end of a cpu suspend, start of the next + m = re.match('smpboot: CPU (?P<cpu>[0-9]*) is now offline', msg) + cpu = 'CPU'+m.group('cpu') + if(cpu not in actions): + actions[cpu] = [] + actions[cpu].append({'begin': cpu_start, 'end': ktime}) + cpu_start = ktime + elif(re.match('CPU(?P<cpu>[0-9]*) is up', msg)): + # end of a cpu resume, start of the next + m = re.match('CPU(?P<cpu>[0-9]*) is up', msg) + cpu = 'CPU'+m.group('cpu') + if(cpu not in actions): + actions[cpu] = [] + actions[cpu].append({'begin': cpu_start, 'end': ktime}) + cpu_start = ktime + prevktime = ktime + + # fill in any missing phases + lp = data.phases[0] + for p in data.phases: + if(data.dmesg[p]['start'] < 0 and data.dmesg[p]['end'] < 0): + print('WARNING: phase "%s" is missing, something went wrong!' % p) + print(' In %s, this dmesg line denotes the start of %s:' % \ + (sysvals.suspendmode, p)) + print(' "%s"' % dm[p]) + if(data.dmesg[p]['start'] < 0): + data.dmesg[p]['start'] = data.dmesg[lp]['end'] + if(p == 'resume_machine'): + data.tSuspended = data.dmesg[lp]['end'] + data.tResumed = data.dmesg[lp]['end'] + data.tLow = 0 + if(data.dmesg[p]['end'] < 0): + data.dmesg[p]['end'] = data.dmesg[p]['start'] + lp = p + + # fill in any actions we've found + for name in actions: + for event in actions[name]: + data.newActionGlobal(name, event['begin'], event['end']) + + data.printDetails() + if(len(sysvals.devicefilter) > 0): + data.deviceFilter(sysvals.devicefilter) + data.fixupInitcallsThatDidntReturn() + return True + +def callgraphHTML(sv, hf, num, cg, title, color, devid): + html_func_top = '<article id="{0}" class="atop" style="background:{1}">\n<input type="checkbox" class="pf" id="f{2}" checked/><label for="f{2}">{3} {4}</label>\n' + html_func_start = '<article>\n<input type="checkbox" class="pf" id="f{0}" checked/><label for="f{0}">{1} {2}</label>\n' + html_func_end = '</article>\n' + html_func_leaf = '<article>{0} {1}</article>\n' + + cgid = devid + if cg.id: + cgid += cg.id + cglen = (cg.end - cg.start) * 1000 + if cglen < sv.mincglen: + return num + + fmt = '<r>(%.3f ms @ '+sv.timeformat+' to '+sv.timeformat+')</r>' + flen = fmt % (cglen, cg.start, cg.end) + hf.write(html_func_top.format(cgid, color, num, title, flen)) + num += 1 + for line in cg.list: + if(line.length < 0.000000001): + flen = '' + else: + fmt = '<n>(%.3f ms @ '+sv.timeformat+')</n>' + flen = fmt % (line.length*1000, line.time) + if(line.freturn and line.fcall): + hf.write(html_func_leaf.format(line.name, flen)) + elif(line.freturn): + hf.write(html_func_end) + else: + hf.write(html_func_start.format(num, line.name, flen)) + num += 1 + hf.write(html_func_end) + return num + +def addCallgraphs(sv, hf, data): + hf.write('<section id="callgraphs" class="callgraph">\n') + # write out the ftrace data converted to html + num = 0 + for p in data.phases: + if sv.cgphase and p != sv.cgphase: + continue + list = data.dmesg[p]['list'] + for devname in data.sortedDevices(p): + dev = list[devname] + color = 'white' + if 'color' in data.dmesg[p]: + color = data.dmesg[p]['color'] + if 'color' in dev: + color = dev['color'] + name = devname + if(devname in sv.devprops): + name = sv.devprops[devname].altName(devname) + if sv.suspendmode in suspendmodename: + name += ' '+p + if('ftrace' in dev): + cg = dev['ftrace'] + num = callgraphHTML(sv, hf, num, cg, + name, color, dev['id']) + if('ftraces' in dev): + for cg in dev['ftraces']: + num = callgraphHTML(sv, hf, num, cg, + name+' → '+cg.name, color, dev['id']) + + hf.write('\n\n </section>\n') + +# Function: createHTMLSummarySimple +# Description: +# Create summary html file for a series of tests +# Arguments: +# testruns: array of Data objects from parseTraceLog +def createHTMLSummarySimple(testruns, htmlfile, folder): + # write the html header first (html head, css code, up to body start) + html = '<!DOCTYPE html>\n<html>\n<head>\n\ + <meta http-equiv="content-type" content="text/html; charset=UTF-8">\n\ + <title>SleepGraph Summary</title>\n\ + <style type=\'text/css\'>\n\ + .stamp {width: 100%;text-align:center;background:#888;line-height:30px;color:white;font: 25px Arial;}\n\ + table {width:100%;border-collapse: collapse;}\n\ + .summary {border:1px solid;}\n\ + th {border: 1px solid black;background:#222;color:white;}\n\ + td {font: 16px "Times New Roman";text-align: center;}\n\ + tr.alt td {background:#ddd;}\n\ + tr.avg td {background:#aaa;}\n\ + </style>\n</head>\n<body>\n' + + # group test header + html += '<div class="stamp">%s (%d tests)</div>\n' % (folder, len(testruns)) + th = '\t<th>{0}</th>\n' + td = '\t<td>{0}</td>\n' + tdlink = '\t<td><a href="{0}">html</a></td>\n' + + # table header + html += '<table class="summary">\n<tr>\n' + th.format('#') +\ + th.format('Mode') + th.format('Host') + th.format('Kernel') +\ + th.format('Test Time') + th.format('Suspend') + th.format('Resume') +\ + th.format('Detail') + '</tr>\n' + + # test data, 1 row per test + avg = '<tr class="avg"><td></td><td></td><td></td><td></td>'+\ + '<td>Average of {0} {1} tests</td><td>{2}</td><td>{3}</td><td></td></tr>\n' + sTimeAvg = rTimeAvg = 0.0 + mode = '' + num = 0 + for data in sorted(testruns, key=lambda v:(v['mode'], v['host'], v['kernel'])): + if mode != data['mode']: + # test average line + if(num > 0): + sTimeAvg /= (num - 1) + rTimeAvg /= (num - 1) + html += avg.format('%d' % (num - 1), mode, + '%3.3f ms' % sTimeAvg, '%3.3f ms' % rTimeAvg) + sTimeAvg = rTimeAvg = 0.0 + mode = data['mode'] + num = 1 + # alternate row color + if num % 2 == 1: + html += '<tr class="alt">\n' + else: + html += '<tr>\n' + html += td.format("%d" % num) + num += 1 + # basic info + for item in ['mode', 'host', 'kernel', 'time']: + val = "unknown" + if(item in data): + val = data[item] + html += td.format(val) + # suspend time + sTime = float(data['suspend']) + sTimeAvg += sTime + html += td.format('%.3f ms' % sTime) + # resume time + rTime = float(data['resume']) + rTimeAvg += rTime + html += td.format('%.3f ms' % rTime) + # link to the output html + html += tdlink.format(data['url']) + '</tr>\n' + # last test average line + if(num > 0): + sTimeAvg /= (num - 1) + rTimeAvg /= (num - 1) + html += avg.format('%d' % (num - 1), mode, + '%3.3f ms' % sTimeAvg, '%3.3f ms' % rTimeAvg) + + # flush the data to file + hf = open(htmlfile, 'w') + hf.write(html+'</table>\n</body>\n</html>\n') + hf.close() + +def ordinal(value): + suffix = 'th' + if value < 10 or value > 19: + if value % 10 == 1: + suffix = 'st' + elif value % 10 == 2: + suffix = 'nd' + elif value % 10 == 3: + suffix = 'rd' + return '%d%s' % (value, suffix) + +# Function: createHTML +# Description: +# Create the output html file from the resident test data +# Arguments: +# testruns: array of Data objects from parseKernelLog or parseTraceLog +# Output: +# True if the html file was created, false if it failed +def createHTML(testruns): + if len(testruns) < 1: + print('ERROR: Not enough test data to build a timeline') + return + + kerror = False + for data in testruns: + if data.kerror: + kerror = True + data.normalizeTime(testruns[-1].tSuspended) + + # html function templates + html_error = '<div id="{1}" title="kernel error/warning" class="err" style="right:{0}%">ERROR→</div>\n' + html_traceevent = '<div title="{0}" class="traceevent{6}" style="left:{1}%;top:{2}px;height:{3}px;width:{4}%;line-height:{3}px;{7}">{5}</div>\n' + html_cpuexec = '<div class="jiffie" style="left:{0}%;top:{1}px;height:{2}px;width:{3}%;background:{4};"></div>\n' + html_legend = '<div id="p{3}" class="square" style="left:{0}%;background:{1}"> {2}</div>\n' + html_timetotal = '<table class="time1">\n<tr>'\ + '<td class="green" title="{3}">{2} Suspend Time: <b>{0} ms</b></td>'\ + '<td class="yellow" title="{4}">{2} Resume Time: <b>{1} ms</b></td>'\ + '</tr>\n</table>\n' + html_timetotal2 = '<table class="time1">\n<tr>'\ + '<td class="green" title="{4}">{3} Suspend Time: <b>{0} ms</b></td>'\ + '<td class="gray" title="time spent in low-power mode with clock running">'+sysvals.suspendmode+' time: <b>{1} ms</b></td>'\ + '<td class="yellow" title="{5}">{3} Resume Time: <b>{2} ms</b></td>'\ + '</tr>\n</table>\n' + html_timetotal3 = '<table class="time1">\n<tr>'\ + '<td class="green">Execution Time: <b>{0} ms</b></td>'\ + '<td class="yellow">Command: <b>{1}</b></td>'\ + '</tr>\n</table>\n' + html_timegroups = '<table class="time2">\n<tr>'\ + '<td class="green" title="time from kernel enter_state({5}) to firmware mode [kernel time only]">{4}Kernel Suspend: {0} ms</td>'\ + '<td class="purple">{4}Firmware Suspend: {1} ms</td>'\ + '<td class="purple">{4}Firmware Resume: {2} ms</td>'\ + '<td class="yellow" title="time from firmware mode to return from kernel enter_state({5}) [kernel time only]">{4}Kernel Resume: {3} ms</td>'\ + '</tr>\n</table>\n' + + # html format variables + scaleH = 20 + if kerror: + scaleH = 40 + + # device timeline + vprint('Creating Device Timeline...') + + devtl = Timeline(30, scaleH) + + # write the test title and general info header + devtl.createHeader(sysvals) + + # Generate the header for this timeline + for data in testruns: + tTotal = data.end - data.start + sktime = (data.dmesg['suspend_machine']['end'] - \ + data.tKernSus) * 1000 + rktime = (data.dmesg['resume_complete']['end'] - \ + data.dmesg['resume_machine']['start']) * 1000 + if(tTotal == 0): + print('ERROR: No timeline data') + sys.exit() + if(data.tLow > 0): + low_time = '%.0f'%(data.tLow*1000) + if sysvals.suspendmode == 'command': + run_time = '%.0f'%((data.end-data.start)*1000) + if sysvals.testcommand: + testdesc = sysvals.testcommand + else: + testdesc = 'unknown' + if(len(testruns) > 1): + testdesc = ordinal(data.testnumber+1)+' '+testdesc + thtml = html_timetotal3.format(run_time, testdesc) + devtl.html += thtml + elif data.fwValid: + suspend_time = '%.0f'%(sktime + (data.fwSuspend/1000000.0)) + resume_time = '%.0f'%(rktime + (data.fwResume/1000000.0)) + testdesc1 = 'Total' + testdesc2 = '' + stitle = 'time from kernel enter_state(%s) to low-power mode [kernel & firmware time]' % sysvals.suspendmode + rtitle = 'time from low-power mode to return from kernel enter_state(%s) [firmware & kernel time]' % sysvals.suspendmode + if(len(testruns) > 1): + testdesc1 = testdesc2 = ordinal(data.testnumber+1) + testdesc2 += ' ' + if(data.tLow == 0): + thtml = html_timetotal.format(suspend_time, \ + resume_time, testdesc1, stitle, rtitle) + else: + thtml = html_timetotal2.format(suspend_time, low_time, \ + resume_time, testdesc1, stitle, rtitle) + devtl.html += thtml + sftime = '%.3f'%(data.fwSuspend / 1000000.0) + rftime = '%.3f'%(data.fwResume / 1000000.0) + devtl.html += html_timegroups.format('%.3f'%sktime, \ + sftime, rftime, '%.3f'%rktime, testdesc2, sysvals.suspendmode) + else: + suspend_time = '%.3f' % sktime + resume_time = '%.3f' % rktime + testdesc = 'Kernel' + stitle = 'time from kernel enter_state(%s) to firmware mode [kernel time only]' % sysvals.suspendmode + rtitle = 'time from firmware mode to return from kernel enter_state(%s) [kernel time only]' % sysvals.suspendmode + if(len(testruns) > 1): + testdesc = ordinal(data.testnumber+1)+' '+testdesc + if(data.tLow == 0): + thtml = html_timetotal.format(suspend_time, \ + resume_time, testdesc, stitle, rtitle) + else: + thtml = html_timetotal2.format(suspend_time, low_time, \ + resume_time, testdesc, stitle, rtitle) + devtl.html += thtml + + # time scale for potentially multiple datasets + t0 = testruns[0].start + tMax = testruns[-1].end + tTotal = tMax - t0 + + # determine the maximum number of rows we need to draw + fulllist = [] + threadlist = [] + pscnt = 0 + devcnt = 0 + for data in testruns: + data.selectTimelineDevices('%f', tTotal, sysvals.mindevlen) + for group in data.devicegroups: + devlist = [] + for phase in group: + for devname in data.tdevlist[phase]: + d = DevItem(data.testnumber, phase, data.dmesg[phase]['list'][devname]) + devlist.append(d) + if d.isa('kth'): + threadlist.append(d) + else: + if d.isa('ps'): + pscnt += 1 + else: + devcnt += 1 + fulllist.append(d) + if sysvals.mixedphaseheight: + devtl.getPhaseRows(devlist) + if not sysvals.mixedphaseheight: + if len(threadlist) > 0 and len(fulllist) > 0: + if pscnt > 0 and devcnt > 0: + msg = 'user processes & device pm callbacks' + elif pscnt > 0: + msg = 'user processes' + else: + msg = 'device pm callbacks' + d = testruns[0].addHorizontalDivider(msg, testruns[-1].end) + fulllist.insert(0, d) + devtl.getPhaseRows(fulllist) + if len(threadlist) > 0: + d = testruns[0].addHorizontalDivider('asynchronous kernel threads', testruns[-1].end) + threadlist.insert(0, d) + devtl.getPhaseRows(threadlist, devtl.rows) + devtl.calcTotalRows() + + # draw the full timeline + devtl.createZoomBox(sysvals.suspendmode, len(testruns)) + phases = {'suspend':[],'resume':[]} + for phase in data.dmesg: + if 'resume' in phase: + phases['resume'].append(phase) + else: + phases['suspend'].append(phase) + + # draw each test run chronologically + for data in testruns: + # now draw the actual timeline blocks + for dir in phases: + # draw suspend and resume blocks separately + bname = '%s%d' % (dir[0], data.testnumber) + if dir == 'suspend': + m0 = data.start + mMax = data.tSuspended + left = '%f' % (((m0-t0)*100.0)/tTotal) + else: + m0 = data.tSuspended + mMax = data.end + # in an x2 run, remove any gap between blocks + if len(testruns) > 1 and data.testnumber == 0: + mMax = testruns[1].start + left = '%f' % ((((m0-t0)*100.0)+sysvals.srgap/2)/tTotal) + mTotal = mMax - m0 + # if a timeline block is 0 length, skip altogether + if mTotal == 0: + continue + width = '%f' % (((mTotal*100.0)-sysvals.srgap/2)/tTotal) + devtl.html += devtl.html_tblock.format(bname, left, width, devtl.scaleH) + for b in sorted(phases[dir]): + # draw the phase color background + phase = data.dmesg[b] + length = phase['end']-phase['start'] + left = '%f' % (((phase['start']-m0)*100.0)/mTotal) + width = '%f' % ((length*100.0)/mTotal) + devtl.html += devtl.html_phase.format(left, width, \ + '%.3f'%devtl.scaleH, '%.3f'%devtl.bodyH, \ + data.dmesg[b]['color'], '') + for e in data.errorinfo[dir]: + # draw red lines for any kernel errors found + t, err = e + right = '%f' % (((mMax-t)*100.0)/mTotal) + devtl.html += html_error.format(right, err) + for b in sorted(phases[dir]): + # draw the devices for this phase + phaselist = data.dmesg[b]['list'] + for d in data.tdevlist[b]: + name = d + drv = '' + dev = phaselist[d] + xtraclass = '' + xtrainfo = '' + xtrastyle = '' + if 'htmlclass' in dev: + xtraclass = dev['htmlclass'] + if 'color' in dev: + xtrastyle = 'background:%s;' % dev['color'] + if(d in sysvals.devprops): + name = sysvals.devprops[d].altName(d) + xtraclass = sysvals.devprops[d].xtraClass() + xtrainfo = sysvals.devprops[d].xtraInfo() + elif xtraclass == ' kth': + xtrainfo = ' kernel_thread' + if('drv' in dev and dev['drv']): + drv = ' {%s}' % dev['drv'] + rowheight = devtl.phaseRowHeight(data.testnumber, b, dev['row']) + rowtop = devtl.phaseRowTop(data.testnumber, b, dev['row']) + top = '%.3f' % (rowtop + devtl.scaleH) + left = '%f' % (((dev['start']-m0)*100)/mTotal) + width = '%f' % (((dev['end']-dev['start'])*100)/mTotal) + length = ' (%0.3f ms) ' % ((dev['end']-dev['start'])*1000) + title = name+drv+xtrainfo+length + if sysvals.suspendmode == 'command': + title += sysvals.testcommand + elif xtraclass == ' ps': + if 'suspend' in b: + title += 'pre_suspend_process' + else: + title += 'post_resume_process' + else: + title += b + devtl.html += devtl.html_device.format(dev['id'], \ + title, left, top, '%.3f'%rowheight, width, \ + d+drv, xtraclass, xtrastyle) + if('cpuexec' in dev): + for t in sorted(dev['cpuexec']): + start, end = t + j = float(dev['cpuexec'][t]) / 5 + if j > 1.0: + j = 1.0 + height = '%.3f' % (rowheight/3) + top = '%.3f' % (rowtop + devtl.scaleH + 2*rowheight/3) + left = '%f' % (((start-m0)*100)/mTotal) + width = '%f' % ((end-start)*100/mTotal) + color = 'rgba(255, 0, 0, %f)' % j + devtl.html += \ + html_cpuexec.format(left, top, height, width, color) + if('src' not in dev): + continue + # draw any trace events for this device + for e in dev['src']: + height = '%.3f' % devtl.rowH + top = '%.3f' % (rowtop + devtl.scaleH + (e.row*devtl.rowH)) + left = '%f' % (((e.time-m0)*100)/mTotal) + width = '%f' % (e.length*100/mTotal) + xtrastyle = '' + if e.color: + xtrastyle = 'background:%s;' % e.color + devtl.html += \ + html_traceevent.format(e.title(), \ + left, top, height, width, e.text(), '', xtrastyle) + # draw the time scale, try to make the number of labels readable + devtl.createTimeScale(m0, mMax, tTotal, dir) + devtl.html += '</div>\n' + + # timeline is finished + devtl.html += '</div>\n</div>\n' + + # draw a legend which describes the phases by color + if sysvals.suspendmode != 'command': + data = testruns[-1] + devtl.html += '<div class="legend">\n' + pdelta = 100.0/len(data.phases) + pmargin = pdelta / 4.0 + for phase in data.phases: + tmp = phase.split('_') + id = tmp[0][0] + if(len(tmp) > 1): + id += tmp[1][0] + order = '%.2f' % ((data.dmesg[phase]['order'] * pdelta) + pmargin) + name = string.replace(phase, '_', ' ') + devtl.html += html_legend.format(order, \ + data.dmesg[phase]['color'], name, id) + devtl.html += '</div>\n' + + hf = open(sysvals.htmlfile, 'w') + + # no header or css if its embedded + if(sysvals.embedded): + hf.write('pass True tSus %.3f tRes %.3f tLow %.3f fwvalid %s tSus %.3f tRes %.3f\n' % + (data.tSuspended-data.start, data.end-data.tSuspended, data.tLow, data.fwValid, \ + data.fwSuspend/1000000, data.fwResume/1000000)) + else: + addCSS(hf, sysvals, len(testruns), kerror) + + # write the device timeline + hf.write(devtl.html) + hf.write('<div id="devicedetailtitle"></div>\n') + hf.write('<div id="devicedetail" style="display:none;">\n') + # draw the colored boxes for the device detail section + for data in testruns: + hf.write('<div id="devicedetail%d">\n' % data.testnumber) + pscolor = 'linear-gradient(to top left, #ccc, #eee)' + hf.write(devtl.html_phaselet.format('pre_suspend_process', \ + '0', '0', pscolor)) + for b in data.phases: + phase = data.dmesg[b] + length = phase['end']-phase['start'] + left = '%.3f' % (((phase['start']-t0)*100.0)/tTotal) + width = '%.3f' % ((length*100.0)/tTotal) + hf.write(devtl.html_phaselet.format(b, left, width, \ + data.dmesg[b]['color'])) + hf.write(devtl.html_phaselet.format('post_resume_process', \ + '0', '0', pscolor)) + if sysvals.suspendmode == 'command': + hf.write(devtl.html_phaselet.format('cmdexec', '0', '0', pscolor)) + hf.write('</div>\n') + hf.write('</div>\n') + + # write the ftrace data (callgraph) + if sysvals.cgtest >= 0 and len(testruns) > sysvals.cgtest: + data = testruns[sysvals.cgtest] + else: + data = testruns[-1] + if(sysvals.usecallgraph and not sysvals.embedded): + addCallgraphs(sysvals, hf, data) + + # add the test log as a hidden div + if sysvals.logmsg: + hf.write('<div id="testlog" style="display:none;">\n'+sysvals.logmsg+'</div>\n') + # add the dmesg log as a hidden div + if sysvals.addlogs and sysvals.dmesgfile: + hf.write('<div id="dmesglog" style="display:none;">\n') + lf = open(sysvals.dmesgfile, 'r') + for line in lf: + line = line.replace('<', '<').replace('>', '>') + hf.write(line) + lf.close() + hf.write('</div>\n') + # add the ftrace log as a hidden div + if sysvals.addlogs and sysvals.ftracefile: + hf.write('<div id="ftracelog" style="display:none;">\n') + lf = open(sysvals.ftracefile, 'r') + for line in lf: + hf.write(line) + lf.close() + hf.write('</div>\n') + + if(not sysvals.embedded): + # write the footer and close + addScriptCode(hf, testruns) + hf.write('</body>\n</html>\n') + else: + # embedded out will be loaded in a page, skip the js + t0 = (testruns[0].start - testruns[-1].tSuspended) * 1000 + tMax = (testruns[-1].end - testruns[-1].tSuspended) * 1000 + # add js code in a div entry for later evaluation + detail = 'var bounds = [%f,%f];\n' % (t0, tMax) + detail += 'var devtable = [\n' + for data in testruns: + topo = data.deviceTopology() + detail += '\t"%s",\n' % (topo) + detail += '];\n' + hf.write('<div id=customcode style=display:none>\n'+detail+'</div>\n') + hf.close() + return True + +def addCSS(hf, sv, testcount=1, kerror=False, extra=''): + kernel = sv.stamp['kernel'] + host = sv.hostname[0].upper()+sv.hostname[1:] + mode = sv.suspendmode + if sv.suspendmode in suspendmodename: + mode = suspendmodename[sv.suspendmode] + title = host+' '+mode+' '+kernel + + # various format changes by flags + cgchk = 'checked' + cgnchk = 'not(:checked)' + if sv.cgexp: + cgchk = 'not(:checked)' + cgnchk = 'checked' + + hoverZ = 'z-index:8;' + if sv.usedevsrc: + hoverZ = '' + + devlistpos = 'absolute' + if testcount > 1: + devlistpos = 'relative' + + scaleTH = 20 + if kerror: + scaleTH = 60 + + # write the html header first (html head, css code, up to body start) + html_header = '<!DOCTYPE html>\n<html>\n<head>\n\ + <meta http-equiv="content-type" content="text/html; charset=UTF-8">\n\ + <title>'+title+'</title>\n\ + <style type=\'text/css\'>\n\ + body {overflow-y:scroll;}\n\ + .stamp {width:100%;text-align:center;background:gray;line-height:30px;color:white;font:25px Arial;}\n\ + .callgraph {margin-top:30px;box-shadow:5px 5px 20px black;}\n\ + .callgraph article * {padding-left:28px;}\n\ + h1 {color:black;font:bold 30px Times;}\n\ + t0 {color:black;font:bold 30px Times;}\n\ + t1 {color:black;font:30px Times;}\n\ + t2 {color:black;font:25px Times;}\n\ + t3 {color:black;font:20px Times;white-space:nowrap;}\n\ + t4 {color:black;font:bold 30px Times;line-height:60px;white-space:nowrap;}\n\ + cS {font:bold 13px Times;}\n\ + table {width:100%;}\n\ + .gray {background:rgba(80,80,80,0.1);}\n\ + .green {background:rgba(204,255,204,0.4);}\n\ + .purple {background:rgba(128,0,128,0.2);}\n\ + .yellow {background:rgba(255,255,204,0.4);}\n\ + .blue {background:rgba(169,208,245,0.4);}\n\ + .time1 {font:22px Arial;border:1px solid;}\n\ + .time2 {font:15px Arial;border-bottom:1px solid;border-left:1px solid;border-right:1px solid;}\n\ + td {text-align:center;}\n\ + r {color:#500000;font:15px Tahoma;}\n\ + n {color:#505050;font:15px Tahoma;}\n\ + .tdhl {color:red;}\n\ + .hide {display:none;}\n\ + .pf {display:none;}\n\ + .pf:'+cgchk+' + label {background:url(\'data:image/svg+xml;utf,<?xml version="1.0" standalone="no"?><svg xmlns="http://www.w3.org/2000/svg" height="18" width="18" version="1.1"><circle cx="9" cy="9" r="8" stroke="black" stroke-width="1" fill="white"/><rect x="4" y="8" width="10" height="2" style="fill:black;stroke-width:0"/><rect x="8" y="4" width="2" height="10" style="fill:black;stroke-width:0"/></svg>\') no-repeat left center;}\n\ + .pf:'+cgnchk+' ~ label {background:url(\'data:image/svg+xml;utf,<?xml version="1.0" standalone="no"?><svg xmlns="http://www.w3.org/2000/svg" height="18" width="18" version="1.1"><circle cx="9" cy="9" r="8" stroke="black" stroke-width="1" fill="white"/><rect x="4" y="8" width="10" height="2" style="fill:black;stroke-width:0"/></svg>\') no-repeat left center;}\n\ + .pf:'+cgchk+' ~ *:not(:nth-child(2)) {display:none;}\n\ + .zoombox {position:relative;width:100%;overflow-x:scroll;-webkit-user-select:none;-moz-user-select:none;user-select:none;}\n\ + .timeline {position:relative;font-size:14px;cursor:pointer;width:100%; overflow:hidden;background:linear-gradient(#cccccc, white);}\n\ + .thread {position:absolute;height:0%;overflow:hidden;z-index:7;line-height:30px;font-size:14px;border:1px solid;text-align:center;white-space:nowrap;}\n\ + .thread.ps {border-radius:3px;background:linear-gradient(to top, #ccc, #eee);}\n\ + .thread:hover {background:white;border:1px solid red;'+hoverZ+'}\n\ + .thread.sec,.thread.sec:hover {background:black;border:0;color:white;line-height:15px;font-size:10px;}\n\ + .hover {background:white;border:1px solid red;'+hoverZ+'}\n\ + .hover.sync {background:white;}\n\ + .hover.bg,.hover.kth,.hover.sync,.hover.ps {background:white;}\n\ + .jiffie {position:absolute;pointer-events: none;z-index:8;}\n\ + .traceevent {position:absolute;font-size:10px;z-index:7;overflow:hidden;color:black;text-align:center;white-space:nowrap;border-radius:5px;border:1px solid black;background:linear-gradient(to bottom right,#CCC,#969696);}\n\ + .traceevent:hover {color:white;font-weight:bold;border:1px solid white;}\n\ + .phase {position:absolute;overflow:hidden;border:0px;text-align:center;}\n\ + .phaselet {float:left;overflow:hidden;border:0px;text-align:center;min-height:100px;font-size:24px;}\n\ + .t {position:absolute;line-height:'+('%d'%scaleTH)+'px;pointer-events:none;top:0;height:100%;border-right:1px solid black;z-index:6;}\n\ + .err {position:absolute;top:0%;height:100%;border-right:3px solid red;color:red;font:bold 14px Times;line-height:18px;}\n\ + .legend {position:relative; width:100%; height:40px; text-align:center;margin-bottom:20px}\n\ + .legend .square {position:absolute;cursor:pointer;top:10px; width:0px;height:20px;border:1px solid;padding-left:20px;}\n\ + button {height:40px;width:200px;margin-bottom:20px;margin-top:20px;font-size:24px;}\n\ + .logbtn {position:relative;float:right;height:25px;width:50px;margin-top:3px;margin-bottom:0;font-size:10px;text-align:center;}\n\ + .devlist {position:'+devlistpos+';width:190px;}\n\ + a:link {color:white;text-decoration:none;}\n\ + a:visited {color:white;}\n\ + a:hover {color:white;}\n\ + a:active {color:white;}\n\ + .version {position:relative;float:left;color:white;font-size:10px;line-height:30px;margin-left:10px;}\n\ + #devicedetail {min-height:100px;box-shadow:5px 5px 20px black;}\n\ + .tblock {position:absolute;height:100%;background:#ddd;}\n\ + .tback {position:absolute;width:100%;background:linear-gradient(#ccc, #ddd);}\n\ + .bg {z-index:1;}\n\ +'+extra+'\ + </style>\n</head>\n<body>\n' + hf.write(html_header) + +# Function: addScriptCode +# Description: +# Adds the javascript code to the output html +# Arguments: +# hf: the open html file pointer +# testruns: array of Data objects from parseKernelLog or parseTraceLog +def addScriptCode(hf, testruns): + t0 = testruns[0].start * 1000 + tMax = testruns[-1].end * 1000 + # create an array in javascript memory with the device details + detail = ' var devtable = [];\n' + for data in testruns: + topo = data.deviceTopology() + detail += ' devtable[%d] = "%s";\n' % (data.testnumber, topo) + detail += ' var bounds = [%f,%f];\n' % (t0, tMax) + # add the code which will manipulate the data in the browser + script_code = \ + '<script type="text/javascript">\n'+detail+\ + ' var resolution = -1;\n'\ + ' var dragval = [0, 0];\n'\ + ' function redrawTimescale(t0, tMax, tS) {\n'\ + ' var rline = \'<div class="t" style="left:0;border-left:1px solid black;border-right:0;">\';\n'\ + ' var tTotal = tMax - t0;\n'\ + ' var list = document.getElementsByClassName("tblock");\n'\ + ' for (var i = 0; i < list.length; i++) {\n'\ + ' var timescale = list[i].getElementsByClassName("timescale")[0];\n'\ + ' var m0 = t0 + (tTotal*parseFloat(list[i].style.left)/100);\n'\ + ' var mTotal = tTotal*parseFloat(list[i].style.width)/100;\n'\ + ' var mMax = m0 + mTotal;\n'\ + ' var html = "";\n'\ + ' var divTotal = Math.floor(mTotal/tS) + 1;\n'\ + ' if(divTotal > 1000) continue;\n'\ + ' var divEdge = (mTotal - tS*(divTotal-1))*100/mTotal;\n'\ + ' var pos = 0.0, val = 0.0;\n'\ + ' for (var j = 0; j < divTotal; j++) {\n'\ + ' var htmlline = "";\n'\ + ' var mode = list[i].id[5];\n'\ + ' if(mode == "s") {\n'\ + ' pos = 100 - (((j)*tS*100)/mTotal) - divEdge;\n'\ + ' val = (j-divTotal+1)*tS;\n'\ + ' if(j == divTotal - 1)\n'\ + ' htmlline = \'<div class="t" style="right:\'+pos+\'%"><cS>S→</cS></div>\';\n'\ + ' else\n'\ + ' htmlline = \'<div class="t" style="right:\'+pos+\'%">\'+val+\'ms</div>\';\n'\ + ' } else {\n'\ + ' pos = 100 - (((j)*tS*100)/mTotal);\n'\ + ' val = (j)*tS;\n'\ + ' htmlline = \'<div class="t" style="right:\'+pos+\'%">\'+val+\'ms</div>\';\n'\ + ' if(j == 0)\n'\ + ' if(mode == "r")\n'\ + ' htmlline = rline+"<cS>←R</cS></div>";\n'\ + ' else\n'\ + ' htmlline = rline+"<cS>0ms</div>";\n'\ + ' }\n'\ + ' html += htmlline;\n'\ + ' }\n'\ + ' timescale.innerHTML = html;\n'\ + ' }\n'\ + ' }\n'\ + ' function zoomTimeline() {\n'\ + ' var dmesg = document.getElementById("dmesg");\n'\ + ' var zoombox = document.getElementById("dmesgzoombox");\n'\ + ' var left = zoombox.scrollLeft;\n'\ + ' var val = parseFloat(dmesg.style.width);\n'\ + ' var newval = 100;\n'\ + ' var sh = window.outerWidth / 2;\n'\ + ' if(this.id == "zoomin") {\n'\ + ' newval = val * 1.2;\n'\ + ' if(newval > 910034) newval = 910034;\n'\ + ' dmesg.style.width = newval+"%";\n'\ + ' zoombox.scrollLeft = ((left + sh) * newval / val) - sh;\n'\ + ' } else if (this.id == "zoomout") {\n'\ + ' newval = val / 1.2;\n'\ + ' if(newval < 100) newval = 100;\n'\ + ' dmesg.style.width = newval+"%";\n'\ + ' zoombox.scrollLeft = ((left + sh) * newval / val) - sh;\n'\ + ' } else {\n'\ + ' zoombox.scrollLeft = 0;\n'\ + ' dmesg.style.width = "100%";\n'\ + ' }\n'\ + ' var tS = [10000, 5000, 2000, 1000, 500, 200, 100, 50, 20, 10, 5, 2, 1];\n'\ + ' var t0 = bounds[0];\n'\ + ' var tMax = bounds[1];\n'\ + ' var tTotal = tMax - t0;\n'\ + ' var wTotal = tTotal * 100.0 / newval;\n'\ + ' var idx = 7*window.innerWidth/1100;\n'\ + ' for(var i = 0; (i < tS.length)&&((wTotal / tS[i]) < idx); i++);\n'\ + ' if(i >= tS.length) i = tS.length - 1;\n'\ + ' if(tS[i] == resolution) return;\n'\ + ' resolution = tS[i];\n'\ + ' redrawTimescale(t0, tMax, tS[i]);\n'\ + ' }\n'\ + ' function deviceName(title) {\n'\ + ' var name = title.slice(0, title.indexOf(" ("));\n'\ + ' return name;\n'\ + ' }\n'\ + ' function deviceHover() {\n'\ + ' var name = deviceName(this.title);\n'\ + ' var dmesg = document.getElementById("dmesg");\n'\ + ' var dev = dmesg.getElementsByClassName("thread");\n'\ + ' var cpu = -1;\n'\ + ' if(name.match("CPU_ON\[[0-9]*\]"))\n'\ + ' cpu = parseInt(name.slice(7));\n'\ + ' else if(name.match("CPU_OFF\[[0-9]*\]"))\n'\ + ' cpu = parseInt(name.slice(8));\n'\ + ' for (var i = 0; i < dev.length; i++) {\n'\ + ' dname = deviceName(dev[i].title);\n'\ + ' var cname = dev[i].className.slice(dev[i].className.indexOf("thread"));\n'\ + ' if((cpu >= 0 && dname.match("CPU_O[NF]*\\\[*"+cpu+"\\\]")) ||\n'\ + ' (name == dname))\n'\ + ' {\n'\ + ' dev[i].className = "hover "+cname;\n'\ + ' } else {\n'\ + ' dev[i].className = cname;\n'\ + ' }\n'\ + ' }\n'\ + ' }\n'\ + ' function deviceUnhover() {\n'\ + ' var dmesg = document.getElementById("dmesg");\n'\ + ' var dev = dmesg.getElementsByClassName("thread");\n'\ + ' for (var i = 0; i < dev.length; i++) {\n'\ + ' dev[i].className = dev[i].className.slice(dev[i].className.indexOf("thread"));\n'\ + ' }\n'\ + ' }\n'\ + ' function deviceTitle(title, total, cpu) {\n'\ + ' var prefix = "Total";\n'\ + ' if(total.length > 3) {\n'\ + ' prefix = "Average";\n'\ + ' total[1] = (total[1]+total[3])/2;\n'\ + ' total[2] = (total[2]+total[4])/2;\n'\ + ' }\n'\ + ' var devtitle = document.getElementById("devicedetailtitle");\n'\ + ' var name = deviceName(title);\n'\ + ' if(cpu >= 0) name = "CPU"+cpu;\n'\ + ' var driver = "";\n'\ + ' var tS = "<t2>(</t2>";\n'\ + ' var tR = "<t2>)</t2>";\n'\ + ' if(total[1] > 0)\n'\ + ' tS = "<t2>("+prefix+" Suspend:</t2><t0> "+total[1].toFixed(3)+" ms</t0> ";\n'\ + ' if(total[2] > 0)\n'\ + ' tR = " <t2>"+prefix+" Resume:</t2><t0> "+total[2].toFixed(3)+" ms<t2>)</t2></t0>";\n'\ + ' var s = title.indexOf("{");\n'\ + ' var e = title.indexOf("}");\n'\ + ' if((s >= 0) && (e >= 0))\n'\ + ' driver = title.slice(s+1, e) + " <t1>@</t1> ";\n'\ + ' if(total[1] > 0 && total[2] > 0)\n'\ + ' devtitle.innerHTML = "<t0>"+driver+name+"</t0> "+tS+tR;\n'\ + ' else\n'\ + ' devtitle.innerHTML = "<t0>"+title+"</t0>";\n'\ + ' return name;\n'\ + ' }\n'\ + ' function deviceDetail() {\n'\ + ' var devinfo = document.getElementById("devicedetail");\n'\ + ' devinfo.style.display = "block";\n'\ + ' var name = deviceName(this.title);\n'\ + ' var cpu = -1;\n'\ + ' if(name.match("CPU_ON\[[0-9]*\]"))\n'\ + ' cpu = parseInt(name.slice(7));\n'\ + ' else if(name.match("CPU_OFF\[[0-9]*\]"))\n'\ + ' cpu = parseInt(name.slice(8));\n'\ + ' var dmesg = document.getElementById("dmesg");\n'\ + ' var dev = dmesg.getElementsByClassName("thread");\n'\ + ' var idlist = [];\n'\ + ' var pdata = [[]];\n'\ + ' if(document.getElementById("devicedetail1"))\n'\ + ' pdata = [[], []];\n'\ + ' var pd = pdata[0];\n'\ + ' var total = [0.0, 0.0, 0.0];\n'\ + ' for (var i = 0; i < dev.length; i++) {\n'\ + ' dname = deviceName(dev[i].title);\n'\ + ' if((cpu >= 0 && dname.match("CPU_O[NF]*\\\[*"+cpu+"\\\]")) ||\n'\ + ' (name == dname))\n'\ + ' {\n'\ + ' idlist[idlist.length] = dev[i].id;\n'\ + ' var tidx = 1;\n'\ + ' if(dev[i].id[0] == "a") {\n'\ + ' pd = pdata[0];\n'\ + ' } else {\n'\ + ' if(pdata.length == 1) pdata[1] = [];\n'\ + ' if(total.length == 3) total[3]=total[4]=0.0;\n'\ + ' pd = pdata[1];\n'\ + ' tidx = 3;\n'\ + ' }\n'\ + ' var info = dev[i].title.split(" ");\n'\ + ' var pname = info[info.length-1];\n'\ + ' pd[pname] = parseFloat(info[info.length-3].slice(1));\n'\ + ' total[0] += pd[pname];\n'\ + ' if(pname.indexOf("suspend") >= 0)\n'\ + ' total[tidx] += pd[pname];\n'\ + ' else\n'\ + ' total[tidx+1] += pd[pname];\n'\ + ' }\n'\ + ' }\n'\ + ' var devname = deviceTitle(this.title, total, cpu);\n'\ + ' var left = 0.0;\n'\ + ' for (var t = 0; t < pdata.length; t++) {\n'\ + ' pd = pdata[t];\n'\ + ' devinfo = document.getElementById("devicedetail"+t);\n'\ + ' var phases = devinfo.getElementsByClassName("phaselet");\n'\ + ' for (var i = 0; i < phases.length; i++) {\n'\ + ' if(phases[i].id in pd) {\n'\ + ' var w = 100.0*pd[phases[i].id]/total[0];\n'\ + ' var fs = 32;\n'\ + ' if(w < 8) fs = 4*w | 0;\n'\ + ' var fs2 = fs*3/4;\n'\ + ' phases[i].style.width = w+"%";\n'\ + ' phases[i].style.left = left+"%";\n'\ + ' phases[i].title = phases[i].id+" "+pd[phases[i].id]+" ms";\n'\ + ' left += w;\n'\ + ' var time = "<t4 style=\\"font-size:"+fs+"px\\">"+pd[phases[i].id]+" ms<br></t4>";\n'\ + ' var pname = "<t3 style=\\"font-size:"+fs2+"px\\">"+phases[i].id.replace(new RegExp("_", "g"), " ")+"</t3>";\n'\ + ' phases[i].innerHTML = time+pname;\n'\ + ' } else {\n'\ + ' phases[i].style.width = "0%";\n'\ + ' phases[i].style.left = left+"%";\n'\ + ' }\n'\ + ' }\n'\ + ' }\n'\ + ' if(typeof devstats !== \'undefined\')\n'\ + ' callDetail(this.id, this.title);\n'\ + ' var cglist = document.getElementById("callgraphs");\n'\ + ' if(!cglist) return;\n'\ + ' var cg = cglist.getElementsByClassName("atop");\n'\ + ' if(cg.length < 10) return;\n'\ + ' for (var i = 0; i < cg.length; i++) {\n'\ + ' cgid = cg[i].id.split("x")[0]\n'\ + ' if(idlist.indexOf(cgid) >= 0) {\n'\ + ' cg[i].style.display = "block";\n'\ + ' } else {\n'\ + ' cg[i].style.display = "none";\n'\ + ' }\n'\ + ' }\n'\ + ' }\n'\ + ' function callDetail(devid, devtitle) {\n'\ + ' if(!(devid in devstats) || devstats[devid].length < 1)\n'\ + ' return;\n'\ + ' var list = devstats[devid];\n'\ + ' var tmp = devtitle.split(" ");\n'\ + ' var name = tmp[0], phase = tmp[tmp.length-1];\n'\ + ' var dd = document.getElementById(phase);\n'\ + ' var total = parseFloat(tmp[1].slice(1));\n'\ + ' var mlist = [];\n'\ + ' var maxlen = 0;\n'\ + ' var info = []\n'\ + ' for(var i in list) {\n'\ + ' if(list[i][0] == "@") {\n'\ + ' info = list[i].split("|");\n'\ + ' continue;\n'\ + ' }\n'\ + ' var tmp = list[i].split("|");\n'\ + ' var t = parseFloat(tmp[0]), f = tmp[1], c = parseInt(tmp[2]);\n'\ + ' var p = (t*100.0/total).toFixed(2);\n'\ + ' mlist[mlist.length] = [f, c, t.toFixed(2), p+"%"];\n'\ + ' if(f.length > maxlen)\n'\ + ' maxlen = f.length;\n'\ + ' }\n'\ + ' var pad = 5;\n'\ + ' if(mlist.length == 0) pad = 30;\n'\ + ' var html = \'<div style="padding-top:\'+pad+\'px"><t3> <b>\'+name+\':</b>\';\n'\ + ' if(info.length > 2)\n'\ + ' html += " start=<b>"+info[1]+"</b>, end=<b>"+info[2]+"</b>";\n'\ + ' if(info.length > 3)\n'\ + ' html += ", length<i>(w/o overhead)</i>=<b>"+info[3]+" ms</b>";\n'\ + ' if(info.length > 4)\n'\ + ' html += ", return=<b>"+info[4]+"</b>";\n'\ + ' html += "</t3></div>";\n'\ + ' if(mlist.length > 0) {\n'\ + ' html += \'<table class=fstat style="padding-top:\'+(maxlen*5)+\'px;"><tr><th>Function</th>\';\n'\ + ' for(var i in mlist)\n'\ + ' html += "<td class=vt>"+mlist[i][0]+"</td>";\n'\ + ' html += "</tr><tr><th>Calls</th>";\n'\ + ' for(var i in mlist)\n'\ + ' html += "<td>"+mlist[i][1]+"</td>";\n'\ + ' html += "</tr><tr><th>Time(ms)</th>";\n'\ + ' for(var i in mlist)\n'\ + ' html += "<td>"+mlist[i][2]+"</td>";\n'\ + ' html += "</tr><tr><th>Percent</th>";\n'\ + ' for(var i in mlist)\n'\ + ' html += "<td>"+mlist[i][3]+"</td>";\n'\ + ' html += "</tr></table>";\n'\ + ' }\n'\ + ' dd.innerHTML = html;\n'\ + ' var height = (maxlen*5)+100;\n'\ + ' dd.style.height = height+"px";\n'\ + ' document.getElementById("devicedetail").style.height = height+"px";\n'\ + ' }\n'\ + ' function callSelect() {\n'\ + ' var cglist = document.getElementById("callgraphs");\n'\ + ' if(!cglist) return;\n'\ + ' var cg = cglist.getElementsByClassName("atop");\n'\ + ' for (var i = 0; i < cg.length; i++) {\n'\ + ' if(this.id == cg[i].id) {\n'\ + ' cg[i].style.display = "block";\n'\ + ' } else {\n'\ + ' cg[i].style.display = "none";\n'\ + ' }\n'\ + ' }\n'\ + ' }\n'\ + ' function devListWindow(e) {\n'\ + ' var win = window.open();\n'\ + ' var html = "<title>"+e.target.innerHTML+"</title>"+\n'\ + ' "<style type=\\"text/css\\">"+\n'\ + ' " ul {list-style-type:circle;padding-left:10px;margin-left:10px;}"+\n'\ + ' "</style>"\n'\ + ' var dt = devtable[0];\n'\ + ' if(e.target.id != "devlist1")\n'\ + ' dt = devtable[1];\n'\ + ' win.document.write(html+dt);\n'\ + ' }\n'\ + ' function errWindow() {\n'\ + ' var text = this.id;\n'\ + ' var win = window.open();\n'\ + ' win.document.write("<pre>"+text+"</pre>");\n'\ + ' win.document.close();\n'\ + ' }\n'\ + ' function logWindow(e) {\n'\ + ' var name = e.target.id.slice(4);\n'\ + ' var win = window.open();\n'\ + ' var log = document.getElementById(name+"log");\n'\ + ' var title = "<title>"+document.title.split(" ")[0]+" "+name+" log</title>";\n'\ + ' win.document.write(title+"<pre>"+log.innerHTML+"</pre>");\n'\ + ' win.document.close();\n'\ + ' }\n'\ + ' function onClickPhase(e) {\n'\ + ' }\n'\ + ' function onMouseDown(e) {\n'\ + ' dragval[0] = e.clientX;\n'\ + ' dragval[1] = document.getElementById("dmesgzoombox").scrollLeft;\n'\ + ' document.onmousemove = onMouseMove;\n'\ + ' }\n'\ + ' function onMouseMove(e) {\n'\ + ' var zoombox = document.getElementById("dmesgzoombox");\n'\ + ' zoombox.scrollLeft = dragval[1] + dragval[0] - e.clientX;\n'\ + ' }\n'\ + ' function onMouseUp(e) {\n'\ + ' document.onmousemove = null;\n'\ + ' }\n'\ + ' function onKeyPress(e) {\n'\ + ' var c = e.charCode;\n'\ + ' if(c != 42 && c != 43 && c != 45) return;\n'\ + ' var click = document.createEvent("Events");\n'\ + ' click.initEvent("click", true, false);\n'\ + ' if(c == 43) \n'\ + ' document.getElementById("zoomin").dispatchEvent(click);\n'\ + ' else if(c == 45)\n'\ + ' document.getElementById("zoomout").dispatchEvent(click);\n'\ + ' else if(c == 42)\n'\ + ' document.getElementById("zoomdef").dispatchEvent(click);\n'\ + ' }\n'\ + ' window.addEventListener("resize", function () {zoomTimeline();});\n'\ + ' window.addEventListener("load", function () {\n'\ + ' var dmesg = document.getElementById("dmesg");\n'\ + ' dmesg.style.width = "100%"\n'\ + ' dmesg.onmousedown = onMouseDown;\n'\ + ' document.onmouseup = onMouseUp;\n'\ + ' document.onkeypress = onKeyPress;\n'\ + ' document.getElementById("zoomin").onclick = zoomTimeline;\n'\ + ' document.getElementById("zoomout").onclick = zoomTimeline;\n'\ + ' document.getElementById("zoomdef").onclick = zoomTimeline;\n'\ + ' var list = document.getElementsByClassName("square");\n'\ + ' for (var i = 0; i < list.length; i++)\n'\ + ' list[i].onclick = onClickPhase;\n'\ + ' var list = document.getElementsByClassName("err");\n'\ + ' for (var i = 0; i < list.length; i++)\n'\ + ' list[i].onclick = errWindow;\n'\ + ' var list = document.getElementsByClassName("logbtn");\n'\ + ' for (var i = 0; i < list.length; i++)\n'\ + ' list[i].onclick = logWindow;\n'\ + ' list = document.getElementsByClassName("devlist");\n'\ + ' for (var i = 0; i < list.length; i++)\n'\ + ' list[i].onclick = devListWindow;\n'\ + ' var dev = dmesg.getElementsByClassName("thread");\n'\ + ' for (var i = 0; i < dev.length; i++) {\n'\ + ' dev[i].onclick = deviceDetail;\n'\ + ' dev[i].onmouseover = deviceHover;\n'\ + ' dev[i].onmouseout = deviceUnhover;\n'\ + ' }\n'\ + ' var dev = dmesg.getElementsByClassName("srccall");\n'\ + ' for (var i = 0; i < dev.length; i++)\n'\ + ' dev[i].onclick = callSelect;\n'\ + ' zoomTimeline();\n'\ + ' });\n'\ + '</script>\n' + hf.write(script_code); + +# Function: executeSuspend +# Description: +# Execute system suspend through the sysfs interface, then copy the output +# dmesg and ftrace files to the test output directory. +def executeSuspend(): + pm = ProcessMonitor() + tp = sysvals.tpath + fwdata = [] + # mark the start point in the kernel ring buffer just as we start + sysvals.initdmesg() + # start ftrace + if(sysvals.usecallgraph or sysvals.usetraceevents): + print('START TRACING') + sysvals.fsetVal('1', 'tracing_on') + if sysvals.useprocmon: + pm.start() + # execute however many s/r runs requested + for count in range(1,sysvals.execcount+1): + # x2delay in between test runs + if(count > 1 and sysvals.x2delay > 0): + sysvals.fsetVal('WAIT %d' % sysvals.x2delay, 'trace_marker') + time.sleep(sysvals.x2delay/1000.0) + sysvals.fsetVal('WAIT END', 'trace_marker') + # start message + if sysvals.testcommand != '': + print('COMMAND START') + else: + if(sysvals.rtcwake): + print('SUSPEND START') + else: + print('SUSPEND START (press a key to resume)') + # set rtcwake + if(sysvals.rtcwake): + print('will issue an rtcwake in %d seconds' % sysvals.rtcwaketime) + sysvals.rtcWakeAlarmOn() + # start of suspend trace marker + if(sysvals.usecallgraph or sysvals.usetraceevents): + sysvals.fsetVal('SUSPEND START', 'trace_marker') + # predelay delay + if(count == 1 and sysvals.predelay > 0): + sysvals.fsetVal('WAIT %d' % sysvals.predelay, 'trace_marker') + time.sleep(sysvals.predelay/1000.0) + sysvals.fsetVal('WAIT END', 'trace_marker') + # initiate suspend or command + if sysvals.testcommand != '': + call(sysvals.testcommand+' 2>&1', shell=True); + else: + pf = open(sysvals.powerfile, 'w') + pf.write(sysvals.suspendmode) + # execution will pause here + try: + pf.close() + except: + pass + if(sysvals.rtcwake): + sysvals.rtcWakeAlarmOff() + # postdelay delay + if(count == sysvals.execcount and sysvals.postdelay > 0): + sysvals.fsetVal('WAIT %d' % sysvals.postdelay, 'trace_marker') + time.sleep(sysvals.postdelay/1000.0) + sysvals.fsetVal('WAIT END', 'trace_marker') + # return from suspend + print('RESUME COMPLETE') + if(sysvals.usecallgraph or sysvals.usetraceevents): + sysvals.fsetVal('RESUME COMPLETE', 'trace_marker') + if(sysvals.suspendmode == 'mem' or sysvals.suspendmode == 'command'): + fwdata.append(getFPDT(False)) + # stop ftrace + if(sysvals.usecallgraph or sysvals.usetraceevents): + if sysvals.useprocmon: + pm.stop() + sysvals.fsetVal('0', 'tracing_on') + print('CAPTURING TRACE') + writeDatafileHeader(sysvals.ftracefile, fwdata) + call('cat '+tp+'trace >> '+sysvals.ftracefile, shell=True) + sysvals.fsetVal('', 'trace') + devProps() + # grab a copy of the dmesg output + print('CAPTURING DMESG') + writeDatafileHeader(sysvals.dmesgfile, fwdata) + sysvals.getdmesg() + +def writeDatafileHeader(filename, fwdata): + fp = open(filename, 'a') + fp.write(sysvals.teststamp+'\n') + if(sysvals.suspendmode == 'mem' or sysvals.suspendmode == 'command'): + for fw in fwdata: + if(fw): + fp.write('# fwsuspend %u fwresume %u\n' % (fw[0], fw[1])) + fp.close() + +# Function: setUSBDevicesAuto +# Description: +# Set the autosuspend control parameter of all USB devices to auto +# This can be dangerous, so use at your own risk, most devices are set +# to always-on since the kernel cant determine if the device can +# properly autosuspend +def setUSBDevicesAuto(): + rootCheck(True) + for dirname, dirnames, filenames in os.walk('/sys/devices'): + if(re.match('.*/usb[0-9]*.*', dirname) and + 'idVendor' in filenames and 'idProduct' in filenames): + call('echo auto > %s/power/control' % dirname, shell=True) + name = dirname.split('/')[-1] + desc = Popen(['cat', '%s/product' % dirname], + stderr=PIPE, stdout=PIPE).stdout.read().replace('\n', '') + ctrl = Popen(['cat', '%s/power/control' % dirname], + stderr=PIPE, stdout=PIPE).stdout.read().replace('\n', '') + print('control is %s for %6s: %s' % (ctrl, name, desc)) + +# Function: yesno +# Description: +# Print out an equivalent Y or N for a set of known parameter values +# Output: +# 'Y', 'N', or ' ' if the value is unknown +def yesno(val): + yesvals = ['auto', 'enabled', 'active', '1'] + novals = ['on', 'disabled', 'suspended', 'forbidden', 'unsupported'] + if val in yesvals: + return 'Y' + elif val in novals: + return 'N' + return ' ' + +# Function: ms2nice +# Description: +# Print out a very concise time string in minutes and seconds +# Output: +# The time string, e.g. "1901m16s" +def ms2nice(val): + ms = 0 + try: + ms = int(val) + except: + return 0.0 + m = ms / 60000 + s = (ms / 1000) - (m * 60) + return '%3dm%2ds' % (m, s) + +# Function: detectUSB +# Description: +# Detect all the USB hosts and devices currently connected and add +# a list of USB device names to sysvals for better timeline readability +def detectUSB(): + field = {'idVendor':'', 'idProduct':'', 'product':'', 'speed':''} + power = {'async':'', 'autosuspend':'', 'autosuspend_delay_ms':'', + 'control':'', 'persist':'', 'runtime_enabled':'', + 'runtime_status':'', 'runtime_usage':'', + 'runtime_active_time':'', + 'runtime_suspended_time':'', + 'active_duration':'', + 'connected_duration':''} + + print('LEGEND') + print('---------------------------------------------------------------------------------------------') + print(' A = async/sync PM queue Y/N D = autosuspend delay (seconds)') + print(' S = autosuspend Y/N rACTIVE = runtime active (min/sec)') + print(' P = persist across suspend Y/N rSUSPEN = runtime suspend (min/sec)') + print(' E = runtime suspend enabled/forbidden Y/N ACTIVE = active duration (min/sec)') + print(' R = runtime status active/suspended Y/N CONNECT = connected duration (min/sec)') + print(' U = runtime usage count') + print('---------------------------------------------------------------------------------------------') + print(' NAME ID DESCRIPTION SPEED A S P E R U D rACTIVE rSUSPEN ACTIVE CONNECT') + print('---------------------------------------------------------------------------------------------') + + for dirname, dirnames, filenames in os.walk('/sys/devices'): + if(re.match('.*/usb[0-9]*.*', dirname) and + 'idVendor' in filenames and 'idProduct' in filenames): + for i in field: + field[i] = Popen(['cat', '%s/%s' % (dirname, i)], + stderr=PIPE, stdout=PIPE).stdout.read().replace('\n', '') + name = dirname.split('/')[-1] + for i in power: + power[i] = Popen(['cat', '%s/power/%s' % (dirname, i)], + stderr=PIPE, stdout=PIPE).stdout.read().replace('\n', '') + if(re.match('usb[0-9]*', name)): + first = '%-8s' % name + else: + first = '%8s' % name + print('%s [%s:%s] %-20s %-4s %1s %1s %1s %1s %1s %1s %1s %s %s %s %s' % \ + (first, field['idVendor'], field['idProduct'], \ + field['product'][0:20], field['speed'], \ + yesno(power['async']), \ + yesno(power['control']), \ + yesno(power['persist']), \ + yesno(power['runtime_enabled']), \ + yesno(power['runtime_status']), \ + power['runtime_usage'], \ + power['autosuspend'], \ + ms2nice(power['runtime_active_time']), \ + ms2nice(power['runtime_suspended_time']), \ + ms2nice(power['active_duration']), \ + ms2nice(power['connected_duration']))) + +# Function: devProps +# Description: +# Retrieve a list of properties for all devices in the trace log +def devProps(data=0): + props = dict() + + if data: + idx = data.index(': ') + 2 + if idx >= len(data): + return + devlist = data[idx:].split(';') + for dev in devlist: + f = dev.split(',') + if len(f) < 3: + continue + dev = f[0] + props[dev] = DevProps() + props[dev].altname = f[1] + if int(f[2]): + props[dev].async = True + else: + props[dev].async = False + sysvals.devprops = props + if sysvals.suspendmode == 'command' and 'testcommandstring' in props: + sysvals.testcommand = props['testcommandstring'].altname + return + + if(os.path.exists(sysvals.ftracefile) == False): + doError('%s does not exist' % sysvals.ftracefile) + + # first get the list of devices we need properties for + msghead = 'Additional data added by AnalyzeSuspend' + alreadystamped = False + tp = TestProps() + tf = open(sysvals.ftracefile, 'r') + for line in tf: + if msghead in line: + alreadystamped = True + continue + # determine the trace data type (required for further parsing) + m = re.match(sysvals.tracertypefmt, line) + if(m): + tp.setTracerType(m.group('t')) + continue + # parse only valid lines, if this is not one move on + m = re.match(tp.ftrace_line_fmt, line) + if(not m or 'device_pm_callback_start' not in line): + continue + m = re.match('.*: (?P<drv>.*) (?P<d>.*), parent: *(?P<p>.*), .*', m.group('msg')); + if(not m): + continue + dev = m.group('d') + if dev not in props: + props[dev] = DevProps() + tf.close() + + if not alreadystamped and sysvals.suspendmode == 'command': + out = '#\n# '+msghead+'\n# Device Properties: ' + out += 'testcommandstring,%s,0;' % (sysvals.testcommand) + with open(sysvals.ftracefile, 'a') as fp: + fp.write(out+'\n') + sysvals.devprops = props + return + + # now get the syspath for each of our target devices + for dirname, dirnames, filenames in os.walk('/sys/devices'): + if(re.match('.*/power', dirname) and 'async' in filenames): + dev = dirname.split('/')[-2] + if dev in props and (not props[dev].syspath or len(dirname) < len(props[dev].syspath)): + props[dev].syspath = dirname[:-6] + + # now fill in the properties for our target devices + for dev in props: + dirname = props[dev].syspath + if not dirname or not os.path.exists(dirname): + continue + with open(dirname+'/power/async') as fp: + text = fp.read() + props[dev].async = False + if 'enabled' in text: + props[dev].async = True + fields = os.listdir(dirname) + if 'product' in fields: + with open(dirname+'/product') as fp: + props[dev].altname = fp.read() + elif 'name' in fields: + with open(dirname+'/name') as fp: + props[dev].altname = fp.read() + elif 'model' in fields: + with open(dirname+'/model') as fp: + props[dev].altname = fp.read() + elif 'description' in fields: + with open(dirname+'/description') as fp: + props[dev].altname = fp.read() + elif 'id' in fields: + with open(dirname+'/id') as fp: + props[dev].altname = fp.read() + elif 'idVendor' in fields and 'idProduct' in fields: + idv, idp = '', '' + with open(dirname+'/idVendor') as fp: + idv = fp.read().strip() + with open(dirname+'/idProduct') as fp: + idp = fp.read().strip() + props[dev].altname = '%s:%s' % (idv, idp) + + if props[dev].altname: + out = props[dev].altname.strip().replace('\n', ' ') + out = out.replace(',', ' ') + out = out.replace(';', ' ') + props[dev].altname = out + + # and now write the data to the ftrace file + if not alreadystamped: + out = '#\n# '+msghead+'\n# Device Properties: ' + for dev in sorted(props): + out += props[dev].out(dev) + with open(sysvals.ftracefile, 'a') as fp: + fp.write(out+'\n') + + sysvals.devprops = props + +# Function: getModes +# Description: +# Determine the supported power modes on this system +# Output: +# A string list of the available modes +def getModes(): + modes = '' + if(os.path.exists(sysvals.powerfile)): + fp = open(sysvals.powerfile, 'r') + modes = string.split(fp.read()) + fp.close() + return modes + +# Function: getFPDT +# Description: +# Read the acpi bios tables and pull out FPDT, the firmware data +# Arguments: +# output: True to output the info to stdout, False otherwise +def getFPDT(output): + rectype = {} + rectype[0] = 'Firmware Basic Boot Performance Record' + rectype[1] = 'S3 Performance Table Record' + prectype = {} + prectype[0] = 'Basic S3 Resume Performance Record' + prectype[1] = 'Basic S3 Suspend Performance Record' + + rootCheck(True) + if(not os.path.exists(sysvals.fpdtpath)): + if(output): + doError('file does not exist: %s' % sysvals.fpdtpath) + return False + if(not os.access(sysvals.fpdtpath, os.R_OK)): + if(output): + doError('file is not readable: %s' % sysvals.fpdtpath) + return False + if(not os.path.exists(sysvals.mempath)): + if(output): + doError('file does not exist: %s' % sysvals.mempath) + return False + if(not os.access(sysvals.mempath, os.R_OK)): + if(output): + doError('file is not readable: %s' % sysvals.mempath) + return False + + fp = open(sysvals.fpdtpath, 'rb') + buf = fp.read() + fp.close() + + if(len(buf) < 36): + if(output): + doError('Invalid FPDT table data, should '+\ + 'be at least 36 bytes') + return False + + table = struct.unpack('4sIBB6s8sI4sI', buf[0:36]) + if(output): + print('') + print('Firmware Performance Data Table (%s)' % table[0]) + print(' Signature : %s' % table[0]) + print(' Table Length : %u' % table[1]) + print(' Revision : %u' % table[2]) + print(' Checksum : 0x%x' % table[3]) + print(' OEM ID : %s' % table[4]) + print(' OEM Table ID : %s' % table[5]) + print(' OEM Revision : %u' % table[6]) + print(' Creator ID : %s' % table[7]) + print(' Creator Revision : 0x%x' % table[8]) + print('') + + if(table[0] != 'FPDT'): + if(output): + doError('Invalid FPDT table') + return False + if(len(buf) <= 36): + return False + i = 0 + fwData = [0, 0] + records = buf[36:] + fp = open(sysvals.mempath, 'rb') + while(i < len(records)): + header = struct.unpack('HBB', records[i:i+4]) + if(header[0] not in rectype): + i += header[1] + continue + if(header[1] != 16): + i += header[1] + continue + addr = struct.unpack('Q', records[i+8:i+16])[0] + try: + fp.seek(addr) + first = fp.read(8) + except: + if(output): + print('Bad address 0x%x in %s' % (addr, sysvals.mempath)) + return [0, 0] + rechead = struct.unpack('4sI', first) + recdata = fp.read(rechead[1]-8) + if(rechead[0] == 'FBPT'): + record = struct.unpack('HBBIQQQQQ', recdata) + if(output): + print('%s (%s)' % (rectype[header[0]], rechead[0])) + print(' Reset END : %u ns' % record[4]) + print(' OS Loader LoadImage Start : %u ns' % record[5]) + print(' OS Loader StartImage Start : %u ns' % record[6]) + print(' ExitBootServices Entry : %u ns' % record[7]) + print(' ExitBootServices Exit : %u ns' % record[8]) + elif(rechead[0] == 'S3PT'): + if(output): + print('%s (%s)' % (rectype[header[0]], rechead[0])) + j = 0 + while(j < len(recdata)): + prechead = struct.unpack('HBB', recdata[j:j+4]) + if(prechead[0] not in prectype): + continue + if(prechead[0] == 0): + record = struct.unpack('IIQQ', recdata[j:j+prechead[1]]) + fwData[1] = record[2] + if(output): + print(' %s' % prectype[prechead[0]]) + print(' Resume Count : %u' % \ + record[1]) + print(' FullResume : %u ns' % \ + record[2]) + print(' AverageResume : %u ns' % \ + record[3]) + elif(prechead[0] == 1): + record = struct.unpack('QQ', recdata[j+4:j+prechead[1]]) + fwData[0] = record[1] - record[0] + if(output): + print(' %s' % prectype[prechead[0]]) + print(' SuspendStart : %u ns' % \ + record[0]) + print(' SuspendEnd : %u ns' % \ + record[1]) + print(' SuspendTime : %u ns' % \ + fwData[0]) + j += prechead[1] + if(output): + print('') + i += header[1] + fp.close() + return fwData + +# Function: statusCheck +# Description: +# Verify that the requested command and options will work, and +# print the results to the terminal +# Output: +# True if the test will work, False if not +def statusCheck(probecheck=False): + status = True + + print('Checking this system (%s)...' % platform.node()) + + # check we have root access + res = sysvals.colorText('NO (No features of this tool will work!)') + if(rootCheck(False)): + res = 'YES' + print(' have root access: %s' % res) + if(res != 'YES'): + print(' Try running this script with sudo') + return False + + # check sysfs is mounted + res = sysvals.colorText('NO (No features of this tool will work!)') + if(os.path.exists(sysvals.powerfile)): + res = 'YES' + print(' is sysfs mounted: %s' % res) + if(res != 'YES'): + return False + + # check target mode is a valid mode + if sysvals.suspendmode != 'command': + res = sysvals.colorText('NO') + modes = getModes() + if(sysvals.suspendmode in modes): + res = 'YES' + else: + status = False + print(' is "%s" a valid power mode: %s' % (sysvals.suspendmode, res)) + if(res == 'NO'): + print(' valid power modes are: %s' % modes) + print(' please choose one with -m') + + # check if ftrace is available + res = sysvals.colorText('NO') + ftgood = sysvals.verifyFtrace() + if(ftgood): + res = 'YES' + elif(sysvals.usecallgraph): + status = False + print(' is ftrace supported: %s' % res) + + # check if kprobes are available + res = sysvals.colorText('NO') + sysvals.usekprobes = sysvals.verifyKprobes() + if(sysvals.usekprobes): + res = 'YES' + else: + sysvals.usedevsrc = False + print(' are kprobes supported: %s' % res) + + # what data source are we using + res = 'DMESG' + if(ftgood): + sysvals.usetraceeventsonly = True + sysvals.usetraceevents = False + for e in sysvals.traceevents: + check = False + if(os.path.exists(sysvals.epath+e)): + check = True + if(not check): + sysvals.usetraceeventsonly = False + if(e == 'suspend_resume' and check): + sysvals.usetraceevents = True + if(sysvals.usetraceevents and sysvals.usetraceeventsonly): + res = 'FTRACE (all trace events found)' + elif(sysvals.usetraceevents): + res = 'DMESG and FTRACE (suspend_resume trace event found)' + print(' timeline data source: %s' % res) + + # check if rtcwake + res = sysvals.colorText('NO') + if(sysvals.rtcpath != ''): + res = 'YES' + elif(sysvals.rtcwake): + status = False + print(' is rtcwake supported: %s' % res) + + if not probecheck: + return status + + # verify kprobes + if sysvals.usekprobes: + for name in sysvals.tracefuncs: + sysvals.defaultKprobe(name, sysvals.tracefuncs[name]) + if sysvals.usedevsrc: + for name in sysvals.dev_tracefuncs: + sysvals.defaultKprobe(name, sysvals.dev_tracefuncs[name]) + sysvals.addKprobes(True) + + return status + +# Function: doError +# Description: +# generic error function for catastrphic failures +# Arguments: +# msg: the error message to print +# help: True if printHelp should be called after, False otherwise +def doError(msg, help=False): + if(help == True): + printHelp() + print('ERROR: %s\n') % msg + sys.exit() + +# Function: rootCheck +# Description: +# quick check to see if we have root access +def rootCheck(fatal): + if(os.access(sysvals.powerfile, os.W_OK)): + return True + if fatal: + doError('This command requires sysfs mount and root access') + return False + +# Function: getArgInt +# Description: +# pull out an integer argument from the command line with checks +def getArgInt(name, args, min, max, main=True): + if main: + try: + arg = args.next() + except: + doError(name+': no argument supplied', True) + else: + arg = args + try: + val = int(arg) + except: + doError(name+': non-integer value given', True) + if(val < min or val > max): + doError(name+': value should be between %d and %d' % (min, max), True) + return val + +# Function: getArgFloat +# Description: +# pull out a float argument from the command line with checks +def getArgFloat(name, args, min, max, main=True): + if main: + try: + arg = args.next() + except: + doError(name+': no argument supplied', True) + else: + arg = args + try: + val = float(arg) + except: + doError(name+': non-numerical value given', True) + if(val < min or val > max): + doError(name+': value should be between %f and %f' % (min, max), True) + return val + +def processData(): + print('PROCESSING DATA') + if(sysvals.usetraceeventsonly): + testruns = parseTraceLog() + if sysvals.dmesgfile: + dmesgtext = loadKernelLog(True) + for data in testruns: + data.extractErrorInfo(dmesgtext) + else: + testruns = loadKernelLog() + for data in testruns: + parseKernelLog(data) + if(sysvals.ftracefile and (sysvals.usecallgraph or sysvals.usetraceevents)): + appendIncompleteTraceLog(testruns) + createHTML(testruns) + +# Function: rerunTest +# Description: +# generate an output from an existing set of ftrace/dmesg logs +def rerunTest(): + if sysvals.ftracefile: + doesTraceLogHaveTraceEvents() + if not sysvals.dmesgfile and not sysvals.usetraceeventsonly: + doError('recreating this html output requires a dmesg file') + sysvals.setOutputFile() + vprint('Output file: %s' % sysvals.htmlfile) + if(os.path.exists(sysvals.htmlfile) and not os.access(sysvals.htmlfile, os.W_OK)): + doError('missing permission to write to %s' % sysvals.htmlfile) + processData() + +# Function: runTest +# Description: +# execute a suspend/resume, gather the logs, and generate the output +def runTest(subdir, testpath=''): + # prepare for the test + sysvals.initFtrace() + sysvals.initTestOutput(subdir, testpath) + vprint('Output files:\n\t%s\n\t%s\n\t%s' % \ + (sysvals.dmesgfile, sysvals.ftracefile, sysvals.htmlfile)) + + # execute the test + executeSuspend() + sysvals.cleanupFtrace() + processData() + + # if running as root, change output dir owner to sudo_user + if os.path.isdir(sysvals.testdir) and os.getuid() == 0 and \ + 'SUDO_USER' in os.environ: + cmd = 'chown -R {0}:{0} {1} > /dev/null 2>&1' + call(cmd.format(os.environ['SUDO_USER'], sysvals.testdir), shell=True) + +def find_in_html(html, strs, div=False): + for str in strs: + l = len(str) + i = html.find(str) + if i >= 0: + break + if i < 0: + return '' + if not div: + return re.search(r'[-+]?\d*\.\d+|\d+', html[i+l:i+l+50]).group() + n = html[i+l:].find('</div>') + if n < 0: + return '' + return html[i+l:i+l+n] + +# Function: runSummary +# Description: +# create a summary of tests in a sub-directory +def runSummary(subdir, local=True): + inpath = os.path.abspath(subdir) + outpath = inpath + if local: + outpath = os.path.abspath('.') + print('Generating a summary of folder "%s"' % inpath) + testruns = [] + for dirname, dirnames, filenames in os.walk(subdir): + for filename in filenames: + if(not re.match('.*.html', filename)): + continue + file = os.path.join(dirname, filename) + html = open(file, 'r').read(10000) + suspend = find_in_html(html, + ['Kernel Suspend: ', 'Kernel Suspend Time: ']) + resume = find_in_html(html, + ['Kernel Resume: ', 'Kernel Resume Time: ']) + line = find_in_html(html, ['<div class="stamp">'], True) + stmp = line.split() + if not suspend or not resume or len(stmp) < 4: + continue + data = { + 'host': stmp[0], + 'kernel': stmp[1], + 'mode': stmp[2], + 'time': string.join(stmp[3:], ' '), + 'suspend': suspend, + 'resume': resume, + 'url': os.path.relpath(file, outpath), + } + if len(stmp) == 7: + data['kernel'] = 'unknown' + data['mode'] = stmp[1] + data['time'] = string.join(stmp[2:], ' ') + testruns.append(data) + outfile = os.path.join(outpath, 'summary.html') + print('Summary file: %s' % outfile) + createHTMLSummarySimple(testruns, outfile, inpath) + +# Function: checkArgBool +# Description: +# check if a boolean string value is true or false +def checkArgBool(value): + yes = ['1', 'true', 'yes', 'on'] + if value.lower() in yes: + return True + return False + +# Function: configFromFile +# Description: +# Configure the script via the info in a config file +def configFromFile(file): + Config = ConfigParser.ConfigParser() + + Config.read(file) + sections = Config.sections() + overridekprobes = False + overridedevkprobes = False + if 'Settings' in sections: + for opt in Config.options('Settings'): + value = Config.get('Settings', opt).lower() + if(opt.lower() == 'verbose'): + sysvals.verbose = checkArgBool(value) + elif(opt.lower() == 'addlogs'): + sysvals.addlogs = checkArgBool(value) + elif(opt.lower() == 'dev'): + sysvals.usedevsrc = checkArgBool(value) + elif(opt.lower() == 'proc'): + sysvals.useprocmon = checkArgBool(value) + elif(opt.lower() == 'x2'): + if checkArgBool(value): + sysvals.execcount = 2 + elif(opt.lower() == 'callgraph'): + sysvals.usecallgraph = checkArgBool(value) + elif(opt.lower() == 'override-timeline-functions'): + overridekprobes = checkArgBool(value) + elif(opt.lower() == 'override-dev-timeline-functions'): + overridedevkprobes = checkArgBool(value) + elif(opt.lower() == 'devicefilter'): + sysvals.setDeviceFilter(value) + elif(opt.lower() == 'expandcg'): + sysvals.cgexp = checkArgBool(value) + elif(opt.lower() == 'srgap'): + if checkArgBool(value): + sysvals.srgap = 5 + elif(opt.lower() == 'mode'): + sysvals.suspendmode = value + elif(opt.lower() == 'command'): + sysvals.testcommand = value + elif(opt.lower() == 'x2delay'): + sysvals.x2delay = getArgInt('-x2delay', value, 0, 60000, False) + elif(opt.lower() == 'predelay'): + sysvals.predelay = getArgInt('-predelay', value, 0, 60000, False) + elif(opt.lower() == 'postdelay'): + sysvals.postdelay = getArgInt('-postdelay', value, 0, 60000, False) + elif(opt.lower() == 'maxdepth'): + sysvals.max_graph_depth = getArgInt('-maxdepth', value, 0, 1000, False) + elif(opt.lower() == 'rtcwake'): + if value.lower() == 'off': + sysvals.rtcwake = False + else: + sysvals.rtcwake = True + sysvals.rtcwaketime = getArgInt('-rtcwake', value, 0, 3600, False) + elif(opt.lower() == 'timeprec'): + sysvals.setPrecision(getArgInt('-timeprec', value, 0, 6, False)) + elif(opt.lower() == 'mindev'): + sysvals.mindevlen = getArgFloat('-mindev', value, 0.0, 10000.0, False) + elif(opt.lower() == 'callloop-maxgap'): + sysvals.callloopmaxgap = getArgFloat('-callloop-maxgap', value, 0.0, 1.0, False) + elif(opt.lower() == 'callloop-maxlen'): + sysvals.callloopmaxgap = getArgFloat('-callloop-maxlen', value, 0.0, 1.0, False) + elif(opt.lower() == 'mincg'): + sysvals.mincglen = getArgFloat('-mincg', value, 0.0, 10000.0, False) + elif(opt.lower() == 'output-dir'): + sysvals.setOutputFolder(value) + + if sysvals.suspendmode == 'command' and not sysvals.testcommand: + doError('No command supplied for mode "command"') + + # compatibility errors + if sysvals.usedevsrc and sysvals.usecallgraph: + doError('-dev is not compatible with -f') + if sysvals.usecallgraph and sysvals.useprocmon: + doError('-proc is not compatible with -f') + + if overridekprobes: + sysvals.tracefuncs = dict() + if overridedevkprobes: + sysvals.dev_tracefuncs = dict() + + kprobes = dict() + kprobesec = 'dev_timeline_functions_'+platform.machine() + if kprobesec in sections: + for name in Config.options(kprobesec): + text = Config.get(kprobesec, name) + kprobes[name] = (text, True) + kprobesec = 'timeline_functions_'+platform.machine() + if kprobesec in sections: + for name in Config.options(kprobesec): + if name in kprobes: + doError('Duplicate timeline function found "%s"' % (name)) + text = Config.get(kprobesec, name) + kprobes[name] = (text, False) + + for name in kprobes: + function = name + format = name + color = '' + args = dict() + text, dev = kprobes[name] + data = text.split() + i = 0 + for val in data: + # bracketted strings are special formatting, read them separately + if val[0] == '[' and val[-1] == ']': + for prop in val[1:-1].split(','): + p = prop.split('=') + if p[0] == 'color': + try: + color = int(p[1], 16) + color = '#'+p[1] + except: + color = p[1] + continue + # first real arg should be the format string + if i == 0: + format = val + # all other args are actual function args + else: + d = val.split('=') + args[d[0]] = d[1] + i += 1 + if not function or not format: + doError('Invalid kprobe: %s' % name) + for arg in re.findall('{(?P<n>[a-z,A-Z,0-9]*)}', format): + if arg not in args: + doError('Kprobe "%s" is missing argument "%s"' % (name, arg)) + if (dev and name in sysvals.dev_tracefuncs) or (not dev and name in sysvals.tracefuncs): + doError('Duplicate timeline function found "%s"' % (name)) + + kp = { + 'name': name, + 'func': function, + 'format': format, + sysvals.archargs: args + } + if color: + kp['color'] = color + if dev: + sysvals.dev_tracefuncs[name] = kp + else: + sysvals.tracefuncs[name] = kp + +# Function: printHelp +# Description: +# print out the help text +def printHelp(): + modes = getModes() + + print('') + print('%s v%s' % (sysvals.title, sysvals.version)) + print('Usage: sudo sleepgraph <options> <commands>') + print('') + print('Description:') + print(' This tool is designed to assist kernel and OS developers in optimizing') + print(' their linux stack\'s suspend/resume time. Using a kernel image built') + print(' with a few extra options enabled, the tool will execute a suspend and') + print(' capture dmesg and ftrace data until resume is complete. This data is') + print(' transformed into a device timeline and an optional callgraph to give') + print(' a detailed view of which devices/subsystems are taking the most') + print(' time in suspend/resume.') + print('') + print(' If no specific command is given, the default behavior is to initiate') + print(' a suspend/resume and capture the dmesg/ftrace output as an html timeline.') + print('') + print(' Generates output files in subdirectory: suspend-mmddyy-HHMMSS') + print(' HTML output: <hostname>_<mode>.html') + print(' raw dmesg output: <hostname>_<mode>_dmesg.txt') + print(' raw ftrace output: <hostname>_<mode>_ftrace.txt') + print('') + print('Options:') + print(' -h Print this help text') + print(' -v Print the current tool version') + print(' -config fn Pull arguments and config options from file fn') + print(' -verbose Print extra information during execution and analysis') + print(' -m mode Mode to initiate for suspend %s (default: %s)') % (modes, sysvals.suspendmode) + print(' -o subdir Override the output subdirectory') + print(' -rtcwake t Wakeup t seconds after suspend, set t to "off" to disable (default: 15)') + print(' -addlogs Add the dmesg and ftrace logs to the html output') + print(' -srgap Add a visible gap in the timeline between sus/res (default: disabled)') + print(' [advanced]') + print(' -cmd {s} Run the timeline over a custom command, e.g. "sync -d"') + print(' -proc Add usermode process info into the timeline (default: disabled)') + print(' -dev Add kernel function calls and threads to the timeline (default: disabled)') + print(' -x2 Run two suspend/resumes back to back (default: disabled)') + print(' -x2delay t Include t ms delay between multiple test runs (default: 0 ms)') + print(' -predelay t Include t ms delay before 1st suspend (default: 0 ms)') + print(' -postdelay t Include t ms delay after last resume (default: 0 ms)') + print(' -mindev ms Discard all device blocks shorter than ms milliseconds (e.g. 0.001 for us)') + print(' -multi n d Execute <n> consecutive tests at <d> seconds intervals. The outputs will') + print(' be created in a new subdirectory with a summary page.') + print(' [debug]') + print(' -f Use ftrace to create device callgraphs (default: disabled)') + print(' -maxdepth N limit the callgraph data to N call levels (default: 0=all)') + print(' -expandcg pre-expand the callgraph data in the html output (default: disabled)') + print(' -fadd file Add functions to be graphed in the timeline from a list in a text file') + print(' -filter "d1,d2,..." Filter out all but this comma-delimited list of device names') + print(' -mincg ms Discard all callgraphs shorter than ms milliseconds (e.g. 0.001 for us)') + print(' -cgphase P Only show callgraph data for phase P (e.g. suspend_late)') + print(' -cgtest N Only show callgraph data for test N (e.g. 0 or 1 in an x2 run)') + print(' -timeprec N Number of significant digits in timestamps (0:S, [3:ms], 6:us)') + print(' [commands]') + print(' -ftrace ftracefile Create HTML output using ftrace input (used with -dmesg)') + print(' -dmesg dmesgfile Create HTML output using dmesg (used with -ftrace)') + print(' -summary directory Create a summary of all test in this dir') + print(' -modes List available suspend modes') + print(' -status Test to see if the system is enabled to run this tool') + print(' -fpdt Print out the contents of the ACPI Firmware Performance Data Table') + print(' -usbtopo Print out the current USB topology with power info') + print(' -usbauto Enable autosuspend for all connected USB devices') + print(' -flist Print the list of functions currently being captured in ftrace') + print(' -flistall Print all functions capable of being captured in ftrace') + print('') + return True + +# ----------------- MAIN -------------------- +# exec start (skipped if script is loaded as library) +if __name__ == '__main__': + cmd = '' + cmdarg = '' + multitest = {'run': False, 'count': 0, 'delay': 0} + simplecmds = ['-modes', '-fpdt', '-flist', '-flistall', '-usbtopo', '-usbauto', '-status'] + # loop through the command line arguments + args = iter(sys.argv[1:]) + for arg in args: + if(arg == '-m'): + try: + val = args.next() + except: + doError('No mode supplied', True) + if val == 'command' and not sysvals.testcommand: + doError('No command supplied for mode "command"', True) + sysvals.suspendmode = val + elif(arg in simplecmds): + cmd = arg[1:] + elif(arg == '-h'): + printHelp() + sys.exit() + elif(arg == '-v'): + print("Version %s" % sysvals.version) + sys.exit() + elif(arg == '-x2'): + sysvals.execcount = 2 + elif(arg == '-x2delay'): + sysvals.x2delay = getArgInt('-x2delay', args, 0, 60000) + elif(arg == '-predelay'): + sysvals.predelay = getArgInt('-predelay', args, 0, 60000) + elif(arg == '-postdelay'): + sysvals.postdelay = getArgInt('-postdelay', args, 0, 60000) + elif(arg == '-f'): + sysvals.usecallgraph = True + elif(arg == '-addlogs'): + sysvals.addlogs = True + elif(arg == '-verbose'): + sysvals.verbose = True + elif(arg == '-proc'): + sysvals.useprocmon = True + elif(arg == '-dev'): + sysvals.usedevsrc = True + elif(arg == '-maxdepth'): + sysvals.max_graph_depth = getArgInt('-maxdepth', args, 0, 1000) + elif(arg == '-rtcwake'): + try: + val = args.next() + except: + doError('No rtcwake time supplied', True) + if val.lower() == 'off': + sysvals.rtcwake = False + else: + sysvals.rtcwake = True + sysvals.rtcwaketime = getArgInt('-rtcwake', val, 0, 3600, False) + elif(arg == '-timeprec'): + sysvals.setPrecision(getArgInt('-timeprec', args, 0, 6)) + elif(arg == '-mindev'): + sysvals.mindevlen = getArgFloat('-mindev', args, 0.0, 10000.0) + elif(arg == '-mincg'): + sysvals.mincglen = getArgFloat('-mincg', args, 0.0, 10000.0) + elif(arg == '-cgtest'): + sysvals.cgtest = getArgInt('-cgtest', args, 0, 1) + elif(arg == '-cgphase'): + try: + val = args.next() + except: + doError('No phase name supplied', True) + d = Data(0) + if val not in d.phases: + doError('Invalid phase, valid phaess are %s' % d.phases, True) + sysvals.cgphase = val + elif(arg == '-callloop-maxgap'): + sysvals.callloopmaxgap = getArgFloat('-callloop-maxgap', args, 0.0, 1.0) + elif(arg == '-callloop-maxlen'): + sysvals.callloopmaxlen = getArgFloat('-callloop-maxlen', args, 0.0, 1.0) + elif(arg == '-cmd'): + try: + val = args.next() + except: + doError('No command string supplied', True) + sysvals.testcommand = val + sysvals.suspendmode = 'command' + elif(arg == '-expandcg'): + sysvals.cgexp = True + elif(arg == '-srgap'): + sysvals.srgap = 5 + elif(arg == '-multi'): + multitest['run'] = True + multitest['count'] = getArgInt('-multi n (exec count)', args, 2, 1000000) + multitest['delay'] = getArgInt('-multi d (delay between tests)', args, 0, 3600) + elif(arg == '-o'): + try: + val = args.next() + except: + doError('No subdirectory name supplied', True) + sysvals.setOutputFolder(val) + elif(arg == '-config'): + try: + val = args.next() + except: + doError('No text file supplied', True) + if(os.path.exists(val) == False): + doError('%s does not exist' % val) + configFromFile(val) + elif(arg == '-fadd'): + try: + val = args.next() + except: + doError('No text file supplied', True) + if(os.path.exists(val) == False): + doError('%s does not exist' % val) + sysvals.addFtraceFilterFunctions(val) + elif(arg == '-dmesg'): + try: + val = args.next() + except: + doError('No dmesg file supplied', True) + sysvals.notestrun = True + sysvals.dmesgfile = val + if(os.path.exists(sysvals.dmesgfile) == False): + doError('%s does not exist' % sysvals.dmesgfile) + elif(arg == '-ftrace'): + try: + val = args.next() + except: + doError('No ftrace file supplied', True) + sysvals.notestrun = True + sysvals.ftracefile = val + if(os.path.exists(sysvals.ftracefile) == False): + doError('%s does not exist' % sysvals.ftracefile) + elif(arg == '-summary'): + try: + val = args.next() + except: + doError('No directory supplied', True) + cmd = 'summary' + cmdarg = val + sysvals.notestrun = True + if(os.path.isdir(val) == False): + doError('%s is not accesible' % val) + elif(arg == '-filter'): + try: + val = args.next() + except: + doError('No devnames supplied', True) + sysvals.setDeviceFilter(val) + else: + doError('Invalid argument: '+arg, True) + + # compatibility errors + if(sysvals.usecallgraph and sysvals.usedevsrc): + doError('-dev is not compatible with -f') + if(sysvals.usecallgraph and sysvals.useprocmon): + doError('-proc is not compatible with -f') + + # callgraph size cannot exceed device size + if sysvals.mincglen < sysvals.mindevlen: + sysvals.mincglen = sysvals.mindevlen + + # just run a utility command and exit + if(cmd != ''): + if(cmd == 'status'): + statusCheck(True) + elif(cmd == 'fpdt'): + getFPDT(True) + elif(cmd == 'usbtopo'): + detectUSB() + elif(cmd == 'modes'): + print getModes() + elif(cmd == 'flist'): + sysvals.getFtraceFilterFunctions(True) + elif(cmd == 'flistall'): + sysvals.getFtraceFilterFunctions(False) + elif(cmd == 'usbauto'): + setUSBDevicesAuto() + elif(cmd == 'summary'): + runSummary(cmdarg, True) + sys.exit() + + # if instructed, re-analyze existing data files + if(sysvals.notestrun): + rerunTest() + sys.exit() + + # verify that we can run a test + if(not statusCheck()): + print('Check FAILED, aborting the test run!') + sys.exit() + + if multitest['run']: + # run multiple tests in a separate subdirectory + s = 'x%d' % multitest['count'] + if not sysvals.outdir: + sysvals.outdir = datetime.now().strftime('suspend-'+s+'-%m%d%y-%H%M%S') + if not os.path.isdir(sysvals.outdir): + os.mkdir(sysvals.outdir) + for i in range(multitest['count']): + if(i != 0): + print('Waiting %d seconds...' % (multitest['delay'])) + time.sleep(multitest['delay']) + print('TEST (%d/%d) START' % (i+1, multitest['count'])) + runTest(sysvals.outdir) + print('TEST (%d/%d) COMPLETE' % (i+1, multitest['count'])) + runSummary(sysvals.outdir, False) + else: + # run the test in the current directory + runTest('.', sysvals.outdir) diff --git a/tools/power/pm-graph/bootgraph.8 b/tools/power/pm-graph/bootgraph.8 new file mode 100644 index 000000000000..55272a67b0e7 --- /dev/null +++ b/tools/power/pm-graph/bootgraph.8 @@ -0,0 +1,132 @@ +.TH BOOTGRAPH 8 +.SH NAME +bootgraph \- Kernel boot timing analysis +.SH SYNOPSIS +.ft B +.B bootgraph +.RB [ OPTIONS ] +.RB [ COMMAND ] +.SH DESCRIPTION +\fBbootgraph \fP reads the dmesg log from kernel boot and +creates an html representation of the initcall timeline up to the start +of the init process. +.PP +If no specific command is given, the tool reads the current dmesg log and +outputs bootgraph.html. +.PP +The tool can also augment the timeline with ftrace data on custom target +functions as well as full trace callgraphs. +.SH OPTIONS +.TP +\fB-h\fR +Print this help text +.TP +\fB-v\fR +Print the current tool version +.TP +\fB-addlogs\fR +Add the dmesg log to the html output. It will be viewable by +clicking a button in the timeline. +.TP +\fB-o \fIfile\fR +Override the HTML output filename (default: bootgraph.html) +.SS "Ftrace Debug" +.TP +\fB-f\fR +Use ftrace to add function detail (default: disabled) +.TP +\fB-callgraph\fR +Use ftrace to create initcall callgraphs (default: disabled). If -filter +is not used there will be one callgraph per initcall. This can produce +very large outputs, i.e. 10MB - 100MB. +.TP +\fB-maxdepth \fIlevel\fR +limit the callgraph trace depth to \fIlevel\fR (default: 2). This is +the best way to limit the output size when using -callgraph. +.TP +\fB-mincg \fIt\fR +Discard all callgraphs shorter than \fIt\fR milliseconds (default: 0=all). +This reduces the html file size as there can be many tiny callgraphs +which are barely visible in the timeline. +The value is a float: e.g. 0.001 represents 1 us. +.TP +\fB-timeprec \fIn\fR +Number of significant digits in timestamps (0:S, 3:ms, [6:us]) +.TP +\fB-expandcg\fR +pre-expand the callgraph data in the html output (default: disabled) +.TP +\fB-filter \fI"func1,func2,..."\fR +Instead of tracing each initcall, trace a custom list of functions (default: do_one_initcall) + +.SH COMMANDS +.TP +\fB-reboot\fR +Reboot the machine and generate a new timeline automatically. Works in 4 steps. + 1. updates grub with the required kernel parameters + 2. installs a cron job which re-runs the tool after reboot + 3. reboots the system + 4. after startup, extracts the data and generates the timeline +.TP +\fB-manual\fR +Show the requirements to generate a new timeline manually. Requires 3 steps. + 1. append the string to the kernel command line via your native boot manager. + 2. reboot the system + 3. after startup, re-run the tool with the same arguments and no command +.TP +\fB-dmesg \fIfile\fR +Create HTML output from an existing dmesg file. +.TP +\fB-ftrace \fIfile\fR +Create HTML output from an existing ftrace file (used with -dmesg). +.TP +\fB-flistall\fR +Print all ftrace functions capable of being captured. These are all the +possible values you can add to trace via the -filter argument. + +.SH EXAMPLES +Create a timeline using the current dmesg log. +.IP +\f(CW$ bootgraph\fR +.PP +Create a timeline using the current dmesg and ftrace log. +.IP +\f(CW$ bootgraph -callgraph\fR +.PP +Create a timeline using the current dmesg, add the log to the html and change the name. +.IP +\f(CW$ bootgraph -addlogs -o myboot.html\fR +.PP +Capture a new boot timeline by automatically rebooting the machine. +.IP +\f(CW$ sudo bootgraph -reboot -addlogs -o latestboot.html\fR +.PP +Capture a new boot timeline with function trace data. +.IP +\f(CW$ sudo bootgraph -reboot -f\fR +.PP +Capture a new boot timeline with trace & callgraph data. Skip callgraphs smaller than 5ms. +.IP +\f(CW$ sudo bootgraph -reboot -callgraph -mincg 5\fR +.PP +Capture a new boot timeline with callgraph data over custom functions. +.IP +\f(CW$ sudo bootgraph -reboot -callgraph -filter "acpi_ps_parse_aml,msleep"\fR +.PP +Capture a brand new boot timeline with manual reboot. +.IP +\f(CW$ sudo bootgraph -callgraph -manual\fR +.IP +\f(CW$ vi /etc/default/grub # add the CMDLINE string to your kernel params\fR +.IP +\f(CW$ sudo reboot # reboot the machine\fR +.IP +\f(CW$ sudo bootgraph -callgraph # re-run the tool after restart\fR +.PP + +.SH "SEE ALSO" +dmesg(1), update-grub(8), crontab(1), reboot(8) +.PP +.SH AUTHOR +.nf +Written by Todd Brandt <todd.e.brandt@linux.intel.com> diff --git a/tools/power/pm-graph/sleepgraph.8 b/tools/power/pm-graph/sleepgraph.8 new file mode 100644 index 000000000000..610e72ebbc06 --- /dev/null +++ b/tools/power/pm-graph/sleepgraph.8 @@ -0,0 +1,243 @@ +.TH SLEEPGRAPH 8 +.SH NAME +sleepgraph \- Suspend/Resume timing analysis +.SH SYNOPSIS +.ft B +.B sleepgraph +.RB [ OPTIONS ] +.RB [ COMMAND ] +.SH DESCRIPTION +\fBsleepgraph \fP is designed to assist kernel and OS developers +in optimizing their linux stack's suspend/resume time. Using a kernel +image built with a few extra options enabled, the tool will execute a +suspend and capture dmesg and ftrace data until resume is complete. +This data is transformed into a device timeline and an optional +callgraph to give a detailed view of which devices/subsystems are +taking the most time in suspend/resume. +.PP +If no specific command is given, the default behavior is to initiate +a suspend/resume. +.PP +Generates output files in subdirectory: suspend-yymmdd-HHMMSS + html timeline : <hostname>_<mode>.html + raw dmesg file : <hostname>_<mode>_dmesg.txt + raw ftrace file : <hostname>_<mode>_ftrace.txt +.SH OPTIONS +.TP +\fB-h\fR +Print the help text. +.TP +\fB-v\fR +Print the current tool version. +.TP +\fB-verbose\fR +Print extra information during execution and analysis. +.TP +\fB-config \fIfile\fR +Pull arguments and config options from a file. +.TP +\fB-m \fImode\fR +Mode to initiate for suspend e.g. standby, freeze, mem (default: mem). +.TP +\fB-o \fIsubdir\fR +Override the output subdirectory. Use {date}, {time}, {hostname} for current values. +.sp +e.g. suspend-{hostname}-{date}-{time} +.TP +\fB-rtcwake \fIt\fR | off +Use rtcwake to autoresume after \fIt\fR seconds (default: 15). Set t to "off" to +disable rtcwake and require a user keypress to resume. +.TP +\fB-addlogs\fR +Add the dmesg and ftrace logs to the html output. They will be viewable by +clicking buttons in the timeline. + +.SS "Advanced" +.TP +\fB-cmd \fIstr\fR +Run the timeline over a custom suspend command, e.g. pm-suspend. By default +the tool forces suspend via /sys/power/state so this allows testing over +an OS's official suspend method. The output file will change to +hostname_command.html and will autodetect which suspend mode was triggered. +.TP +\fB-filter \fI"d1,d2,..."\fR +Filter out all but these device callbacks. These strings can be device names +or module names. e.g. 0000:00:02.0, ata5, i915, usb, etc. +.TP +\fB-mindev \fIt\fR +Discard all device callbacks shorter than \fIt\fR milliseconds (default: 0.0). +This reduces the html file size as there can be many tiny callbacks which are barely +visible. The value is a float: e.g. 0.001 represents 1 us. +.TP +\fB-proc\fR +Add usermode process info into the timeline (default: disabled). +.TP +\fB-dev\fR +Add kernel source calls and threads to the timeline (default: disabled). +.TP +\fB-x2\fR +Run two suspend/resumes back to back (default: disabled). +.TP +\fB-x2delay \fIt\fR +Include \fIt\fR ms delay between multiple test runs (default: 0 ms). +.TP +\fB-predelay \fIt\fR +Include \fIt\fR ms delay before 1st suspend (default: 0 ms). +.TP +\fB-postdelay \fIt\fR +Include \fIt\fR ms delay after last resume (default: 0 ms). +.TP +\fB-multi \fIn d\fR +Execute \fIn\fR consecutive tests at \fId\fR seconds intervals. The outputs will +be created in a new subdirectory with a summary page: suspend-xN-{date}-{time}. + +.SS "Ftrace Debug" +.TP +\fB-f\fR +Use ftrace to create device callgraphs (default: disabled). This can produce +very large outputs, i.e. 10MB - 100MB. +.TP +\fB-maxdepth \fIlevel\fR +limit the callgraph trace depth to \fIlevel\fR (default: 0=all). This is +the best way to limit the output size when using callgraphs via -f. +.TP +\fB-expandcg\fR +pre-expand the callgraph data in the html output (default: disabled) +.TP +\fB-fadd \fIfile\fR +Add functions to be graphed in the timeline from a list in a text file +.TP +\fB-mincg \fIt\fR +Discard all callgraphs shorter than \fIt\fR milliseconds (default: 0.0). +This reduces the html file size as there can be many tiny callgraphs +which are barely visible in the timeline. +The value is a float: e.g. 0.001 represents 1 us. +.TP +\fB-cgphase \fIp\fR +Only show callgraph data for phase \fIp\fR (e.g. suspend_late). +.TP +\fB-cgtest \fIn\fR +In an x2 run, only show callgraph data for test \fIn\fR (e.g. 0 or 1). +.TP +\fB-timeprec \fIn\fR +Number of significant digits in timestamps (0:S, [3:ms], 6:us). + +.SH COMMANDS +.TP +\fB-ftrace \fIfile\fR +Create HTML output from an existing ftrace file. +.TP +\fB-dmesg \fIfile\fR +Create HTML output from an existing dmesg file. +.TP +\fB-summary \fIindir\fR +Create a summary page of all tests in \fIindir\fR. Creates summary.html +in the current folder. The output page is a table of tests with +suspend and resume values sorted by suspend mode, host, and kernel. +Includes test averages by mode and links to the test html files. +.TP +\fB-modes\fR +List available suspend modes. +.TP +\fB-status\fR +Test to see if the system is able to run this tool. Use this along +with any options you intend to use to see if they will work. +.TP +\fB-fpdt\fR +Print out the contents of the ACPI Firmware Performance Data Table. +.TP +\fB-usbtopo\fR +Print out the current USB topology with power info. +.TP +\fB-usbauto\fR +Enable autosuspend for all connected USB devices. +.TP +\fB-flist\fR +Print the list of ftrace functions currently being captured. Functions +that are not available as symbols in the current kernel are shown in red. +By default, the tool traces a list of important suspend/resume functions +in order to better fill out the timeline. If the user has added their own +with -fadd they will also be checked. +.TP +\fB-flistall\fR +Print all ftrace functions capable of being captured. These are all the +possible values you can add to trace via the -fadd argument. + +.SH EXAMPLES +.SS "Simple Commands" +Check which suspend modes are currently supported. +.IP +\f(CW$ sleepgraph -modes\fR +.PP +Read the Firmware Performance Data Table (FPDT) +.IP +\f(CW$ sudo sleepgraph -fpdt\fR +.PP +Print out the current USB power topology +.IP +\f(CW$ sleepgraph -usbtopo +.PP +Verify that you can run a command with a set of arguments +.IP +\f(CW$ sudo sleepgraph -f -rtcwake 30 -status +.PP +Generate a summary of all timelines in a particular folder. +.IP +\f(CW$ sleepgraph -summary ~/workspace/myresults/\fR +.PP +Re-generate the html output from a previous run's dmesg and ftrace log. +.IP +\f(CW$ sleepgraph -dmesg myhost_mem_dmesg.txt -ftrace myhost_mem_ftrace.txt\fR +.PP + +.SS "Capturing Simple Timelines" +Execute a mem suspend with a 15 second wakeup. Include the logs in the html. +.IP +\f(CW$ sudo sleepgraph -rtcwake 15 -addlogs\fR +.PP +Execute a standby with a 15 second wakeup. Change the output folder name. +.IP +\f(CW$ sudo sleepgraph -m standby -rtcwake 15 -o "standby-{hostname}-{date}-{time}"\fR +.PP +Execute a freeze with no wakeup (require keypress). Change output folder name. +.IP +\f(CW$ sudo sleepgraph -m freeze -rtcwake off -o "freeze-{hostname}-{date}-{time}"\fR +.PP + +.SS "Capturing Advanced Timelines" +Execute a suspend & include dev mode source calls, limit callbacks to 5ms or larger. +.IP +\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -dev -mindev 5\fR +.PP +Run two suspends back to back, include a 500ms delay before, after, and in between runs. +.IP +\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -x2 -predelay 500 -x2delay 500 -postdelay 500\fR +.PP +Do a batch run of 10 freezes with 30 seconds delay between runs. +.IP +\f(CW$ sudo sleepgraph -m freeze -rtcwake 15 -multi 10 30\fR +.PP +Execute a suspend using a custom command. +.IP +\f(CW$ sudo sleepgraph -cmd "echo mem > /sys/power/state" -rtcwake 15\fR +.PP + + +.SS "Capturing Timelines with Callgraph Data" +Add device callgraphs. Limit the trace depth and only show callgraphs 10ms or larger. +.IP +\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -f -maxdepth 5 -mincg 10\fR +.PP +Capture a full callgraph across all suspend, then filter the html by a single phase. +.IP +\f(CW$ sudo sleepgraph -m mem -rtcwake 15 -f\fR +.IP +\f(CW$ sleepgraph -dmesg host_mem_dmesg.txt -ftrace host_mem_ftrace.txt -f -cgphase resume +.PP + +.SH "SEE ALSO" +dmesg(1) +.PP +.SH AUTHOR +.nf +Written by Todd Brandt <todd.e.brandt@linux.intel.com> diff --git a/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py b/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py new file mode 100755 index 000000000000..0b24dd9d01ff --- /dev/null +++ b/tools/power/x86/intel_pstate_tracer/intel_pstate_tracer.py @@ -0,0 +1,586 @@ +#!/usr/bin/python +# -*- coding: utf-8 -*- +# +""" This utility can be used to debug and tune the performance of the +intel_pstate driver. This utility can be used in two ways: +- If there is Linux trace file with pstate_sample events enabled, then +this utility can parse the trace file and generate performance plots. +- If user has not specified a trace file as input via command line parameters, +then this utility enables and collects trace data for a user specified interval +and generates performance plots. + +Prerequisites: + Python version 2.7.x + gnuplot 5.0 or higher + gnuplot-py 1.8 + (Most of the distributions have these required packages. They may be called + gnuplot-py, phython-gnuplot. ) + + HWP (Hardware P-States are disabled) + Kernel config for Linux trace is enabled + + see print_help(): for Usage and Output details + +""" +from __future__ import print_function +from datetime import datetime +import subprocess +import os +import time +import re +import sys +import getopt +import Gnuplot +from numpy import * +from decimal import * + +__author__ = "Srinivas Pandruvada" +__copyright__ = " Copyright (c) 2017, Intel Corporation. " +__license__ = "GPL version 2" + + +MAX_CPUS = 256 + +# Define the csv file columns +C_COMM = 18 +C_GHZ = 17 +C_ELAPSED = 16 +C_SAMPLE = 15 +C_DURATION = 14 +C_LOAD = 13 +C_BOOST = 12 +C_FREQ = 11 +C_TSC = 10 +C_APERF = 9 +C_MPERF = 8 +C_TO = 7 +C_FROM = 6 +C_SCALED = 5 +C_CORE = 4 +C_USEC = 3 +C_SEC = 2 +C_CPU = 1 + +global sample_num, last_sec_cpu, last_usec_cpu, start_time, testname + +# 11 digits covers uptime to 115 days +getcontext().prec = 11 + +sample_num =0 +last_sec_cpu = [0] * MAX_CPUS +last_usec_cpu = [0] * MAX_CPUS + +def print_help(): + print('intel_pstate_tracer.py:') + print(' Usage:') + print(' If the trace file is available, then to simply parse and plot, use (sudo not required):') + print(' ./intel_pstate_tracer.py [-c cpus] -t <trace_file> -n <test_name>') + print(' Or') + print(' ./intel_pstate_tracer.py [--cpu cpus] ---trace_file <trace_file> --name <test_name>') + print(' To generate trace file, parse and plot, use (sudo required):') + print(' sudo ./intel_pstate_tracer.py [-c cpus] -i <interval> -n <test_name>') + print(' Or') + print(' sudo ./intel_pstate_tracer.py [--cpu cpus] --interval <interval> --name <test_name>') + print(' Optional argument:') + print(' cpus: comma separated list of CPUs') + print(' Output:') + print(' If not already present, creates a "results/test_name" folder in the current working directory with:') + print(' cpu.csv - comma seperated values file with trace contents and some additional calculations.') + print(' cpu???.csv - comma seperated values file for CPU number ???.') + print(' *.png - a variety of PNG format plot files created from the trace contents and the additional calculations.') + print(' Notes:') + print(' Avoid the use of _ (underscore) in test names, because in gnuplot it is a subscript directive.') + print(' Maximum number of CPUs is {0:d}. If there are more the script will abort with an error.'.format(MAX_CPUS)) + print(' Off-line CPUs cause the script to list some warnings, and create some empty files. Use the CPU mask feature for a clean run.') + print(' Empty y range warnings for autoscaled plots can occur and can be ignored.') + +def plot_perf_busy_with_sample(cpu_index): + """ Plot method to per cpu information """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_perf_busy_vs_samples.png" % cpu_index + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:40]') + g_plot('set y2range [0:200]') + g_plot('set y2tics 0, 10') + g_plot('set title "{} : cpu perf busy vs. sample : CPU {:0>3} : {:%F %H:%M}"'.format(testname, cpu_index, datetime.now())) +# Override common + g_plot('set xlabel "Samples"') + g_plot('set ylabel "P-State"') + g_plot('set y2label "Scaled Busy/performance/io-busy(%)"') + set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y2 title "performance",\\'.format(C_SAMPLE, C_CORE)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 2 axis x1y2 title "scaled-busy",\\'.format(C_SAMPLE, C_SCALED)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 3 axis x1y2 title "io-boost",\\'.format(C_SAMPLE, C_BOOST)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 4 axis x1y1 title "P-State"'.format(C_SAMPLE, C_TO)) + +def plot_perf_busy(cpu_index): + """ Plot some per cpu information """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_perf_busy.png" % cpu_index + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:40]') + g_plot('set y2range [0:200]') + g_plot('set y2tics 0, 10') + g_plot('set title "{} : perf busy : CPU {:0>3} : {:%F %H:%M}"'.format(testname, cpu_index, datetime.now())) + g_plot('set ylabel "P-State"') + g_plot('set y2label "Scaled Busy/performance/io-busy(%)"') + set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y2 title "performance",\\'.format(C_ELAPSED, C_CORE)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 2 axis x1y2 title "scaled-busy",\\'.format(C_ELAPSED, C_SCALED)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 3 axis x1y2 title "io-boost",\\'.format(C_ELAPSED, C_BOOST)) + g_plot('"' + file_name + '" using {:d}:{:d} with linespoints linestyle 4 axis x1y1 title "P-State"'.format(C_ELAPSED, C_TO)) + +def plot_durations(cpu_index): + """ Plot per cpu durations """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_durations.png" % cpu_index + g_plot = common_all_gnuplot_settings(output_png) +# Should autoscale be used here? Should seconds be used here? + g_plot('set yrange [0:5000]') + g_plot('set ytics 0, 500') + g_plot('set title "{} : durations : CPU {:0>3} : {:%F %H:%M}"'.format(testname, cpu_index, datetime.now())) + g_plot('set ylabel "Timer Duration (MilliSeconds)"') +# override common + g_plot('set key off') + set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y1'.format(C_ELAPSED, C_DURATION)) + +def plot_loads(cpu_index): + """ Plot per cpu loads """ + + file_name = 'cpu{:0>3}.csv'.format(cpu_index) + if os.path.exists(file_name): + output_png = "cpu%03d_loads.png" % cpu_index + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:100]') + g_plot('set ytics 0, 10') + g_plot('set title "{} : loads : CPU {:0>3} : {:%F %H:%M}"'.format(testname, cpu_index, datetime.now())) + g_plot('set ylabel "CPU load (percent)"') +# override common + g_plot('set key off') + set_4_plot_linestyles(g_plot) + g_plot('plot "' + file_name + '" using {:d}:{:d} with linespoints linestyle 1 axis x1y1'.format(C_ELAPSED, C_LOAD)) + +def plot_pstate_cpu_with_sample(): + """ Plot all cpu information """ + + if os.path.exists('cpu.csv'): + output_png = 'all_cpu_pstates_vs_samples.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:40]') +# override common + g_plot('set xlabel "Samples"') + g_plot('set ylabel "P-State"') + g_plot('set title "{} : cpu pstate vs. sample : {:%F %H:%M}"'.format(testname, datetime.now())) + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_SAMPLE, C_TO) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_pstate_cpu(): + """ Plot all cpu information from csv files """ + + output_png = 'all_cpu_pstates.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:40]') + g_plot('set ylabel "P-State"') + g_plot('set title "{} : cpu pstates : {:%F %H:%M}"'.format(testname, datetime.now())) + +# the following command is really cool, but doesn't work with the CPU masking option because it aborts on the first missing file. +# plot_str = 'plot for [i=0:*] file=sprintf("cpu%03d.csv",i) title_s=sprintf("cpu%03d",i) file using 16:7 pt 7 ps 1 title title_s' +# + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_TO) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_load_cpu(): + """ Plot all cpu loads """ + + output_png = 'all_cpu_loads.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:100]') + g_plot('set ylabel "CPU load (percent)"') + g_plot('set title "{} : cpu loads : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_LOAD) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_frequency_cpu(): + """ Plot all cpu frequencies """ + + output_png = 'all_cpu_frequencies.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:4]') + g_plot('set ylabel "CPU Frequency (GHz)"') + g_plot('set title "{} : cpu frequencies : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_FREQ) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_duration_cpu(): + """ Plot all cpu durations """ + + output_png = 'all_cpu_durations.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:5000]') + g_plot('set ytics 0, 500') + g_plot('set ylabel "Timer Duration (MilliSeconds)"') + g_plot('set title "{} : cpu durations : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_DURATION) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_scaled_cpu(): + """ Plot all cpu scaled busy """ + + output_png = 'all_cpu_scaled.png' + g_plot = common_all_gnuplot_settings(output_png) +# autoscale this one, no set y range + g_plot('set ylabel "Scaled Busy (Unitless)"') + g_plot('set title "{} : cpu scaled busy : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_SCALED) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_boost_cpu(): + """ Plot all cpu IO Boosts """ + + output_png = 'all_cpu_boost.png' + g_plot = common_all_gnuplot_settings(output_png) + g_plot('set yrange [0:100]') + g_plot('set ylabel "CPU IO Boost (percent)"') + g_plot('set title "{} : cpu io boost : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_BOOST) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def plot_ghz_cpu(): + """ Plot all cpu tsc ghz """ + + output_png = 'all_cpu_ghz.png' + g_plot = common_all_gnuplot_settings(output_png) +# autoscale this one, no set y range + g_plot('set ylabel "TSC Frequency (GHz)"') + g_plot('set title "{} : cpu TSC Frequencies (Sanity check calculation) : {:%F %H:%M}"'.format(testname, datetime.now())) + + title_list = subprocess.check_output('ls cpu???.csv | sed -e \'s/.csv//\'',shell=True).replace('\n', ' ') + plot_str = "plot for [i in title_list] i.'.csv' using {:d}:{:d} pt 7 ps 1 title i".format(C_ELAPSED, C_GHZ) + g_plot('title_list = "{}"'.format(title_list)) + g_plot(plot_str) + +def common_all_gnuplot_settings(output_png): + """ common gnuplot settings for multiple CPUs one one graph. """ + + g_plot = common_gnuplot_settings() + g_plot('set output "' + output_png + '"') + return(g_plot) + +def common_gnuplot_settings(): + """ common gnuplot settings. """ + + g_plot = Gnuplot.Gnuplot(persist=1) +# The following line is for rigor only. It seems to be assumed for .csv files + g_plot('set datafile separator \",\"') + g_plot('set ytics nomirror') + g_plot('set xtics nomirror') + g_plot('set xtics font ", 10"') + g_plot('set ytics font ", 10"') + g_plot('set tics out scale 1.0') + g_plot('set grid') + g_plot('set key out horiz') + g_plot('set key bot center') + g_plot('set key samplen 2 spacing .8 font ", 9"') + g_plot('set term png size 1200, 600') + g_plot('set title font ", 11"') + g_plot('set ylabel font ", 10"') + g_plot('set xlabel font ", 10"') + g_plot('set xlabel offset 0, 0.5') + g_plot('set xlabel "Elapsed Time (Seconds)"') + return(g_plot) + +def set_4_plot_linestyles(g_plot): + """ set the linestyles used for 4 plots in 1 graphs. """ + + g_plot('set style line 1 linetype 1 linecolor rgb "green" pointtype -1') + g_plot('set style line 2 linetype 1 linecolor rgb "red" pointtype -1') + g_plot('set style line 3 linetype 1 linecolor rgb "purple" pointtype -1') + g_plot('set style line 4 linetype 1 linecolor rgb "blue" pointtype -1') + +def store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz): + """ Store master csv file information """ + + global graph_data_present + + if cpu_mask[cpu_int] == 0: + return + + try: + f_handle = open('cpu.csv', 'a') + string_buffer = "CPU_%03u, %05u, %06u, %u, %u, %u, %u, %u, %u, %u, %.4f, %u, %.2f, %.3f, %u, %.3f, %.3f, %s\n" % (cpu_int, int(time_pre_dec), int(time_post_dec), int(core_busy), int(scaled), int(_from), int(_to), int(mperf), int(aperf), int(tsc), freq_ghz, int(io_boost), load, duration_ms, sample_num, elapsed_time, tsc_ghz, common_comm) + f_handle.write(string_buffer); + f_handle.close() + except: + print('IO error cpu.csv') + return + + graph_data_present = True; + +def split_csv(): + """ seperate the all csv file into per CPU csv files. """ + + global current_max_cpu + + if os.path.exists('cpu.csv'): + for index in range(0, current_max_cpu + 1): + if cpu_mask[int(index)] != 0: + os.system('grep -m 1 common_cpu cpu.csv > cpu{:0>3}.csv'.format(index)) + os.system('grep CPU_{:0>3} cpu.csv >> cpu{:0>3}.csv'.format(index, index)) + +def fix_ownership(path): + """Change the owner of the file to SUDO_UID, if required""" + + uid = os.environ.get('SUDO_UID') + gid = os.environ.get('SUDO_GID') + if uid is not None: + os.chown(path, int(uid), int(gid)) + +def cleanup_data_files(): + """ clean up existing data files """ + + if os.path.exists('cpu.csv'): + os.remove('cpu.csv') + f_handle = open('cpu.csv', 'a') + f_handle.write('common_cpu, common_secs, common_usecs, core_busy, scaled_busy, from, to, mperf, aperf, tsc, freq, boost, load, duration_ms, sample_num, elapsed_time, tsc_ghz, common_comm') + f_handle.write('\n') + f_handle.close() + +def clear_trace_file(): + """ Clear trace file """ + + try: + f_handle = open('/sys/kernel/debug/tracing/trace', 'w') + f_handle.close() + except: + print('IO error clearing trace file ') + quit() + +def enable_trace(): + """ Enable trace """ + + try: + open('/sys/kernel/debug/tracing/events/power/pstate_sample/enable' + , 'w').write("1") + except: + print('IO error enabling trace ') + quit() + +def disable_trace(): + """ Disable trace """ + + try: + open('/sys/kernel/debug/tracing/events/power/pstate_sample/enable' + , 'w').write("0") + except: + print('IO error disabling trace ') + quit() + +def set_trace_buffer_size(): + """ Set trace buffer size """ + + try: + open('/sys/kernel/debug/tracing/buffer_size_kb' + , 'w').write("10240") + except: + print('IO error setting trace buffer size ') + quit() + +def read_trace_data(filename): + """ Read and parse trace data """ + + global current_max_cpu + global sample_num, last_sec_cpu, last_usec_cpu, start_time + + try: + data = open(filename, 'r').read() + except: + print('Error opening ', filename) + quit() + + for line in data.splitlines(): + search_obj = \ + re.search(r'(^(.*?)\[)((\d+)[^\]])(.*?)(\d+)([.])(\d+)(.*?core_busy=)(\d+)(.*?scaled=)(\d+)(.*?from=)(\d+)(.*?to=)(\d+)(.*?mperf=)(\d+)(.*?aperf=)(\d+)(.*?tsc=)(\d+)(.*?freq=)(\d+)' + , line) + + if search_obj: + cpu = search_obj.group(3) + cpu_int = int(cpu) + cpu = str(cpu_int) + + time_pre_dec = search_obj.group(6) + time_post_dec = search_obj.group(8) + core_busy = search_obj.group(10) + scaled = search_obj.group(12) + _from = search_obj.group(14) + _to = search_obj.group(16) + mperf = search_obj.group(18) + aperf = search_obj.group(20) + tsc = search_obj.group(22) + freq = search_obj.group(24) + common_comm = search_obj.group(2).replace(' ', '') + + # Not all kernel versions have io_boost field + io_boost = '0' + search_obj = re.search(r'.*?io_boost=(\d+)', line) + if search_obj: + io_boost = search_obj.group(1) + + if sample_num == 0 : + start_time = Decimal(time_pre_dec) + Decimal(time_post_dec) / Decimal(1000000) + sample_num += 1 + + if last_sec_cpu[cpu_int] == 0 : + last_sec_cpu[cpu_int] = time_pre_dec + last_usec_cpu[cpu_int] = time_post_dec + else : + duration_us = (int(time_pre_dec) - int(last_sec_cpu[cpu_int])) * 1000000 + (int(time_post_dec) - int(last_usec_cpu[cpu_int])) + duration_ms = Decimal(duration_us) / Decimal(1000) + last_sec_cpu[cpu_int] = time_pre_dec + last_usec_cpu[cpu_int] = time_post_dec + elapsed_time = Decimal(time_pre_dec) + Decimal(time_post_dec) / Decimal(1000000) - start_time + load = Decimal(int(mperf)*100)/ Decimal(tsc) + freq_ghz = Decimal(freq)/Decimal(1000000) +# Sanity check calculation, typically anomalies indicate missed samples +# However, check for 0 (should never occur) + tsc_ghz = Decimal(0) + if duration_ms != Decimal(0) : + tsc_ghz = Decimal(tsc)/duration_ms/Decimal(1000000) + store_csv(cpu_int, time_pre_dec, time_post_dec, core_busy, scaled, _from, _to, mperf, aperf, tsc, freq_ghz, io_boost, common_comm, load, duration_ms, sample_num, elapsed_time, tsc_ghz) + + if cpu_int > current_max_cpu: + current_max_cpu = cpu_int +# End of for each trace line loop +# Now seperate the main overall csv file into per CPU csv files. + split_csv() + +interval = "" +filename = "" +cpu_list = "" +testname = "" +graph_data_present = False; + +valid1 = False +valid2 = False + +cpu_mask = zeros((MAX_CPUS,), dtype=int) + +try: + opts, args = getopt.getopt(sys.argv[1:],"ht:i:c:n:",["help","trace_file=","interval=","cpu=","name="]) +except getopt.GetoptError: + print_help() + sys.exit(2) +for opt, arg in opts: + if opt == '-h': + print() + sys.exit() + elif opt in ("-t", "--trace_file"): + valid1 = True + location = os.path.realpath(os.path.join(os.getcwd(), os.path.dirname(__file__))) + filename = os.path.join(location, arg) + elif opt in ("-i", "--interval"): + valid1 = True + interval = arg + elif opt in ("-c", "--cpu"): + cpu_list = arg + elif opt in ("-n", "--name"): + valid2 = True + testname = arg + +if not (valid1 and valid2): + print_help() + sys.exit() + +if cpu_list: + for p in re.split("[,]", cpu_list): + if int(p) < MAX_CPUS : + cpu_mask[int(p)] = 1 +else: + for i in range (0, MAX_CPUS): + cpu_mask[i] = 1 + +if not os.path.exists('results'): + os.mkdir('results') + # The regular user needs to own the directory, not root. + fix_ownership('results') + +os.chdir('results') +if os.path.exists(testname): + print('The test name directory already exists. Please provide a unique test name. Test re-run not supported, yet.') + sys.exit() +os.mkdir(testname) +# The regular user needs to own the directory, not root. +fix_ownership(testname) +os.chdir(testname) + +# Temporary (or perhaps not) +cur_version = sys.version_info +print('python version (should be >= 2.7):') +print(cur_version) + +# Left as "cleanup" for potential future re-run ability. +cleanup_data_files() + +if interval: + filename = "/sys/kernel/debug/tracing/trace" + clear_trace_file() + set_trace_buffer_size() + enable_trace() + print('Sleeping for ', interval, 'seconds') + time.sleep(int(interval)) + disable_trace() + +current_max_cpu = 0 + +read_trace_data(filename) + +if graph_data_present == False: + print('No valid data to plot') + sys.exit(2) + +for cpu_no in range(0, current_max_cpu + 1): + plot_perf_busy_with_sample(cpu_no) + plot_perf_busy(cpu_no) + plot_durations(cpu_no) + plot_loads(cpu_no) + +plot_pstate_cpu_with_sample() +plot_pstate_cpu() +plot_load_cpu() +plot_frequency_cpu() +plot_duration_cpu() +plot_scaled_cpu() +plot_boost_cpu() +plot_ghz_cpu() + +# It is preferrable, but not necessary, that the regular user owns the files, not root. +for root, dirs, files in os.walk('.'): + for f in files: + fix_ownership(f) + +os.chdir('../../') diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index 03cb639b292e..ccf2a69365cc 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -16,9 +16,9 @@ idle power-state statistics, temperature and power on X86 processors. There are two ways to invoke turbostat. The first method is to supply a \fBcommand\fP, which is forked and statistics are printed -upon its completion. +in one-shot upon its completion. The second method is to omit the command, -and turbostat displays statistics every 5 seconds. +and turbostat displays statistics every 5 seconds interval. The 5-second interval can be changed using the --interval option. .PP Some information is not available on older processors. @@ -28,9 +28,10 @@ name as necessary to disambiguate it from others is necessary. Note that option .PP \fB--add attributes\fP add column with counter having specified 'attributes'. The 'location' attribute is required, all others are optional. .nf - location: {\fBmsrDDD\fP | \fBmsr0xXXX\fP} + location: {\fBmsrDDD\fP | \fBmsr0xXXX\fP | \fB/sys/path...\fP} msrDDD is a decimal offset, eg. msr16 msr0xXXX is a hex offset, eg. msr0x10 + /sys/path... is an absolute path to a sysfs attribute scope: {\fBcpu\fP | \fBcore\fP | \fBpackage\fP} sample and print the counter for every cpu, core, or package. @@ -45,12 +46,21 @@ name as necessary to disambiguate it from others is necessary. Note that option 'delta' shows the difference in values during the measurement interval. 'percent' shows the delta as a percentage of the cycles elapsed. default: delta + + name: "name_string" + Any string that does not match a key-word above is used + as the column header. .fi .PP +\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 +.PP +\fB--hide column\fP do not show the specified 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. +.PP +\fB--show column\fP show only the specified 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. +.PP \fB--Dump\fP displays the raw counter values. .PP -\fB--debug\fP displays additional system configuration information. Invoking this parameter -more than once may also enable internal turbostat debug information. +\fB--quiet\fP Do not decode and print the system configuration header information. .PP \fB--interval seconds\fP overrides the default 5.0 second measurement interval. .PP @@ -61,9 +71,7 @@ The file is truncated if it already exists, and it is created if it does not exi .PP \fB--Joules\fP displays energy in Joules, rather than dividing Joules by time to print power in Watts. .PP -\fB--Package\fP limits output to the system summary plus the 1st thread in each Package. -.PP -\fB--processor\fP limits output to the system summary plus the 1st thread in each processor of each package. Ie. it skips hyper-threaded siblings. +\fB--list\fP display column header names available for use by --show and --hide, then exit. .PP \fB--Summary\fP limits output to a 1-line System Summary for each interval. .PP @@ -74,24 +82,27 @@ The file is truncated if it already exists, and it is created if it does not exi The \fBcommand\fP parameter forks \fBcommand\fP, and upon its exit, displays the statistics gathered since it was forked. .PP -.SH DEFAULT FIELD DESCRIPTIONS +.SH ROW DESCRIPTIONS +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. +.SH COLUMN DESCRIPTIONS .nf +\fBCore\fP processor core number. Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology (HT). \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. -\fBAVG_MHz\fP number of cycles executed divided by time elapsed. -\fBBusy%\fP percent of the interval that the CPU retired instructions, aka. % of time in "C0" state. -\fBBzy_MHz\fP average clock rate while the CPU was busy (in "c0" state). +\fBPackage\fP processor package number -- not present on systems with a single processor package. +\fBAvg_MHz\fP number of cycles executed divided by time elapsed. Note that this includes idle-time when 0 instructions are executed. +\fBBusy%\fP percent of the measurement interval that the CPU executes instructions, aka. % of time in "C0" state. +\fBBzy_MHz\fP average clock rate while the CPU was not idle (ie. in "c0" state). \fBTSC_MHz\fP average MHz that the TSC ran during the entire interval. -.fi -.PP -.SH DEBUG FIELD DESCRIPTIONS -.nf -\fBPackage\fP processor package number. -\fBCore\fP processor core number. -Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading Technology (HT). -\fBCPU%c1, CPU%c3, CPU%c6, CPU%c7\fP show the percentage residency in hardware core idle states. +\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. +\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. +\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. +\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. +\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. \fBCoreTmp\fP Degrees Celsius reported by the per-core Digital Thermal Sensor. \fBPkgTtmp\fP Degrees Celsius reported by the per-package Package Thermal Monitor. -\fBPkg%pc2, Pkg%pc3, Pkg%pc6, Pkg%pc7\fP percentage residency in hardware package idle states. +\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. +\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. +\fBPkg%pc2, Pkg%pc3, Pkg%pc6, Pkg%pc7\fP percentage residency in hardware package idle states. These numbers are from hardware residency counters. \fBPkgWatt\fP Watts consumed by the whole package. \fBCorWatt\fP Watts consumed by the core part of the package. \fBGFXWatt\fP Watts consumed by the Graphics part of the package -- available only on client processors. @@ -99,51 +110,110 @@ Note that multiple CPUs per core indicate support for Intel(R) Hyper-Threading T \fBPKG_%\fP percent of the interval that RAPL throttling was active on the Package. \fBRAM_%\fP percent of the interval that RAPL throttling was active on DRAM. .fi +.SH TOO MUCH INFORMATION EXAMPLE +By default, turbostat dumps all possible information -- a system configuration header, followed by columns for all counters. +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. .PP -.SH PERIODIC EXAMPLE -Without any parameters, turbostat displays statistics ever 5 seconds. -Periodic output goes to stdout, by default, unless --out is used to specify an output file. -The 5-second interval can be changed with th "-i sec" option. -Or a command may be specified as in "FORK EXAMPLE" below. +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. +.PP +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: +.nf +sudo ./turbostat --show sysfs --quiet sleep 10 +10.003837 sec + C1 C1E C3 C6 C7s C1% C1E% C3% C6% C7s% + 4 21 2 2 459 0.14 0.82 0.00 0.00 98.93 + 1 17 2 2 130 0.00 0.02 0.00 0.00 99.80 + 0 0 0 0 31 0.00 0.00 0.00 0.00 99.95 + 2 1 0 0 52 1.14 6.49 0.00 0.00 92.21 + 1 2 0 0 52 0.00 0.08 0.00 0.00 99.86 + 0 0 0 0 71 0.00 0.00 0.00 0.00 99.89 + 0 0 0 0 25 0.00 0.00 0.00 0.00 99.96 + 0 0 0 0 74 0.00 0.00 0.00 0.00 99.94 + 0 1 0 0 24 0.00 0.00 0.00 0.00 99.84 +.fi +.PP +.SH ONE SHOT COMMAND EXAMPLE +If turbostat is invoked with a command, it will fork that command +and output the statistics gathered after the command exits. +In this case, turbostat output goes to stderr, by default. +Output can instead be saved to a file using the --out option. +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. + .nf -[root@hsw]# ./turbostat - CPU Avg_MHz Busy% Bzy_MHz TSC_MHz - - 488 12.51 3898 3498 - 0 0 0.01 3885 3498 - 4 3897 99.99 3898 3498 - 1 0 0.00 3861 3498 - 5 0 0.00 3882 3498 - 2 1 0.02 3894 3498 - 6 2 0.06 3898 3498 - 3 0 0.00 3849 3498 - 7 0 0.00 3877 3498 +[root@hsw]# ./turbostat -o ts.out sleep 10 +[root@hsw]# +.fi +.SH PERIODIC INTERVAL EXAMPLE +Without a command to fork, turbostat displays statistics ever 5 seconds. +Periodic output goes to stdout, by default, unless --out is used to specify an output file. +The 5-second interval can be changed with the "-i sec" option. +.nf +sudo ./turbostat --quiet --hide sysfs,IRQ,SMI,CoreTmp,PkgTmp,GFX%rc6,GFXMHz,PkgWatt,CorWatt,GFXWatt + Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7 + - - 488 12.52 3900 3498 12.50 0.00 0.00 74.98 + 0 0 5 0.13 3900 3498 99.87 0.00 0.00 0.00 + 0 4 3897 99.99 3900 3498 0.01 + 1 1 0 0.00 3856 3498 0.01 0.00 0.00 99.98 + 1 5 0 0.00 3861 3498 0.01 + 2 2 1 0.02 3889 3498 0.03 0.00 0.00 99.95 + 2 6 0 0.00 3863 3498 0.05 + 3 3 0 0.01 3869 3498 0.02 0.00 0.00 99.97 + 3 7 0 0.00 3878 3498 0.03 + Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz CPU%c1 CPU%c3 CPU%c6 CPU%c7 + - - 491 12.59 3900 3498 12.42 0.00 0.00 74.99 + 0 0 27 0.69 3900 3498 99.31 0.00 0.00 0.00 + 0 4 3898 99.99 3900 3498 0.01 + 1 1 0 0.00 3883 3498 0.01 0.00 0.00 99.99 + 1 5 0 0.00 3898 3498 0.01 + 2 2 0 0.01 3889 3498 0.02 0.00 0.00 99.98 + 2 6 0 0.00 3889 3498 0.02 + 3 3 0 0.00 3856 3498 0.01 0.00 0.00 99.99 + 3 7 0 0.00 3897 3498 0.01 .fi -.SH DEBUG EXAMPLE -The "--debug" option prints additional system information before measurements: +This example also shows the use of the --hide option to skip columns that are not wanted. +Note that cpu4 in this example is 99.99% busy, while the other CPUs are all under 1% busy. +Notice that cpu4's HT sibling is cpu0, which is under 1% busy, but can get into CPU%c1 only, +because its cpu4's activity on shared hardware keeps it from entering a deeper C-state. -The first row of statistics is a summary for the entire system. -For residency % columns, the summary is a weighted average. -For Temperature columns, the summary is the column maximum. -For Watts columns, the summary is a system total. -Subsequent rows show per-CPU statistics. +.SH SYSTEM CONFIGURATION INFORMATION EXAMPLE + +By default, turbostat always dumps system configuration information +before taking measurements. In the example above, "--quiet" is used +to suppress that output. Here is an example of the configuration information: .nf -turbostat version 4.1 10-Feb, 2015 - Len Brown <lenb@kernel.org> +turbostat version 2017.02.15 - Len Brown <lenb@kernel.org> CPUID(0): GenuineIntel 13 CPUID levels; family:model:stepping 0x6:3c:3 (6:60:3) -CPUID(6): APERF, DTS, PTM, EPB +CPUID(1): SSE3 MONITOR - EIST TM2 TSC MSR ACPI-TM TM +CPUID(6): APERF, TURBO, DTS, PTM, No-HWP, No-HWPnotify, No-HWPwindow, No-HWPepp, No-HWPpkg, EPB +cpu4: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST No-MWAIT PREFETCH TURBO) +CPUID(7): No-SGX +cpu4: MSR_MISC_PWR_MGMT: 0x00400000 (ENable-EIST_Coordination DISable-EPB DISable-OOB) RAPL: 3121 sec. Joule Counter Range, at 84 Watts -cpu0: MSR_NHM_PLATFORM_INFO: 0x80838f3012300 -8 * 100 = 800 MHz max efficiency -35 * 100 = 3500 MHz TSC frequency -cpu0: MSR_IA32_POWER_CTL: 0x0004005d (C1E auto-promotion: DISabled) -cpu0: MSR_NHM_SNB_PKG_CST_CFG_CTL: 0x1e000400 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, UNlocked: pkg-cstate-limit=0: pc0) -cpu0: MSR_TURBO_RATIO_LIMIT: 0x25262727 -37 * 100 = 3700 MHz max turbo 4 active cores -38 * 100 = 3800 MHz max turbo 3 active cores -39 * 100 = 3900 MHz max turbo 2 active cores -39 * 100 = 3900 MHz max turbo 1 active cores +cpu4: MSR_PLATFORM_INFO: 0x80838f3012300 +8 * 100.0 = 800.0 MHz max efficiency frequency +35 * 100.0 = 3500.0 MHz base frequency +cpu4: MSR_IA32_POWER_CTL: 0x0004005d (C1E auto-promotion: DISabled) +cpu4: MSR_TURBO_RATIO_LIMIT: 0x25262727 +37 * 100.0 = 3700.0 MHz max turbo 4 active cores +38 * 100.0 = 3800.0 MHz max turbo 3 active cores +39 * 100.0 = 3900.0 MHz max turbo 2 active cores +39 * 100.0 = 3900.0 MHz max turbo 1 active cores +cpu4: MSR_CONFIG_TDP_NOMINAL: 0x00000023 (base_ratio=35) +cpu4: MSR_CONFIG_TDP_LEVEL_1: 0x00000000 () +cpu4: MSR_CONFIG_TDP_LEVEL_2: 0x00000000 () +cpu4: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1) +cpu4: MSR_TURBO_ACTIVATION_RATIO: 0x00000000 (MAX_NON_TURBO_RATIO=0 lock=0) +cpu4: MSR_PKG_CST_CONFIG_CONTROL: 0x1e000400 (UNdemote-C3, UNdemote-C1, demote-C3, demote-C1, UNlocked: pkg-cstate-limit=0: pc0) +cpu4: POLL: CPUIDLE CORE POLL IDLE +cpu4: C1: MWAIT 0x00 +cpu4: C1E: MWAIT 0x01 +cpu4: C3: MWAIT 0x10 +cpu4: C6: MWAIT 0x20 +cpu4: C7s: MWAIT 0x32 +cpu4: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2-Prefetch L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch) cpu0: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced) -cpu0: MSR_CORE_PERF_LIMIT_REASONS, 0x31200000 (Active: ) (Logged: Auto-HWP, Amps, MultiCoreTurbo, Transitions, ) +cpu0: MSR_CORE_PERF_LIMIT_REASONS, 0x31200000 (Active: ) (Logged: Transitions, MultiCoreTurbo, Amps, Auto-HWP, ) cpu0: MSR_GFX_PERF_LIMIT_REASONS, 0x00000000 (Active: ) (Logged: ) cpu0: MSR_RING_PERF_LIMIT_REASONS, 0x0d000000 (Active: ) (Logged: Amps, PkgPwrL1, PkgPwrL2, ) cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061 Joules, 0.000977 sec.) @@ -158,23 +228,14 @@ cpu0: MSR_PP1_POLICY: 0 cpu0: MSR_PP1_POWER_LIMIT: 0x00000000 (UNlocked) cpu0: GFX Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled) cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00641400 (100 C) -cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x88340800 (48 C) -cpu0: MSR_IA32_THERM_STATUS: 0x88340000 (48 C +/- 1) -cpu1: MSR_IA32_THERM_STATUS: 0x88440000 (32 C +/- 1) -cpu2: MSR_IA32_THERM_STATUS: 0x88450000 (31 C +/- 1) -cpu3: MSR_IA32_THERM_STATUS: 0x88490000 (27 C +/- 1) - Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz SMI CPU%c1 CPU%c3 CPU%c6 CPU%c7 CoreTmp PkgTmp PkgWatt CorWatt GFXWatt - - - 493 12.64 3898 3498 0 12.64 0.00 0.00 74.72 47 47 21.62 13.74 0.00 - 0 0 4 0.11 3894 3498 0 99.89 0.00 0.00 0.00 47 47 21.62 13.74 0.00 - 0 4 3897 99.98 3898 3498 0 0.02 - 1 1 7 0.17 3887 3498 0 0.04 0.00 0.00 99.79 32 - 1 5 0 0.00 3885 3498 0 0.21 - 2 2 29 0.76 3895 3498 0 0.10 0.01 0.01 99.13 32 - 2 6 2 0.06 3896 3498 0 0.80 - 3 3 1 0.02 3832 3498 0 0.03 0.00 0.00 99.95 28 - 3 7 0 0.00 3879 3498 0 0.04 -^C - +cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x884c0800 (24 C) +cpu0: MSR_IA32_THERM_STATUS: 0x884c0000 (24 C +/- 1) +cpu1: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1) +cpu2: MSR_IA32_THERM_STATUS: 0x884e0000 (22 C +/- 1) +cpu3: MSR_IA32_THERM_STATUS: 0x88510000 (19 C +/- 1) +cpu4: MSR_PKGC3_IRTL: 0x00008842 (valid, 67584 ns) +cpu4: MSR_PKGC6_IRTL: 0x00008873 (valid, 117760 ns) +cpu4: MSR_PKGC7_IRTL: 0x00008891 (valid, 148480 ns) .fi The \fBmax efficiency\fP frequency, a.k.a. Low Frequency Mode, is the frequency available at the minimum package voltage. The \fBTSC frequency\fP is the base @@ -184,42 +245,22 @@ should be sustainable on all CPUs indefinitely, given nominal power and cooling. The remaining rows show what maximum turbo frequency is possible depending on the number of idle cores. Note that not all information is available on all processors. -.PP -The --debug option adds additional columns to the measurement ouput, including CPU idle power-state residency processor temperature sensor readinds. -See the field definitions above. -.SH FORK EXAMPLE -If turbostat is invoked with a command, it will fork that command -and output the statistics gathered after the command exits. -In this case, turbostat output goes to stderr, by default. -Output can instead be saved to a file using the --out option. -eg. Here a cycle soaker is run on 1 CPU (see %c0) for a few seconds -until ^C while the other CPUs are mostly idle: - +.SH ADD COUNTER EXAMPLE +Here we limit turbostat to showing just the CPU number for cpu0 - cpu3. +We add a counter showing the 32-bit raw value of MSR 0x199 (MSR_IA32_PERF_CTL), +labeling it with the column header, "PRF_CTRL", and display it only once, +afte the conclusion of a 0.1 second sleep. .nf -root@hsw: turbostat cat /dev/zero > /dev/null -^C - CPU Avg_MHz Busy% Bzy_MHz TSC_MHz - - 482 12.51 3854 3498 - 0 0 0.01 1960 3498 - 4 0 0.00 2128 3498 - 1 0 0.00 3003 3498 - 5 3854 99.98 3855 3498 - 2 0 0.01 3504 3498 - 6 3 0.08 3884 3498 - 3 0 0.00 2553 3498 - 7 0 0.00 2126 3498 -10.783983 sec +sudo ./turbostat --quiet --cpu 0-3 --show CPU --add msr0x199,u32,raw,PRF_CTRL sleep .1 +0.101604 sec +CPU PRF_CTRL +- 0x00000000 +0 0x00000c00 +1 0x00000800 +2 0x00000a00 +3 0x00000800 .fi -Above the cycle soaker drives cpu5 up its 3.9 GHz turbo limit. -The first row shows the average MHz and Busy% across all the processors in the system. - -Note that the Avg_MHz column reflects the total number of cycles executed -divided by the measurement interval. If the Busy% column is 100%, -then the processor was running at that speed the entire interval. -The Avg_MHz multiplied by the Busy% results in the Bzy_MHz -- -which is the average frequency while the processor was executing -- -not including any non-busy idle time. .SH NOTES diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index f13f61b065c6..b11294730771 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -49,17 +49,14 @@ FILE *outf; int *fd_percpu; struct timespec interval_ts = {5, 0}; unsigned int debug; +unsigned int quiet; +unsigned int sums_need_wide_columns; unsigned int rapl_joules; unsigned int summary_only; +unsigned int list_header_only; unsigned int dump_only; -unsigned int do_nhm_cstates; unsigned int do_snb_cstates; unsigned int do_knl_cstates; -unsigned int do_pc2; -unsigned int do_pc3; -unsigned int do_pc6; -unsigned int do_pc7; -unsigned int do_c8_c9_c10; unsigned int do_skl_residency; unsigned int do_slm_cstates; unsigned int use_c1_residency_msr; @@ -71,25 +68,19 @@ unsigned int units = 1000000; /* MHz etc */ unsigned int genuine_intel; unsigned int has_invariant_tsc; unsigned int do_nhm_platform_info; +unsigned int no_MSR_MISC_PWR_MGMT; unsigned int aperf_mperf_multiplier = 1; -int do_irq = 1; -int do_smi; double bclk; double base_hz; unsigned int has_base_hz; double tsc_tweak = 1.0; -unsigned int show_pkg; -unsigned int show_core; -unsigned int show_cpu; unsigned int show_pkg_only; unsigned int show_core_only; char *output_buffer, *outp; unsigned int do_rapl; unsigned int do_dts; unsigned int do_ptm; -unsigned int do_gfx_rc6_ms; unsigned long long gfx_cur_rc6_ms; -unsigned int do_gfx_mhz; unsigned int gfx_cur_mhz; unsigned int tcc_activation_temp; unsigned int tcc_activation_temp_override; @@ -109,6 +100,7 @@ unsigned int has_hwp_notify; /* IA32_HWP_INTERRUPT */ unsigned int has_hwp_activity_window; /* IA32_HWP_REQUEST[bits 41:32] */ unsigned int has_hwp_epp; /* IA32_HWP_REQUEST[bits 31:24] */ unsigned int has_hwp_pkg; /* IA32_HWP_REQUEST_PKG */ +unsigned int has_misc_feature_control; #define RAPL_PKG (1 << 0) /* 0x610 MSR_PKG_POWER_LIMIT */ @@ -148,34 +140,38 @@ unsigned int has_hwp_pkg; /* IA32_HWP_REQUEST_PKG */ * Usually truncated to 7 characters, but also handles 18 columns for raw 64-bit counters */ #define NAME_BYTES 20 +#define PATH_BYTES 128 int backwards_count; char *progname; -cpu_set_t *cpu_present_set, *cpu_affinity_set; -size_t cpu_present_setsize, cpu_affinity_setsize; +#define CPU_SUBSET_MAXCPUS 1024 /* need to use before probe... */ +cpu_set_t *cpu_present_set, *cpu_affinity_set, *cpu_subset; +size_t cpu_present_setsize, cpu_affinity_setsize, cpu_subset_size; +#define MAX_ADDED_COUNTERS 16 struct thread_data { unsigned long long tsc; unsigned long long aperf; unsigned long long mperf; unsigned long long c1; - unsigned int irq_count; + unsigned long long irq_count; unsigned int smi_count; unsigned int cpu_id; unsigned int flags; #define CPU_IS_FIRST_THREAD_IN_CORE 0x2 #define CPU_IS_FIRST_CORE_IN_PACKAGE 0x4 - unsigned long long counter[1]; + unsigned long long counter[MAX_ADDED_COUNTERS]; } *thread_even, *thread_odd; struct core_data { unsigned long long c3; unsigned long long c6; unsigned long long c7; + unsigned long long mc6_us; /* duplicate as per-core for now, even though per module */ unsigned int core_temp_c; unsigned int core_id; - unsigned long long counter[1]; + unsigned long long counter[MAX_ADDED_COUNTERS]; } *core_even, *core_odd; struct pkg_data { @@ -200,7 +196,7 @@ struct pkg_data { unsigned int rapl_pkg_perf_status; /* MSR_PKG_PERF_STATUS */ unsigned int rapl_dram_perf_status; /* MSR_DRAM_PERF_STATUS */ unsigned int pkg_temp_c; - unsigned long long counter[1]; + unsigned long long counter[MAX_ADDED_COUNTERS]; } *package_even, *package_odd; #define ODD_COUNTERS thread_odd, core_odd, package_odd @@ -215,22 +211,27 @@ struct pkg_data { #define GET_PKG(pkg_base, pkg_no) (pkg_base + pkg_no) enum counter_scope {SCOPE_CPU, SCOPE_CORE, SCOPE_PACKAGE}; -enum counter_type {COUNTER_CYCLES, COUNTER_SECONDS}; +enum counter_type {COUNTER_ITEMS, COUNTER_CYCLES, COUNTER_SECONDS, COUNTER_USEC}; enum counter_format {FORMAT_RAW, FORMAT_DELTA, FORMAT_PERCENT}; struct msr_counter { unsigned int msr_num; char name[NAME_BYTES]; + char path[PATH_BYTES]; unsigned int width; enum counter_type type; enum counter_format format; struct msr_counter *next; + unsigned int flags; +#define FLAGS_HIDE (1 << 0) +#define FLAGS_SHOW (1 << 1) +#define SYSFS_PERCPU (1 << 1) }; struct sys_counters { - unsigned int thread_counter_bytes; - unsigned int core_counter_bytes; - unsigned int package_counter_bytes; + unsigned int added_thread_counters; + unsigned int added_core_counters; + unsigned int added_package_counters; struct msr_counter *tp; struct msr_counter *cp; struct msr_counter *pp; @@ -334,147 +335,333 @@ int get_msr(int cpu, off_t offset, unsigned long long *msr) retval = pread(get_msr_fd(cpu), msr, sizeof(*msr), offset); if (retval != sizeof *msr) - err(-1, "msr %d offset 0x%llx read failed", cpu, (unsigned long long)offset); + err(-1, "cpu%d: msr offset 0x%llx read failed", cpu, (unsigned long long)offset); return 0; } /* - * Example Format w/ field column widths: - * - * Package Core CPU Avg_MHz Bzy_MHz TSC_MHz IRQ SMI Busy% CPU_%c1 CPU_%c3 CPU_%c6 CPU_%c7 ThreadC CoreTmp CoreCnt PkgTmp GFXMHz Pkg%pc2 Pkg%pc3 Pkg%pc6 Pkg%pc7 PkgWatt CorWatt GFXWatt PkgCnt - * 12345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678123456781234567812345678 + * Each string in this array is compared in --show and --hide cmdline. + * Thus, strings that are proper sub-sets must follow their more specific peers. + */ +struct msr_counter bic[] = { + { 0x0, "Package" }, + { 0x0, "Avg_MHz" }, + { 0x0, "Bzy_MHz" }, + { 0x0, "TSC_MHz" }, + { 0x0, "IRQ" }, + { 0x0, "SMI", "", 32, 0, FORMAT_DELTA, NULL}, + { 0x0, "Busy%" }, + { 0x0, "CPU%c1" }, + { 0x0, "CPU%c3" }, + { 0x0, "CPU%c6" }, + { 0x0, "CPU%c7" }, + { 0x0, "ThreadC" }, + { 0x0, "CoreTmp" }, + { 0x0, "CoreCnt" }, + { 0x0, "PkgTmp" }, + { 0x0, "GFX%rc6" }, + { 0x0, "GFXMHz" }, + { 0x0, "Pkg%pc2" }, + { 0x0, "Pkg%pc3" }, + { 0x0, "Pkg%pc6" }, + { 0x0, "Pkg%pc7" }, + { 0x0, "Pkg%pc8" }, + { 0x0, "Pkg%pc9" }, + { 0x0, "Pkg%pc10" }, + { 0x0, "PkgWatt" }, + { 0x0, "CorWatt" }, + { 0x0, "GFXWatt" }, + { 0x0, "PkgCnt" }, + { 0x0, "RAMWatt" }, + { 0x0, "PKG_%" }, + { 0x0, "RAM_%" }, + { 0x0, "Pkg_J" }, + { 0x0, "Cor_J" }, + { 0x0, "GFX_J" }, + { 0x0, "RAM_J" }, + { 0x0, "Core" }, + { 0x0, "CPU" }, + { 0x0, "Mod%c6" }, + { 0x0, "sysfs" }, +}; + +#define MAX_BIC (sizeof(bic) / sizeof(struct msr_counter)) +#define BIC_Package (1ULL << 0) +#define BIC_Avg_MHz (1ULL << 1) +#define BIC_Bzy_MHz (1ULL << 2) +#define BIC_TSC_MHz (1ULL << 3) +#define BIC_IRQ (1ULL << 4) +#define BIC_SMI (1ULL << 5) +#define BIC_Busy (1ULL << 6) +#define BIC_CPU_c1 (1ULL << 7) +#define BIC_CPU_c3 (1ULL << 8) +#define BIC_CPU_c6 (1ULL << 9) +#define BIC_CPU_c7 (1ULL << 10) +#define BIC_ThreadC (1ULL << 11) +#define BIC_CoreTmp (1ULL << 12) +#define BIC_CoreCnt (1ULL << 13) +#define BIC_PkgTmp (1ULL << 14) +#define BIC_GFX_rc6 (1ULL << 15) +#define BIC_GFXMHz (1ULL << 16) +#define BIC_Pkgpc2 (1ULL << 17) +#define BIC_Pkgpc3 (1ULL << 18) +#define BIC_Pkgpc6 (1ULL << 19) +#define BIC_Pkgpc7 (1ULL << 20) +#define BIC_Pkgpc8 (1ULL << 21) +#define BIC_Pkgpc9 (1ULL << 22) +#define BIC_Pkgpc10 (1ULL << 23) +#define BIC_PkgWatt (1ULL << 24) +#define BIC_CorWatt (1ULL << 25) +#define BIC_GFXWatt (1ULL << 26) +#define BIC_PkgCnt (1ULL << 27) +#define BIC_RAMWatt (1ULL << 28) +#define BIC_PKG__ (1ULL << 29) +#define BIC_RAM__ (1ULL << 30) +#define BIC_Pkg_J (1ULL << 31) +#define BIC_Cor_J (1ULL << 32) +#define BIC_GFX_J (1ULL << 33) +#define BIC_RAM_J (1ULL << 34) +#define BIC_Core (1ULL << 35) +#define BIC_CPU (1ULL << 36) +#define BIC_Mod_c6 (1ULL << 37) +#define BIC_sysfs (1ULL << 38) + +unsigned long long bic_enabled = 0xFFFFFFFFFFFFFFFFULL; +unsigned long long bic_present = BIC_sysfs; + +#define DO_BIC(COUNTER_NAME) (bic_enabled & bic_present & COUNTER_NAME) +#define BIC_PRESENT(COUNTER_BIT) (bic_present |= COUNTER_BIT) +#define BIC_NOT_PRESENT(COUNTER_BIT) (bic_present &= ~COUNTER_BIT) + +#define MAX_DEFERRED 16 +char *deferred_skip_names[MAX_DEFERRED]; +int deferred_skip_index; + +/* + * HIDE_LIST - hide this list of counters, show the rest [default] + * SHOW_LIST - show this list of counters, hide the rest */ +enum show_hide_mode { SHOW_LIST, HIDE_LIST } global_show_hide_mode = HIDE_LIST; -void print_header(void) +void help(void) { - struct msr_counter *mp; + fprintf(outf, + "Usage: turbostat [OPTIONS][(--interval seconds) | COMMAND ...]\n" + "\n" + "Turbostat forks the specified COMMAND and prints statistics\n" + "when COMMAND completes.\n" + "If no COMMAND is specified, turbostat wakes every 5-seconds\n" + "to print statistics, until interrupted.\n" + "--add add a counter\n" + " eg. --add msr0x10,u64,cpu,delta,MY_TSC\n" + "--cpu cpu-set limit output to summary plus cpu-set:\n" + " {core | package | j,k,l..m,n-p }\n" + "--quiet skip decoding system configuration header\n" + "--interval sec Override default 5-second measurement interval\n" + "--help print this help message\n" + "--list list column headers only\n" + "--out file create or truncate \"file\" for all output\n" + "--version print version information\n" + "\n" + "For more help, run \"man turbostat\"\n"); +} - if (show_pkg) - outp += sprintf(outp, "\tPackage"); - if (show_core) - outp += sprintf(outp, "\tCore"); - if (show_cpu) - outp += sprintf(outp, "\tCPU"); - if (has_aperf) - outp += sprintf(outp, "\tAvg_MHz"); - if (has_aperf) - outp += sprintf(outp, "\tBusy%%"); - if (has_aperf) - outp += sprintf(outp, "\tBzy_MHz"); - outp += sprintf(outp, "\tTSC_MHz"); +/* + * bic_lookup + * for all the strings in comma separate name_list, + * set the approprate bit in return value. + */ +unsigned long long bic_lookup(char *name_list, enum show_hide_mode mode) +{ + int i; + unsigned long long retval = 0; - if (!debug) - goto done; + while (name_list) { + char *comma; - if (do_irq) - outp += sprintf(outp, "\tIRQ"); - if (do_smi) - outp += sprintf(outp, "\tSMI"); - - if (do_nhm_cstates) - outp += sprintf(outp, "\tCPU%%c1"); - if (do_nhm_cstates && !do_slm_cstates && !do_knl_cstates) - outp += sprintf(outp, "\tCPU%%c3"); - if (do_nhm_cstates) - outp += sprintf(outp, "\tCPU%%c6"); - if (do_snb_cstates) - outp += sprintf(outp, "\tCPU%%c7"); + comma = strchr(name_list, ','); + + if (comma) + *comma = '\0'; + + for (i = 0; i < MAX_BIC; ++i) { + if (!strcmp(name_list, bic[i].name)) { + retval |= (1ULL << i); + break; + } + } + if (i == MAX_BIC) { + if (mode == SHOW_LIST) { + fprintf(stderr, "Invalid counter name: %s\n", name_list); + exit(-1); + } + deferred_skip_names[deferred_skip_index++] = name_list; + if (debug) + fprintf(stderr, "deferred \"%s\"\n", name_list); + if (deferred_skip_index >= MAX_DEFERRED) { + fprintf(stderr, "More than max %d un-recognized --skip options '%s'\n", + MAX_DEFERRED, name_list); + help(); + exit(1); + } + } + + name_list = comma; + if (name_list) + name_list++; + + } + return retval; +} + + +void print_header(char *delim) +{ + struct msr_counter *mp; + int printed = 0; + + if (DO_BIC(BIC_Package)) + outp += sprintf(outp, "%sPackage", (printed++ ? delim : "")); + if (DO_BIC(BIC_Core)) + outp += sprintf(outp, "%sCore", (printed++ ? delim : "")); + if (DO_BIC(BIC_CPU)) + outp += sprintf(outp, "%sCPU", (printed++ ? delim : "")); + if (DO_BIC(BIC_Avg_MHz)) + outp += sprintf(outp, "%sAvg_MHz", (printed++ ? delim : "")); + if (DO_BIC(BIC_Busy)) + outp += sprintf(outp, "%sBusy%%", (printed++ ? delim : "")); + if (DO_BIC(BIC_Bzy_MHz)) + outp += sprintf(outp, "%sBzy_MHz", (printed++ ? delim : "")); + if (DO_BIC(BIC_TSC_MHz)) + outp += sprintf(outp, "%sTSC_MHz", (printed++ ? delim : "")); + + if (DO_BIC(BIC_IRQ)) { + if (sums_need_wide_columns) + outp += sprintf(outp, "%s IRQ", (printed++ ? delim : "")); + else + outp += sprintf(outp, "%sIRQ", (printed++ ? delim : "")); + } + + if (DO_BIC(BIC_SMI)) + outp += sprintf(outp, "%sSMI", (printed++ ? delim : "")); for (mp = sys.tp; mp; mp = mp->next) { + if (mp->format == FORMAT_RAW) { if (mp->width == 64) - outp += sprintf(outp, "\t%18.18s", mp->name); + outp += sprintf(outp, "%s%18.18s", (printed++ ? delim : ""), mp->name); else - outp += sprintf(outp, "\t%10.10s", mp->name); + outp += sprintf(outp, "%s%10.10s", (printed++ ? delim : ""), mp->name); } else { - outp += sprintf(outp, "\t%-7.7s", mp->name); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8s", (printed++ ? delim : ""), mp->name); + else + outp += sprintf(outp, "%s%s", (printed++ ? delim : ""), mp->name); } } - if (do_dts) - outp += sprintf(outp, "\tCoreTmp"); + if (DO_BIC(BIC_CPU_c1)) + outp += sprintf(outp, "%sCPU%%c1", (printed++ ? delim : "")); + if (DO_BIC(BIC_CPU_c3) && !do_slm_cstates && !do_knl_cstates) + outp += sprintf(outp, "%sCPU%%c3", (printed++ ? delim : "")); + if (DO_BIC(BIC_CPU_c6)) + outp += sprintf(outp, "%sCPU%%c6", (printed++ ? delim : "")); + if (DO_BIC(BIC_CPU_c7)) + outp += sprintf(outp, "%sCPU%%c7", (printed++ ? delim : "")); + + if (DO_BIC(BIC_Mod_c6)) + outp += sprintf(outp, "%sMod%%c6", (printed++ ? delim : "")); + + if (DO_BIC(BIC_CoreTmp)) + outp += sprintf(outp, "%sCoreTmp", (printed++ ? delim : "")); for (mp = sys.cp; mp; mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 64) - outp += sprintf(outp, "\t%18.18s", mp->name); + outp += sprintf(outp, "%s%18.18s", delim, mp->name); else - outp += sprintf(outp, "\t%10.10s", mp->name); + outp += sprintf(outp, "%s%10.10s", delim, mp->name); } else { - outp += sprintf(outp, "\t%-7.7s", mp->name); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8s", delim, mp->name); + else + outp += sprintf(outp, "%s%s", delim, mp->name); } } - if (do_ptm) - outp += sprintf(outp, "\tPkgTmp"); + if (DO_BIC(BIC_PkgTmp)) + outp += sprintf(outp, "%sPkgTmp", (printed++ ? delim : "")); - if (do_gfx_rc6_ms) - outp += sprintf(outp, "\tGFX%%rc6"); + if (DO_BIC(BIC_GFX_rc6)) + outp += sprintf(outp, "%sGFX%%rc6", (printed++ ? delim : "")); - if (do_gfx_mhz) - outp += sprintf(outp, "\tGFXMHz"); + if (DO_BIC(BIC_GFXMHz)) + outp += sprintf(outp, "%sGFXMHz", (printed++ ? delim : "")); if (do_skl_residency) { - outp += sprintf(outp, "\tTotl%%C0"); - outp += sprintf(outp, "\tAny%%C0"); - outp += sprintf(outp, "\tGFX%%C0"); - outp += sprintf(outp, "\tCPUGFX%%"); - } - - if (do_pc2) - outp += sprintf(outp, "\tPkg%%pc2"); - if (do_pc3) - outp += sprintf(outp, "\tPkg%%pc3"); - if (do_pc6) - outp += sprintf(outp, "\tPkg%%pc6"); - if (do_pc7) - outp += sprintf(outp, "\tPkg%%pc7"); - if (do_c8_c9_c10) { - outp += sprintf(outp, "\tPkg%%pc8"); - outp += sprintf(outp, "\tPkg%%pc9"); - outp += sprintf(outp, "\tPk%%pc10"); + outp += sprintf(outp, "%sTotl%%C0", (printed++ ? delim : "")); + outp += sprintf(outp, "%sAny%%C0", (printed++ ? delim : "")); + outp += sprintf(outp, "%sGFX%%C0", (printed++ ? delim : "")); + outp += sprintf(outp, "%sCPUGFX%%", (printed++ ? delim : "")); } + if (DO_BIC(BIC_Pkgpc2)) + outp += sprintf(outp, "%sPkg%%pc2", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc3)) + outp += sprintf(outp, "%sPkg%%pc3", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc6)) + outp += sprintf(outp, "%sPkg%%pc6", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc7)) + outp += sprintf(outp, "%sPkg%%pc7", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc8)) + outp += sprintf(outp, "%sPkg%%pc8", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc9)) + outp += sprintf(outp, "%sPkg%%pc9", (printed++ ? delim : "")); + if (DO_BIC(BIC_Pkgpc10)) + outp += sprintf(outp, "%sPk%%pc10", (printed++ ? delim : "")); + if (do_rapl && !rapl_joules) { - if (do_rapl & RAPL_PKG) - outp += sprintf(outp, "\tPkgWatt"); - if (do_rapl & RAPL_CORES_ENERGY_STATUS) - outp += sprintf(outp, "\tCorWatt"); - if (do_rapl & RAPL_GFX) - outp += sprintf(outp, "\tGFXWatt"); - if (do_rapl & RAPL_DRAM) - outp += sprintf(outp, "\tRAMWatt"); - if (do_rapl & RAPL_PKG_PERF_STATUS) - outp += sprintf(outp, "\tPKG_%%"); - if (do_rapl & RAPL_DRAM_PERF_STATUS) - outp += sprintf(outp, "\tRAM_%%"); + if (DO_BIC(BIC_PkgWatt)) + outp += sprintf(outp, "%sPkgWatt", (printed++ ? delim : "")); + if (DO_BIC(BIC_CorWatt)) + outp += sprintf(outp, "%sCorWatt", (printed++ ? delim : "")); + if (DO_BIC(BIC_GFXWatt)) + outp += sprintf(outp, "%sGFXWatt", (printed++ ? delim : "")); + if (DO_BIC(BIC_RAMWatt)) + outp += sprintf(outp, "%sRAMWatt", (printed++ ? delim : "")); + if (DO_BIC(BIC_PKG__)) + outp += sprintf(outp, "%sPKG_%%", (printed++ ? delim : "")); + if (DO_BIC(BIC_RAM__)) + outp += sprintf(outp, "%sRAM_%%", (printed++ ? delim : "")); } else if (do_rapl && rapl_joules) { - if (do_rapl & RAPL_PKG) - outp += sprintf(outp, "\tPkg_J"); - if (do_rapl & RAPL_CORES_ENERGY_STATUS) - outp += sprintf(outp, "\tCor_J"); - if (do_rapl & RAPL_GFX) - outp += sprintf(outp, "\tGFX_J"); - if (do_rapl & RAPL_DRAM) - outp += sprintf(outp, "\tRAM_J"); - if (do_rapl & RAPL_PKG_PERF_STATUS) - outp += sprintf(outp, "\tPKG_%%"); - if (do_rapl & RAPL_DRAM_PERF_STATUS) - outp += sprintf(outp, "\tRAM_%%"); + if (DO_BIC(BIC_Pkg_J)) + outp += sprintf(outp, "%sPkg_J", (printed++ ? delim : "")); + if (DO_BIC(BIC_Cor_J)) + outp += sprintf(outp, "%sCor_J", (printed++ ? delim : "")); + if (DO_BIC(BIC_GFX_J)) + outp += sprintf(outp, "%sGFX_J", (printed++ ? delim : "")); + if (DO_BIC(BIC_RAM_J)) + outp += sprintf(outp, "%sRAM_J", (printed++ ? delim : "")); + if (DO_BIC(BIC_PKG__)) + outp += sprintf(outp, "%sPKG_%%", (printed++ ? delim : "")); + if (DO_BIC(BIC_RAM__)) + outp += sprintf(outp, "%sRAM_%%", (printed++ ? delim : "")); } for (mp = sys.pp; mp; mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 64) - outp += sprintf(outp, "\t%18.18s", mp->name); + outp += sprintf(outp, "%s%18.18s", delim, mp->name); else - outp += sprintf(outp, "\t%10.10s", mp->name); + outp += sprintf(outp, "%s%10.10s", delim, mp->name); } else { - outp += sprintf(outp, "\t%-7.7s", mp->name); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8s", delim, mp->name); + else + outp += sprintf(outp, "%s%s", delim, mp->name); } } -done: outp += sprintf(outp, "\n"); } @@ -494,10 +681,10 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "mperf: %016llX\n", t->mperf); outp += sprintf(outp, "c1: %016llX\n", t->c1); - if (do_irq) - outp += sprintf(outp, "IRQ: %08X\n", t->irq_count); - if (do_smi) - outp += sprintf(outp, "SMI: %08X\n", t->smi_count); + if (DO_BIC(BIC_IRQ)) + outp += sprintf(outp, "IRQ: %lld\n", t->irq_count); + if (DO_BIC(BIC_SMI)) + outp += sprintf(outp, "SMI: %d\n", t->smi_count); for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { outp += sprintf(outp, "tADDED [%d] msr0x%x: %08llX\n", @@ -516,6 +703,7 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "cADDED [%d] msr0x%x: %08llX\n", i, mp->msr_num, c->counter[i]); } + outp += sprintf(outp, "mc6_us: %016llX\n", c->mc6_us); } if (p) { @@ -527,11 +715,11 @@ int dump_counters(struct thread_data *t, struct core_data *c, outp += sprintf(outp, "CPU + GFX: %016llX\n", p->pkg_both_core_gfxe_c0); outp += sprintf(outp, "pc2: %016llX\n", p->pc2); - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) outp += sprintf(outp, "pc3: %016llX\n", p->pc3); - if (do_pc6) + if (DO_BIC(BIC_Pkgpc6)) outp += sprintf(outp, "pc6: %016llX\n", p->pc6); - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) outp += sprintf(outp, "pc7: %016llX\n", p->pc7); outp += sprintf(outp, "pc8: %016llX\n", p->pc8); outp += sprintf(outp, "pc9: %016llX\n", p->pc9); @@ -563,10 +751,12 @@ int dump_counters(struct thread_data *t, struct core_data *c, int format_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { - double interval_float; + double interval_float, tsc; char *fmt8; int i; struct msr_counter *mp; + char *delim = "\t"; + int printed = 0; /* if showing only 1st thread in core and this isn't one, bail out */ if (show_core_only && !(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) @@ -576,106 +766,126 @@ int format_counters(struct thread_data *t, struct core_data *c, if (show_pkg_only && !(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) return 0; + /*if not summary line and --cpu is used */ + if ((t != &average.threads) && + (cpu_subset && !CPU_ISSET_S(t->cpu_id, cpu_subset_size, cpu_subset))) + return 0; + interval_float = tv_delta.tv_sec + tv_delta.tv_usec/1000000.0; + tsc = t->tsc * tsc_tweak; + /* topo columns, print blanks on 1st (average) line */ if (t == &average.threads) { - if (show_pkg) - outp += sprintf(outp, "\t-"); - if (show_core) - outp += sprintf(outp, "\t-"); - if (show_cpu) - outp += sprintf(outp, "\t-"); + if (DO_BIC(BIC_Package)) + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); + if (DO_BIC(BIC_Core)) + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); + if (DO_BIC(BIC_CPU)) + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); } else { - if (show_pkg) { + if (DO_BIC(BIC_Package)) { if (p) - outp += sprintf(outp, "\t%d", p->package_id); + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->package_id); else - outp += sprintf(outp, "\t-"); + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); } - if (show_core) { + if (DO_BIC(BIC_Core)) { if (c) - outp += sprintf(outp, "\t%d", c->core_id); + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), c->core_id); else - outp += sprintf(outp, "\t-"); + outp += sprintf(outp, "%s-", (printed++ ? delim : "")); } - if (show_cpu) - outp += sprintf(outp, "\t%d", t->cpu_id); + if (DO_BIC(BIC_CPU)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), t->cpu_id); } - /* Avg_MHz */ - if (has_aperf) - outp += sprintf(outp, "\t%.0f", + if (DO_BIC(BIC_Avg_MHz)) + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), 1.0 / units * t->aperf / interval_float); - /* Busy% */ - if (has_aperf) - outp += sprintf(outp, "\t%.2f", 100.0 * t->mperf/t->tsc/tsc_tweak); + if (DO_BIC(BIC_Busy)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->mperf/tsc); - /* Bzy_MHz */ - if (has_aperf) { + if (DO_BIC(BIC_Bzy_MHz)) { if (has_base_hz) - outp += sprintf(outp, "\t%.0f", base_hz / units * t->aperf / t->mperf); + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), base_hz / units * t->aperf / t->mperf); else - outp += sprintf(outp, "\t%.0f", - 1.0 * t->tsc / units * t->aperf / t->mperf / interval_float); + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), + tsc / units * t->aperf / t->mperf / interval_float); } - /* TSC_MHz */ - outp += sprintf(outp, "\t%.0f", 1.0 * t->tsc/units/interval_float); - - if (!debug) - goto done; + if (DO_BIC(BIC_TSC_MHz)) + outp += sprintf(outp, "%s%.0f", (printed++ ? delim : ""), 1.0 * t->tsc/units/interval_float); /* IRQ */ - if (do_irq) - outp += sprintf(outp, "\t%d", t->irq_count); + if (DO_BIC(BIC_IRQ)) { + if (sums_need_wide_columns) + outp += sprintf(outp, "%s%8lld", (printed++ ? delim : ""), t->irq_count); + else + outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), t->irq_count); + } /* SMI */ - if (do_smi) - outp += sprintf(outp, "\t%d", t->smi_count); - - if (do_nhm_cstates) - outp += sprintf(outp, "\t%.2f", 100.0 * t->c1/t->tsc); - - /* print per-core data only for 1st thread in core */ - if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) - goto done; - - if (do_nhm_cstates && !do_slm_cstates && !do_knl_cstates) - outp += sprintf(outp, "\t%.2f", 100.0 * c->c3/t->tsc); - if (do_nhm_cstates) - outp += sprintf(outp, "\t%.2f", 100.0 * c->c6/t->tsc); - if (do_snb_cstates) - outp += sprintf(outp, "\t%.2f", 100.0 * c->c7/t->tsc); + if (DO_BIC(BIC_SMI)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), t->smi_count); + /* Added counters */ for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "\t0x%08lx", (unsigned long) t->counter[i]); + outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) t->counter[i]); else - outp += sprintf(outp, "\t0x%016llx", t->counter[i]); + outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), t->counter[i]); } else if (mp->format == FORMAT_DELTA) { - outp += sprintf(outp, "\t%8lld", t->counter[i]); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8lld", (printed++ ? delim : ""), t->counter[i]); + else + outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), t->counter[i]); } else if (mp->format == FORMAT_PERCENT) { - outp += sprintf(outp, "\t%.2f", 100.0 * t->counter[i]/t->tsc); + if (mp->type == COUNTER_USEC) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), t->counter[i]/interval_float/10000); + else + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->counter[i]/tsc); } } + /* C1 */ + if (DO_BIC(BIC_CPU_c1)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * t->c1/tsc); - if (do_dts) - outp += sprintf(outp, "\t%d", c->core_temp_c); + + /* print per-core data only for 1st thread in core */ + if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) + goto done; + + if (DO_BIC(BIC_CPU_c3) && !do_slm_cstates && !do_knl_cstates) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c3/tsc); + if (DO_BIC(BIC_CPU_c6)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c6/tsc); + if (DO_BIC(BIC_CPU_c7)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->c7/tsc); + + /* Mod%c6 */ + if (DO_BIC(BIC_Mod_c6)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->mc6_us / tsc); + + if (DO_BIC(BIC_CoreTmp)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), c->core_temp_c); for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "\t0x%08lx", (unsigned long) c->counter[i]); + outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) c->counter[i]); else - outp += sprintf(outp, "\t0x%016llx", c->counter[i]); + outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), c->counter[i]); } else if (mp->format == FORMAT_DELTA) { - outp += sprintf(outp, "\t%8lld", c->counter[i]); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8lld", (printed++ ? delim : ""), c->counter[i]); + else + outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), c->counter[i]); } else if (mp->format == FORMAT_PERCENT) { - outp += sprintf(outp, "\t%.2f", 100.0 * c->counter[i]/t->tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * c->counter[i]/tsc); } } @@ -684,95 +894,89 @@ int format_counters(struct thread_data *t, struct core_data *c, goto done; /* PkgTmp */ - if (do_ptm) - outp += sprintf(outp, "\t%d", p->pkg_temp_c); + if (DO_BIC(BIC_PkgTmp)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->pkg_temp_c); /* GFXrc6 */ - if (do_gfx_rc6_ms) { + if (DO_BIC(BIC_GFX_rc6)) { if (p->gfx_rc6_ms == -1) { /* detect GFX counter reset */ - outp += sprintf(outp, "\t**.**"); + outp += sprintf(outp, "%s**.**", (printed++ ? delim : "")); } else { - outp += sprintf(outp, "\t%.2f", + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), p->gfx_rc6_ms / 10.0 / interval_float); } } /* GFXMHz */ - if (do_gfx_mhz) - outp += sprintf(outp, "\t%d", p->gfx_mhz); + if (DO_BIC(BIC_GFXMHz)) + outp += sprintf(outp, "%s%d", (printed++ ? delim : ""), p->gfx_mhz); /* Totl%C0, Any%C0 GFX%C0 CPUGFX% */ if (do_skl_residency) { - outp += sprintf(outp, "\t%.2f", 100.0 * p->pkg_wtd_core_c0/t->tsc); - outp += sprintf(outp, "\t%.2f", 100.0 * p->pkg_any_core_c0/t->tsc); - outp += sprintf(outp, "\t%.2f", 100.0 * p->pkg_any_gfxe_c0/t->tsc); - outp += sprintf(outp, "\t%.2f", 100.0 * p->pkg_both_core_gfxe_c0/t->tsc); - } - - if (do_pc2) - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc2/t->tsc); - if (do_pc3) - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc3/t->tsc); - if (do_pc6) - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc6/t->tsc); - if (do_pc7) - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc7/t->tsc); - if (do_c8_c9_c10) { - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc8/t->tsc); - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc9/t->tsc); - outp += sprintf(outp, "\t%.2f", 100.0 * p->pc10/t->tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_wtd_core_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_core_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_any_gfxe_c0/tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pkg_both_core_gfxe_c0/tsc); } + if (DO_BIC(BIC_Pkgpc2)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc2/tsc); + if (DO_BIC(BIC_Pkgpc3)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc3/tsc); + if (DO_BIC(BIC_Pkgpc6)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc6/tsc); + if (DO_BIC(BIC_Pkgpc7)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc7/tsc); + if (DO_BIC(BIC_Pkgpc8)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc8/tsc); + if (DO_BIC(BIC_Pkgpc9)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc9/tsc); + if (DO_BIC(BIC_Pkgpc10)) + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->pc10/tsc); + /* * If measurement interval exceeds minimum RAPL Joule Counter range, * indicate that results are suspect by printing "**" in fraction place. */ if (interval_float < rapl_joule_counter_range) - fmt8 = "\t%.2f"; + fmt8 = "%s%.2f"; else fmt8 = "%6.0f**"; - if (do_rapl && !rapl_joules) { - if (do_rapl & RAPL_PKG) - outp += sprintf(outp, fmt8, p->energy_pkg * rapl_energy_units / interval_float); - if (do_rapl & RAPL_CORES_ENERGY_STATUS) - outp += sprintf(outp, fmt8, p->energy_cores * rapl_energy_units / interval_float); - if (do_rapl & RAPL_GFX) - outp += sprintf(outp, fmt8, p->energy_gfx * rapl_energy_units / interval_float); - if (do_rapl & RAPL_DRAM) - outp += sprintf(outp, fmt8, p->energy_dram * rapl_dram_energy_units / interval_float); - if (do_rapl & RAPL_PKG_PERF_STATUS) - outp += sprintf(outp, fmt8, 100.0 * p->rapl_pkg_perf_status * rapl_time_units / interval_float); - if (do_rapl & RAPL_DRAM_PERF_STATUS) - outp += sprintf(outp, fmt8, 100.0 * p->rapl_dram_perf_status * rapl_time_units / interval_float); - } else if (do_rapl && rapl_joules) { - if (do_rapl & RAPL_PKG) - outp += sprintf(outp, fmt8, - p->energy_pkg * rapl_energy_units); - if (do_rapl & RAPL_CORES) - outp += sprintf(outp, fmt8, - p->energy_cores * rapl_energy_units); - if (do_rapl & RAPL_GFX) - outp += sprintf(outp, fmt8, - p->energy_gfx * rapl_energy_units); - if (do_rapl & RAPL_DRAM) - outp += sprintf(outp, fmt8, - p->energy_dram * rapl_dram_energy_units); - if (do_rapl & RAPL_PKG_PERF_STATUS) - outp += sprintf(outp, fmt8, 100.0 * p->rapl_pkg_perf_status * rapl_time_units / interval_float); - if (do_rapl & RAPL_DRAM_PERF_STATUS) - outp += sprintf(outp, fmt8, 100.0 * p->rapl_dram_perf_status * rapl_time_units / interval_float); - } + if (DO_BIC(BIC_PkgWatt)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_pkg * rapl_energy_units / interval_float); + if (DO_BIC(BIC_CorWatt)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_cores * rapl_energy_units / interval_float); + if (DO_BIC(BIC_GFXWatt)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_gfx * rapl_energy_units / interval_float); + if (DO_BIC(BIC_RAMWatt)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_dram * rapl_dram_energy_units / interval_float); + if (DO_BIC(BIC_Pkg_J)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_pkg * rapl_energy_units); + if (DO_BIC(BIC_Cor_J)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_cores * rapl_energy_units); + if (DO_BIC(BIC_GFX_J)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_gfx * rapl_energy_units); + if (DO_BIC(BIC_RAM_J)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), p->energy_dram * rapl_dram_energy_units); + if (DO_BIC(BIC_PKG__)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), 100.0 * p->rapl_pkg_perf_status * rapl_time_units / interval_float); + if (DO_BIC(BIC_RAM__)) + outp += sprintf(outp, fmt8, (printed++ ? delim : ""), 100.0 * p->rapl_dram_perf_status * rapl_time_units / interval_float); + for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) { if (mp->width == 32) - outp += sprintf(outp, "\t0x%08lx", (unsigned long) p->counter[i]); + outp += sprintf(outp, "%s0x%08x", (printed++ ? delim : ""), (unsigned int) p->counter[i]); else - outp += sprintf(outp, "\t0x%016llx", p->counter[i]); + outp += sprintf(outp, "%s0x%016llx", (printed++ ? delim : ""), p->counter[i]); } else if (mp->format == FORMAT_DELTA) { - outp += sprintf(outp, "\t%8lld", p->counter[i]); + if ((mp->type == COUNTER_ITEMS) && sums_need_wide_columns) + outp += sprintf(outp, "%s%8lld", (printed++ ? delim : ""), p->counter[i]); + else + outp += sprintf(outp, "%s%lld", (printed++ ? delim : ""), p->counter[i]); } else if (mp->format == FORMAT_PERCENT) { - outp += sprintf(outp, "\t%.2f", 100.0 * p->counter[i]/t->tsc); + outp += sprintf(outp, "%s%.2f", (printed++ ? delim : ""), 100.0 * p->counter[i]/tsc); } } @@ -807,7 +1011,7 @@ void format_all_counters(struct thread_data *t, struct core_data *c, struct pkg_ static int printed; if (!printed || !summary_only) - print_header(); + print_header("\t"); if (topo.num_cpus > 1) format_counters(&average.threads, &average.cores, @@ -841,11 +1045,11 @@ delta_package(struct pkg_data *new, struct pkg_data *old) old->pkg_both_core_gfxe_c0 = new->pkg_both_core_gfxe_c0 - old->pkg_both_core_gfxe_c0; } old->pc2 = new->pc2 - old->pc2; - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) old->pc3 = new->pc3 - old->pc3; - if (do_pc6) + if (DO_BIC(BIC_Pkgpc6)) old->pc6 = new->pc6 - old->pc6; - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) old->pc7 = new->pc7 - old->pc7; old->pc8 = new->pc8 - old->pc8; old->pc9 = new->pc9 - old->pc9; @@ -887,6 +1091,7 @@ delta_core(struct core_data *new, struct core_data *old) old->c6 = new->c6 - old->c6; old->c7 = new->c7 - old->c7; old->core_temp_c = new->core_temp_c; + old->mc6_us = new->mc6_us - old->mc6_us; for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) @@ -916,7 +1121,7 @@ delta_thread(struct thread_data *new, struct thread_data *old, old->c1 = new->c1 - old->c1; - if (has_aperf) { + if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz)) { if ((new->aperf > old->aperf) && (new->mperf > old->mperf)) { old->aperf = new->aperf - old->aperf; old->mperf = new->mperf - old->mperf; @@ -937,11 +1142,11 @@ delta_thread(struct thread_data *new, struct thread_data *old, * it is possible for mperf's non-halted cycles + idle states * to exceed TSC's all cycles: show c1 = 0% in that case. */ - if ((old->mperf + core_delta->c3 + core_delta->c6 + core_delta->c7) > old->tsc) + if ((old->mperf + core_delta->c3 + core_delta->c6 + core_delta->c7) > (old->tsc * tsc_tweak)) old->c1 = 0; else { /* normal case, derive c1 */ - old->c1 = old->tsc - old->mperf - core_delta->c3 + old->c1 = (old->tsc * tsc_tweak) - old->mperf - core_delta->c3 - core_delta->c6 - core_delta->c7; } } @@ -952,10 +1157,10 @@ delta_thread(struct thread_data *new, struct thread_data *old, old->mperf = 1; /* divide by 0 protection */ } - if (do_irq) + if (DO_BIC(BIC_IRQ)) old->irq_count = new->irq_count - old->irq_count; - if (do_smi) + if (DO_BIC(BIC_SMI)) old->smi_count = new->smi_count - old->smi_count; for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { @@ -1008,6 +1213,7 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data c->c3 = 0; c->c6 = 0; c->c7 = 0; + c->mc6_us = 0; c->core_temp_c = 0; p->pkg_wtd_core_c0 = 0; @@ -1016,11 +1222,11 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data p->pkg_both_core_gfxe_c0 = 0; p->pc2 = 0; - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) p->pc3 = 0; - if (do_pc6) + if (DO_BIC(BIC_Pkgpc6)) p->pc6 = 0; - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) p->pc7 = 0; p->pc8 = 0; p->pc9 = 0; @@ -1036,7 +1242,6 @@ void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data p->gfx_rc6_ms = 0; p->gfx_mhz = 0; - for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) t->counter[i] = 0; @@ -1073,6 +1278,7 @@ int sum_counters(struct thread_data *t, struct core_data *c, average.cores.c3 += c->c3; average.cores.c6 += c->c6; average.cores.c7 += c->c7; + average.cores.mc6_us += c->mc6_us; average.cores.core_temp_c = MAX(average.cores.core_temp_c, c->core_temp_c); @@ -1094,11 +1300,11 @@ int sum_counters(struct thread_data *t, struct core_data *c, } average.packages.pc2 += p->pc2; - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) average.packages.pc3 += p->pc3; - if (do_pc6) + if (DO_BIC(BIC_Pkgpc6)) average.packages.pc6 += p->pc6; - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) average.packages.pc7 += p->pc7; average.packages.pc8 += p->pc8; average.packages.pc9 += p->pc9; @@ -1143,9 +1349,13 @@ void compute_average(struct thread_data *t, struct core_data *c, average.threads.mperf /= topo.num_cpus; average.threads.c1 /= topo.num_cpus; + if (average.threads.irq_count > 9999999) + sums_need_wide_columns = 1; + average.cores.c3 /= topo.num_cores; average.cores.c6 /= topo.num_cores; average.cores.c7 /= topo.num_cores; + average.cores.mc6_us /= topo.num_cores; if (do_skl_residency) { average.packages.pkg_wtd_core_c0 /= topo.num_packages; @@ -1155,11 +1365,11 @@ void compute_average(struct thread_data *t, struct core_data *c, } average.packages.pc2 /= topo.num_packages; - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) average.packages.pc3 /= topo.num_packages; - if (do_pc6) + if (DO_BIC(BIC_Pkgpc6)) average.packages.pc6 /= topo.num_packages; - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) average.packages.pc7 /= topo.num_packages; average.packages.pc8 /= topo.num_packages; @@ -1169,16 +1379,29 @@ void compute_average(struct thread_data *t, struct core_data *c, for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) continue; + if (mp->type == COUNTER_ITEMS) { + if (average.threads.counter[i] > 9999999) + sums_need_wide_columns = 1; + continue; + } average.threads.counter[i] /= topo.num_cpus; } for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) continue; + if (mp->type == COUNTER_ITEMS) { + if (average.cores.counter[i] > 9999999) + sums_need_wide_columns = 1; + } average.cores.counter[i] /= topo.num_cores; } for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { if (mp->format == FORMAT_RAW) continue; + if (mp->type == COUNTER_ITEMS) { + if (average.packages.counter[i] > 9999999) + sums_need_wide_columns = 1; + } average.packages.counter[i] /= topo.num_packages; } } @@ -1193,6 +1416,60 @@ static unsigned long long rdtsc(void) } /* + * Open a file, and exit on failure + */ +FILE *fopen_or_die(const char *path, const char *mode) +{ + FILE *filep = fopen(path, mode); + + if (!filep) + err(1, "%s: open failed", path); + return filep; +} +/* + * snapshot_sysfs_counter() + * + * return snapshot of given counter + */ +unsigned long long snapshot_sysfs_counter(char *path) +{ + FILE *fp; + int retval; + unsigned long long counter; + + fp = fopen_or_die(path, "r"); + + retval = fscanf(fp, "%lld", &counter); + if (retval != 1) + err(1, "snapshot_sysfs_counter(%s)", path); + + fclose(fp); + + return counter; +} + +int get_mp(int cpu, struct msr_counter *mp, unsigned long long *counterp) +{ + if (mp->msr_num != 0) { + if (get_msr(cpu, mp->msr_num, counterp)) + return -1; + } else { + char path[128]; + + if (mp->flags & SYSFS_PERCPU) { + sprintf(path, "/sys/devices/system/cpu/cpu%d/%s", + cpu, mp->path); + + *counterp = snapshot_sysfs_counter(path); + } else { + *counterp = snapshot_sysfs_counter(mp->path); + } + } + + return 0; +} + +/* * get_counters(...) * migrate to cpu * acquire and record local counters for that cpu @@ -1213,7 +1490,7 @@ int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) retry: t->tsc = rdtsc(); /* we are running on local CPU of interest */ - if (has_aperf) { + if (DO_BIC(BIC_Avg_MHz) || DO_BIC(BIC_Busy) || DO_BIC(BIC_Bzy_MHz)) { unsigned long long tsc_before, tsc_between, tsc_after, aperf_time, mperf_time; /* @@ -1269,35 +1546,33 @@ retry: t->mperf = t->mperf * aperf_mperf_multiplier; } - if (do_irq) + if (DO_BIC(BIC_IRQ)) t->irq_count = irqs_per_cpu[cpu]; - if (do_smi) { + if (DO_BIC(BIC_SMI)) { if (get_msr(cpu, MSR_SMI_COUNT, &msr)) return -5; t->smi_count = msr & 0xFFFFFFFF; } - - if (use_c1_residency_msr) { + if (DO_BIC(BIC_CPU_c1) && use_c1_residency_msr) { if (get_msr(cpu, MSR_CORE_C1_RES, &t->c1)) return -6; } for (i = 0, mp = sys.tp; mp; i++, mp = mp->next) { - if (get_msr(cpu, mp->msr_num, &t->counter[i])) + if (get_mp(cpu, mp, &t->counter[i])) return -10; } - /* collect core counters only for 1st thread in core */ if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) return 0; - if (do_nhm_cstates && !do_slm_cstates && !do_knl_cstates) { + if (DO_BIC(BIC_CPU_c3) && !do_slm_cstates && !do_knl_cstates) { if (get_msr(cpu, MSR_CORE_C3_RESIDENCY, &c->c3)) return -6; } - if (do_nhm_cstates && !do_knl_cstates) { + if (DO_BIC(BIC_CPU_c6) && !do_knl_cstates) { if (get_msr(cpu, MSR_CORE_C6_RESIDENCY, &c->c6)) return -7; } else if (do_knl_cstates) { @@ -1305,18 +1580,22 @@ retry: return -7; } - if (do_snb_cstates) + if (DO_BIC(BIC_CPU_c7)) if (get_msr(cpu, MSR_CORE_C7_RESIDENCY, &c->c7)) return -8; - if (do_dts) { + if (DO_BIC(BIC_Mod_c6)) + if (get_msr(cpu, MSR_MODULE_C6_RES_MS, &c->mc6_us)) + return -8; + + if (DO_BIC(BIC_CoreTmp)) { if (get_msr(cpu, MSR_IA32_THERM_STATUS, &msr)) return -9; c->core_temp_c = tcc_activation_temp - ((msr >> 16) & 0x7F); } for (i = 0, mp = sys.cp; mp; i++, mp = mp->next) { - if (get_msr(cpu, mp->msr_num, &c->counter[i])) + if (get_mp(cpu, mp, &c->counter[i])) return -10; } @@ -1334,26 +1613,35 @@ retry: if (get_msr(cpu, MSR_PKG_BOTH_CORE_GFXE_C0_RES, &p->pkg_both_core_gfxe_c0)) return -13; } - if (do_pc3) + if (DO_BIC(BIC_Pkgpc3)) if (get_msr(cpu, MSR_PKG_C3_RESIDENCY, &p->pc3)) return -9; - if (do_pc6) - if (get_msr(cpu, MSR_PKG_C6_RESIDENCY, &p->pc6)) - return -10; - if (do_pc2) + if (DO_BIC(BIC_Pkgpc6)) { + if (do_slm_cstates) { + if (get_msr(cpu, MSR_ATOM_PKG_C6_RESIDENCY, &p->pc6)) + return -10; + } else { + if (get_msr(cpu, MSR_PKG_C6_RESIDENCY, &p->pc6)) + return -10; + } + } + + if (DO_BIC(BIC_Pkgpc2)) if (get_msr(cpu, MSR_PKG_C2_RESIDENCY, &p->pc2)) return -11; - if (do_pc7) + if (DO_BIC(BIC_Pkgpc7)) if (get_msr(cpu, MSR_PKG_C7_RESIDENCY, &p->pc7)) return -12; - if (do_c8_c9_c10) { + if (DO_BIC(BIC_Pkgpc8)) if (get_msr(cpu, MSR_PKG_C8_RESIDENCY, &p->pc8)) return -13; + if (DO_BIC(BIC_Pkgpc9)) if (get_msr(cpu, MSR_PKG_C9_RESIDENCY, &p->pc9)) return -13; + if (DO_BIC(BIC_Pkgpc10)) if (get_msr(cpu, MSR_PKG_C10_RESIDENCY, &p->pc10)) return -13; - } + if (do_rapl & RAPL_PKG) { if (get_msr(cpu, MSR_PKG_ENERGY_STATUS, &msr)) return -13; @@ -1384,20 +1672,20 @@ retry: return -16; p->rapl_dram_perf_status = msr & 0xFFFFFFFF; } - if (do_ptm) { + if (DO_BIC(BIC_PkgTmp)) { if (get_msr(cpu, MSR_IA32_PACKAGE_THERM_STATUS, &msr)) return -17; p->pkg_temp_c = tcc_activation_temp - ((msr >> 16) & 0x7F); } - if (do_gfx_rc6_ms) + if (DO_BIC(BIC_GFX_rc6)) p->gfx_rc6_ms = gfx_cur_rc6_ms; - if (do_gfx_mhz) + if (DO_BIC(BIC_GFXMHz)) p->gfx_mhz = gfx_cur_mhz; for (i = 0, mp = sys.pp; mp; i++, mp = mp->next) { - if (get_msr(cpu, mp->msr_num, &p->counter[i])) + if (get_mp(cpu, mp, &p->counter[i])) return -10; } @@ -1433,8 +1721,8 @@ char *pkg_cstate_limit_strings[] = { "reserved", "unknown", "pc0", "pc1", "pc2", int nhm_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCL__3, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; int snb_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCL__7, PCL_7S, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; int hsw_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL__3, PCL__6, PCL__7, PCL_7S, PCL__8, PCL__9, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int slv_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCLRSV, PCLRSV, PCL__4, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; -int amt_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCL__2, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; +int slv_pkg_cstate_limits[16] = {PCL__0, PCL__1, PCLRSV, PCLRSV, PCL__4, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7}; +int amt_pkg_cstate_limits[16] = {PCLUNL, PCL__1, PCL__2, PCLRSV, PCLRSV, PCLRSV, PCL__6, PCL__7, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; int phi_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; int bxt_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; int skx_pkg_cstate_limits[16] = {PCL__0, PCL__2, PCL_6N, PCL_6R, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLUNL, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV, PCLRSV}; @@ -1457,11 +1745,11 @@ dump_nhm_platform_info(void) fprintf(outf, "cpu%d: MSR_PLATFORM_INFO: 0x%08llx\n", base_cpu, msr); ratio = (msr >> 40) & 0xFF; - fprintf(outf, "%d * %.0f = %.0f MHz max efficiency frequency\n", + fprintf(outf, "%d * %.1f = %.1f MHz max efficiency frequency\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0xFF; - fprintf(outf, "%d * %.0f = %.0f MHz base frequency\n", + fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", ratio, bclk, ratio * bclk); get_msr(base_cpu, MSR_IA32_POWER_CTL, &msr); @@ -1483,12 +1771,12 @@ dump_hsw_turbo_ratio_limits(void) ratio = (msr >> 8) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 18 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 18 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 0) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 17 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 17 active cores\n", ratio, bclk, ratio * bclk); return; } @@ -1505,99 +1793,175 @@ dump_ivt_turbo_ratio_limits(void) ratio = (msr >> 56) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 16 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 16 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 48) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 15 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 15 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 40) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 14 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 14 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 32) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 13 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 13 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 24) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 12 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 12 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 16) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 11 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 11 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 8) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 10 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 10 active cores\n", ratio, bclk, ratio * bclk); ratio = (msr >> 0) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 9 active cores\n", + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 9 active cores\n", ratio, bclk, ratio * bclk); return; } +int has_turbo_ratio_group_limits(int family, int model) +{ + + if (!genuine_intel) + return 0; + + switch (model) { + case INTEL_FAM6_ATOM_GOLDMONT: + case INTEL_FAM6_SKYLAKE_X: + case INTEL_FAM6_ATOM_DENVERTON: + return 1; + } + return 0; +} static void -dump_nhm_turbo_ratio_limits(void) +dump_turbo_ratio_limits(int family, int model) { - unsigned long long msr; - unsigned int ratio; + unsigned long long msr, core_counts; + unsigned int ratio, group_size; get_msr(base_cpu, MSR_TURBO_RATIO_LIMIT, &msr); - fprintf(outf, "cpu%d: MSR_TURBO_RATIO_LIMIT: 0x%08llx\n", base_cpu, msr); + if (has_turbo_ratio_group_limits(family, model)) { + get_msr(base_cpu, MSR_TURBO_RATIO_LIMIT1, &core_counts); + fprintf(outf, "cpu%d: MSR_TURBO_RATIO_LIMIT1: 0x%08llx\n", base_cpu, core_counts); + } else { + core_counts = 0x0807060504030201; + } + ratio = (msr >> 56) & 0xFF; + group_size = (core_counts >> 56) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 8 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 48) & 0xFF; + group_size = (core_counts >> 48) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 7 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 40) & 0xFF; + group_size = (core_counts >> 40) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 6 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 32) & 0xFF; + group_size = (core_counts >> 32) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 5 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 24) & 0xFF; + group_size = (core_counts >> 24) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 4 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 16) & 0xFF; + group_size = (core_counts >> 16) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 3 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 8) & 0xFF; + group_size = (core_counts >> 8) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 2 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); ratio = (msr >> 0) & 0xFF; + group_size = (core_counts >> 0) & 0xFF; if (ratio) - fprintf(outf, "%d * %.0f = %.0f MHz max turbo 1 active cores\n", - ratio, bclk, ratio * bclk); + fprintf(outf, "%d * %.1f = %.1f MHz max turbo %d active cores\n", + ratio, bclk, ratio * bclk, group_size); return; } static void +dump_atom_turbo_ratio_limits(void) +{ + unsigned long long msr; + unsigned int ratio; + + get_msr(base_cpu, MSR_ATOM_CORE_RATIOS, &msr); + fprintf(outf, "cpu%d: MSR_ATOM_CORE_RATIOS: 0x%08llx\n", base_cpu, msr & 0xFFFFFFFF); + + ratio = (msr >> 0) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz minimum operating frequency\n", + ratio, bclk, ratio * bclk); + + ratio = (msr >> 8) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz low frequency mode (LFM)\n", + ratio, bclk, ratio * bclk); + + ratio = (msr >> 16) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz base frequency\n", + ratio, bclk, ratio * bclk); + + get_msr(base_cpu, MSR_ATOM_CORE_TURBO_RATIOS, &msr); + fprintf(outf, "cpu%d: MSR_ATOM_CORE_TURBO_RATIOS: 0x%08llx\n", base_cpu, msr & 0xFFFFFFFF); + + ratio = (msr >> 24) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 4 active cores\n", + ratio, bclk, ratio * bclk); + + ratio = (msr >> 16) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 3 active cores\n", + ratio, bclk, ratio * bclk); + + ratio = (msr >> 8) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 2 active cores\n", + ratio, bclk, ratio * bclk); + + ratio = (msr >> 0) & 0x3F; + if (ratio) + fprintf(outf, "%d * %.1f = %.1f MHz max turbo 1 active core\n", + ratio, bclk, ratio * bclk); +} + +static void dump_knl_turbo_ratio_limits(void) { const unsigned int buckets_no = 7; @@ -1652,7 +2016,7 @@ dump_knl_turbo_ratio_limits(void) for (i = buckets_no - 1; i >= 0; i--) if (i > 0 ? ratio[i] != ratio[i - 1] : 1) fprintf(outf, - "%d * %.0f = %.0f MHz max turbo %d active cores\n", + "%d * %.1f = %.1f MHz max turbo %d active cores\n", ratio[i], bclk, ratio[i] * bclk, cores[i]); } @@ -1661,12 +2025,12 @@ dump_nhm_cst_cfg(void) { unsigned long long msr; - get_msr(base_cpu, MSR_NHM_SNB_PKG_CST_CFG_CTL, &msr); + get_msr(base_cpu, MSR_PKG_CST_CONFIG_CONTROL, &msr); #define SNB_C1_AUTO_UNDEMOTE (1UL << 27) #define SNB_C3_AUTO_UNDEMOTE (1UL << 28) - fprintf(outf, "cpu%d: MSR_NHM_SNB_PKG_CST_CFG_CTL: 0x%08llx", base_cpu, msr); + fprintf(outf, "cpu%d: MSR_PKG_CST_CONFIG_CONTROL: 0x%08llx", base_cpu, msr); fprintf(outf, " (%s%s%s%s%slocked: pkg-cstate-limit=%d: %s)\n", (msr & SNB_C3_AUTO_UNDEMOTE) ? "UNdemote-C3, " : "", @@ -1810,16 +2174,6 @@ void free_all_buffers(void) free(irqs_per_cpu); } -/* - * Open a file, and exit on failure - */ -FILE *fopen_or_die(const char *path, const char *mode) -{ - FILE *filep = fopen(path, mode); - if (!filep) - err(1, "%s: open failed", path); - return filep; -} /* * Parse a file containing a single int. @@ -2131,8 +2485,10 @@ int snapshot_gfx_mhz(void) if (fp == NULL) fp = fopen_or_die("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", "r"); - else + else { rewind(fp); + fflush(fp); + } retval = fscanf(fp, "%d", &gfx_cur_mhz); if (retval != 1) @@ -2148,13 +2504,14 @@ int snapshot_gfx_mhz(void) */ int snapshot_proc_sysfs_files(void) { - if (snapshot_proc_interrupts()) - return 1; + if (DO_BIC(BIC_IRQ)) + if (snapshot_proc_interrupts()) + return 1; - if (do_gfx_rc6_ms) + if (DO_BIC(BIC_GFX_rc6)) snapshot_gfx_rc6_ms(); - if (do_gfx_mhz) + if (DO_BIC(BIC_GFXMHz)) snapshot_gfx_mhz(); return 0; @@ -2283,7 +2640,9 @@ void check_permissions() * MSR_SMI_COUNT 0x00000034 * * MSR_PLATFORM_INFO 0x000000ce - * MSR_NHM_SNB_PKG_CST_CFG_CTL 0x000000e2 + * MSR_PKG_CST_CONFIG_CONTROL 0x000000e2 + * + * MSR_MISC_PWR_MGMT 0x000001aa * * MSR_PKG_C3_RESIDENCY 0x000003f8 * MSR_PKG_C6_RESIDENCY 0x000003f9 @@ -2291,7 +2650,8 @@ void check_permissions() * MSR_CORE_C6_RESIDENCY 0x000003fd * * Side effect: - * sets global pkg_cstate_limit to decode MSR_NHM_SNB_PKG_CST_CFG_CTL + * sets global pkg_cstate_limit to decode MSR_PKG_CST_CONFIG_CONTROL + * sets has_misc_feature_control */ int probe_nhm_msrs(unsigned int family, unsigned int model) { @@ -2322,6 +2682,7 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) case INTEL_FAM6_IVYBRIDGE: /* IVB */ case INTEL_FAM6_IVYBRIDGE_X: /* IVB Xeon */ pkg_cstate_limits = snb_pkg_cstate_limits; + has_misc_feature_control = 1; break; case INTEL_FAM6_HASWELL_CORE: /* HSW */ case INTEL_FAM6_HASWELL_X: /* HSX */ @@ -2336,29 +2697,34 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) case INTEL_FAM6_KABYLAKE_MOBILE: /* KBL */ case INTEL_FAM6_KABYLAKE_DESKTOP: /* KBL */ pkg_cstate_limits = hsw_pkg_cstate_limits; + has_misc_feature_control = 1; break; case INTEL_FAM6_SKYLAKE_X: /* SKX */ pkg_cstate_limits = skx_pkg_cstate_limits; + has_misc_feature_control = 1; break; case INTEL_FAM6_ATOM_SILVERMONT1: /* BYT */ + no_MSR_MISC_PWR_MGMT = 1; case INTEL_FAM6_ATOM_SILVERMONT2: /* AVN */ pkg_cstate_limits = slv_pkg_cstate_limits; break; case INTEL_FAM6_ATOM_AIRMONT: /* AMT */ pkg_cstate_limits = amt_pkg_cstate_limits; + no_MSR_MISC_PWR_MGMT = 1; break; case INTEL_FAM6_XEON_PHI_KNL: /* PHI */ case INTEL_FAM6_XEON_PHI_KNM: pkg_cstate_limits = phi_pkg_cstate_limits; break; case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_ATOM_GEMINI_LAKE: case INTEL_FAM6_ATOM_DENVERTON: /* DNV */ pkg_cstate_limits = bxt_pkg_cstate_limits; break; default: return 0; } - get_msr(base_cpu, MSR_NHM_SNB_PKG_CST_CFG_CTL, &msr); + get_msr(base_cpu, MSR_PKG_CST_CONFIG_CONTROL, &msr); pkg_cstate_limit = pkg_cstate_limits[msr & 0xF]; get_msr(base_cpu, MSR_PLATFORM_INFO, &msr); @@ -2368,8 +2734,69 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) has_base_hz = 1; return 1; } -int has_nhm_turbo_ratio_limit(unsigned int family, unsigned int model) +/* + * SLV client has support for unique MSRs: + * + * MSR_CC6_DEMOTION_POLICY_CONFIG + * MSR_MC6_DEMOTION_POLICY_CONFIG + */ + +int has_slv_msrs(unsigned int family, unsigned int model) { + if (!genuine_intel) + return 0; + + switch (model) { + case INTEL_FAM6_ATOM_SILVERMONT1: + case INTEL_FAM6_ATOM_MERRIFIELD: + case INTEL_FAM6_ATOM_MOOREFIELD: + return 1; + } + return 0; +} +int is_dnv(unsigned int family, unsigned int model) +{ + + if (!genuine_intel) + return 0; + + switch (model) { + case INTEL_FAM6_ATOM_DENVERTON: + return 1; + } + return 0; +} +int is_bdx(unsigned int family, unsigned int model) +{ + + if (!genuine_intel) + return 0; + + switch (model) { + case INTEL_FAM6_BROADWELL_X: + case INTEL_FAM6_BROADWELL_XEON_D: + return 1; + } + return 0; +} +int is_skx(unsigned int family, unsigned int model) +{ + + if (!genuine_intel) + return 0; + + switch (model) { + case INTEL_FAM6_SKYLAKE_X: + return 1; + } + return 0; +} + +int has_turbo_ratio_limit(unsigned int family, unsigned int model) +{ + if (has_slv_msrs(family, model)) + return 0; + switch (model) { /* Nehalem compatible, but do not include turbo-ratio limit support */ case INTEL_FAM6_NEHALEM_EX: /* Nehalem-EX Xeon - Beckton */ @@ -2381,6 +2808,13 @@ int has_nhm_turbo_ratio_limit(unsigned int family, unsigned int model) return 1; } } +int has_atom_turbo_ratio_limit(unsigned int family, unsigned int model) +{ + if (has_slv_msrs(family, model)) + return 1; + + return 0; +} int has_ivt_turbo_ratio_limit(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -2429,6 +2863,22 @@ int has_knl_turbo_ratio_limit(unsigned int family, unsigned int model) return 0; } } +int has_glm_turbo_ratio_limit(unsigned int family, unsigned int model) +{ + if (!genuine_intel) + return 0; + + if (family != 6) + return 0; + + switch (model) { + case INTEL_FAM6_ATOM_GOLDMONT: + case INTEL_FAM6_SKYLAKE_X: + return 1; + default: + return 0; + } +} int has_config_tdp(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -2475,8 +2925,11 @@ dump_cstate_pstate_config_info(unsigned int family, unsigned int model) if (has_ivt_turbo_ratio_limit(family, model)) dump_ivt_turbo_ratio_limits(); - if (has_nhm_turbo_ratio_limit(family, model)) - dump_nhm_turbo_ratio_limits(); + if (has_turbo_ratio_limit(family, model)) + dump_turbo_ratio_limits(family, model); + + if (has_atom_turbo_ratio_limit(family, model)) + dump_atom_turbo_ratio_limits(); if (has_knl_turbo_ratio_limit(family, model)) dump_knl_turbo_ratio_limits(); @@ -2487,6 +2940,96 @@ dump_cstate_pstate_config_info(unsigned int family, unsigned int model) dump_nhm_cst_cfg(); } +static void +dump_sysfs_cstate_config(void) +{ + char path[64]; + char name_buf[16]; + char desc[64]; + FILE *input; + int state; + char *sp; + + if (!DO_BIC(BIC_sysfs)) + return; + + for (state = 0; state < 10; ++state) { + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", + base_cpu, state); + input = fopen(path, "r"); + if (input == NULL) + continue; + fgets(name_buf, sizeof(name_buf), input); + + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + sp = strchr(name_buf, '-'); + if (!sp) + sp = strchrnul(name_buf, '\n'); + *sp = '\0'; + + fclose(input); + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/desc", + base_cpu, state); + input = fopen(path, "r"); + if (input == NULL) + continue; + fgets(desc, sizeof(desc), input); + + fprintf(outf, "cpu%d: %s: %s", base_cpu, name_buf, desc); + fclose(input); + } +} +static void +dump_sysfs_pstate_config(void) +{ + char path[64]; + char driver_buf[64]; + char governor_buf[64]; + FILE *input; + int turbo; + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_driver", + base_cpu); + input = fopen(path, "r"); + if (input == NULL) { + fprintf(stderr, "NSFOD %s\n", path); + return; + } + fgets(driver_buf, sizeof(driver_buf), input); + fclose(input); + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpufreq/scaling_governor", + base_cpu); + input = fopen(path, "r"); + if (input == NULL) { + fprintf(stderr, "NSFOD %s\n", path); + return; + } + fgets(governor_buf, sizeof(governor_buf), input); + fclose(input); + + fprintf(outf, "cpu%d: cpufreq driver: %s", base_cpu, driver_buf); + fprintf(outf, "cpu%d: cpufreq governor: %s", base_cpu, governor_buf); + + sprintf(path, "/sys/devices/system/cpu/cpufreq/boost"); + input = fopen(path, "r"); + if (input != NULL) { + fscanf(input, "%d", &turbo); + fprintf(outf, "cpufreq boost: %d\n", turbo); + fclose(input); + } + + sprintf(path, "/sys/devices/system/cpu/intel_pstate/no_turbo"); + input = fopen(path, "r"); + if (input != NULL) { + fscanf(input, "%d", &turbo); + fprintf(outf, "cpufreq intel_pstate no_turbo: %d\n", turbo); + fclose(input); + } +} + /* * print_epb() @@ -2570,7 +3113,7 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; fprintf(outf, "cpu%d: MSR_HWP_CAPABILITIES: 0x%08llx " - "(high 0x%x guar 0x%x eff 0x%x low 0x%x)\n", + "(high %d guar %d eff %d low %d)\n", cpu, msr, (unsigned int)HWP_HIGHEST_PERF(msr), (unsigned int)HWP_GUARANTEED_PERF(msr), @@ -2581,7 +3124,7 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; fprintf(outf, "cpu%d: MSR_HWP_REQUEST: 0x%08llx " - "(min 0x%x max 0x%x des 0x%x epp 0x%x window 0x%x pkg 0x%x)\n", + "(min %d max %d des %d epp 0x%x window 0x%x pkg 0x%x)\n", cpu, msr, (unsigned int)(((msr) >> 0) & 0xff), (unsigned int)(((msr) >> 8) & 0xff), @@ -2595,7 +3138,7 @@ int print_hwp(struct thread_data *t, struct core_data *c, struct pkg_data *p) return 0; fprintf(outf, "cpu%d: MSR_HWP_REQUEST_PKG: 0x%08llx " - "(min 0x%x max 0x%x des 0x%x epp 0x%x window 0x%x)\n", + "(min %d max %d des %d epp 0x%x window 0x%x)\n", cpu, msr, (unsigned int)(((msr) >> 0) & 0xff), (unsigned int)(((msr) >> 8) & 0xff), @@ -2790,15 +3333,42 @@ void rapl_probe(unsigned int family, unsigned int model) case INTEL_FAM6_BROADWELL_CORE: /* BDW */ case INTEL_FAM6_BROADWELL_GT3E: /* BDW */ do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_GFX | RAPL_PKG_POWER_INFO; + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + BIC_PRESENT(BIC_GFX_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); + BIC_PRESENT(BIC_GFXWatt); + } break; case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_ATOM_GEMINI_LAKE: do_rapl = RAPL_PKG | RAPL_PKG_POWER_INFO; + if (rapl_joules) + BIC_PRESENT(BIC_Pkg_J); + else + BIC_PRESENT(BIC_PkgWatt); break; case INTEL_FAM6_SKYLAKE_MOBILE: /* SKL */ case INTEL_FAM6_SKYLAKE_DESKTOP: /* SKL */ case INTEL_FAM6_KABYLAKE_MOBILE: /* KBL */ case INTEL_FAM6_KABYLAKE_DESKTOP: /* KBL */ - do_rapl = RAPL_PKG | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO; + do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_GFX | RAPL_PKG_POWER_INFO; + BIC_PRESENT(BIC_PKG__); + BIC_PRESENT(BIC_RAM__); + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + BIC_PRESENT(BIC_RAM_J); + BIC_PRESENT(BIC_GFX_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); + BIC_PRESENT(BIC_RAMWatt); + BIC_PRESENT(BIC_GFXWatt); + } break; case INTEL_FAM6_HASWELL_X: /* HSX */ case INTEL_FAM6_BROADWELL_X: /* BDX */ @@ -2807,17 +3377,55 @@ void rapl_probe(unsigned int family, unsigned int model) case INTEL_FAM6_XEON_PHI_KNL: /* KNL */ case INTEL_FAM6_XEON_PHI_KNM: do_rapl = RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO; + BIC_PRESENT(BIC_PKG__); + BIC_PRESENT(BIC_RAM__); + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_RAM_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_RAMWatt); + } break; case INTEL_FAM6_SANDYBRIDGE_X: case INTEL_FAM6_IVYBRIDGE_X: do_rapl = RAPL_PKG | RAPL_CORES | RAPL_CORE_POLICY | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_PKG_PERF_STATUS | RAPL_DRAM_PERF_STATUS | RAPL_PKG_POWER_INFO; + BIC_PRESENT(BIC_PKG__); + BIC_PRESENT(BIC_RAM__); + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + BIC_PRESENT(BIC_RAM_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); + BIC_PRESENT(BIC_RAMWatt); + } break; case INTEL_FAM6_ATOM_SILVERMONT1: /* BYT */ case INTEL_FAM6_ATOM_SILVERMONT2: /* AVN */ do_rapl = RAPL_PKG | RAPL_CORES; + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); + } break; case INTEL_FAM6_ATOM_DENVERTON: /* DNV */ do_rapl = RAPL_PKG | RAPL_DRAM | RAPL_DRAM_POWER_INFO | RAPL_DRAM_PERF_STATUS | RAPL_PKG_PERF_STATUS | RAPL_PKG_POWER_INFO | RAPL_CORES_ENERGY_STATUS; + BIC_PRESENT(BIC_PKG__); + BIC_PRESENT(BIC_RAM__); + if (rapl_joules) { + BIC_PRESENT(BIC_Pkg_J); + BIC_PRESENT(BIC_Cor_J); + BIC_PRESENT(BIC_RAM_J); + } else { + BIC_PRESENT(BIC_PkgWatt); + BIC_PRESENT(BIC_CorWatt); + BIC_PRESENT(BIC_RAMWatt); + } break; default: return; @@ -2844,7 +3452,7 @@ void rapl_probe(unsigned int family, unsigned int model) tdp = get_tdp(model); rapl_joule_counter_range = 0xFFFFFFFF * rapl_energy_units / tdp; - if (debug) + if (!quiet) fprintf(outf, "RAPL: %.0f sec. Joule Counter Range, at %.0f Watts\n", rapl_joule_counter_range, tdp); return; @@ -2874,7 +3482,7 @@ void perf_limit_reasons_probe(unsigned int family, unsigned int model) int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p) { unsigned long long msr; - unsigned int dts; + unsigned int dts, dts2; int cpu; if (!(do_dts || do_ptm)) @@ -2899,7 +3507,6 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p fprintf(outf, "cpu%d: MSR_IA32_PACKAGE_THERM_STATUS: 0x%08llx (%d C)\n", cpu, msr, tcc_activation_temp - dts); -#ifdef THERM_DEBUG if (get_msr(cpu, MSR_IA32_PACKAGE_THERM_INTERRUPT, &msr)) return 0; @@ -2907,11 +3514,10 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p dts2 = (msr >> 8) & 0x7F; fprintf(outf, "cpu%d: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x%08llx (%d C, %d C)\n", cpu, msr, tcc_activation_temp - dts, tcc_activation_temp - dts2); -#endif } - if (do_dts) { + if (do_dts && debug) { unsigned int resolution; if (get_msr(cpu, MSR_IA32_THERM_STATUS, &msr)) @@ -2922,7 +3528,6 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p fprintf(outf, "cpu%d: MSR_IA32_THERM_STATUS: 0x%08llx (%d C +/- %d)\n", cpu, msr, tcc_activation_temp - dts, resolution); -#ifdef THERM_DEBUG if (get_msr(cpu, MSR_IA32_THERM_INTERRUPT, &msr)) return 0; @@ -2930,7 +3535,6 @@ int print_thermal(struct thread_data *t, struct core_data *c, struct pkg_data *p dts2 = (msr >> 8) & 0x7F; fprintf(outf, "cpu%d: MSR_IA32_THERM_INTERRUPT: 0x%08llx (%d C, %d C)\n", cpu, msr, tcc_activation_temp - dts, tcc_activation_temp - dts2); -#endif } return 0; @@ -2969,11 +3573,9 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (get_msr(cpu, MSR_RAPL_POWER_UNIT, &msr)) return -1; - if (debug) { - fprintf(outf, "cpu%d: MSR_RAPL_POWER_UNIT: 0x%08llx " - "(%f Watts, %f Joules, %f sec.)\n", cpu, msr, - rapl_power_units, rapl_energy_units, rapl_time_units); - } + fprintf(outf, "cpu%d: MSR_RAPL_POWER_UNIT: 0x%08llx (%f Watts, %f Joules, %f sec.)\n", cpu, msr, + rapl_power_units, rapl_energy_units, rapl_time_units); + if (do_rapl & RAPL_PKG_POWER_INFO) { if (get_msr(cpu, MSR_PKG_POWER_INFO, &msr)) @@ -2994,7 +3596,7 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) return -9; fprintf(outf, "cpu%d: MSR_PKG_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 63) & 1 ? "": "UN"); + cpu, msr, (msr >> 63) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "PKG Limit #1"); fprintf(outf, "cpu%d: PKG Limit #2: %sabled (%f Watts, %f* sec, clamp %sabled)\n", @@ -3020,40 +3622,34 @@ int print_rapl(struct thread_data *t, struct core_data *c, struct pkg_data *p) if (get_msr(cpu, MSR_DRAM_POWER_LIMIT, &msr)) return -9; fprintf(outf, "cpu%d: MSR_DRAM_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "": "UN"); + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); print_power_limit_msr(cpu, msr, "DRAM Limit"); } if (do_rapl & RAPL_CORE_POLICY) { - if (debug) { - if (get_msr(cpu, MSR_PP0_POLICY, &msr)) - return -7; + if (get_msr(cpu, MSR_PP0_POLICY, &msr)) + return -7; - fprintf(outf, "cpu%d: MSR_PP0_POLICY: %lld\n", cpu, msr & 0xF); - } + fprintf(outf, "cpu%d: MSR_PP0_POLICY: %lld\n", cpu, msr & 0xF); } if (do_rapl & RAPL_CORES_POWER_LIMIT) { - if (debug) { - if (get_msr(cpu, MSR_PP0_POWER_LIMIT, &msr)) - return -9; - fprintf(outf, "cpu%d: MSR_PP0_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "": "UN"); - print_power_limit_msr(cpu, msr, "Cores Limit"); - } + if (get_msr(cpu, MSR_PP0_POWER_LIMIT, &msr)) + return -9; + fprintf(outf, "cpu%d: MSR_PP0_POWER_LIMIT: 0x%08llx (%slocked)\n", + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); + print_power_limit_msr(cpu, msr, "Cores Limit"); } if (do_rapl & RAPL_GFX) { - if (debug) { - if (get_msr(cpu, MSR_PP1_POLICY, &msr)) - return -8; + if (get_msr(cpu, MSR_PP1_POLICY, &msr)) + return -8; - fprintf(outf, "cpu%d: MSR_PP1_POLICY: %lld\n", cpu, msr & 0xF); + fprintf(outf, "cpu%d: MSR_PP1_POLICY: %lld\n", cpu, msr & 0xF); - if (get_msr(cpu, MSR_PP1_POWER_LIMIT, &msr)) - return -9; - fprintf(outf, "cpu%d: MSR_PP1_POWER_LIMIT: 0x%08llx (%slocked)\n", - cpu, msr, (msr >> 31) & 1 ? "": "UN"); - print_power_limit_msr(cpu, msr, "GFX Limit"); - } + if (get_msr(cpu, MSR_PP1_POWER_LIMIT, &msr)) + return -9; + fprintf(outf, "cpu%d: MSR_PP1_POWER_LIMIT: 0x%08llx (%slocked)\n", + cpu, msr, (msr >> 31) & 1 ? "" : "UN"); + print_power_limit_msr(cpu, msr, "GFX Limit"); } return 0; } @@ -3090,6 +3686,7 @@ int has_snb_msrs(unsigned int family, unsigned int model) case INTEL_FAM6_KABYLAKE_DESKTOP: /* KBL */ case INTEL_FAM6_SKYLAKE_X: /* SKX */ case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_ATOM_GEMINI_LAKE: case INTEL_FAM6_ATOM_DENVERTON: /* DNV */ return 1; } @@ -3121,6 +3718,7 @@ int has_hsw_msrs(unsigned int family, unsigned int model) case INTEL_FAM6_KABYLAKE_MOBILE: /* KBL */ case INTEL_FAM6_KABYLAKE_DESKTOP: /* KBL */ case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_ATOM_GEMINI_LAKE: return 1; } return 0; @@ -3149,8 +3747,6 @@ int has_skl_msrs(unsigned int family, unsigned int model) return 0; } - - int is_slm(unsigned int family, unsigned int model) { if (!genuine_intel) @@ -3201,7 +3797,8 @@ double slm_bclk(void) } freq = slm_freq_table[i]; - fprintf(outf, "SLM BCLK: %.1f Mhz\n", freq); + if (!quiet) + fprintf(outf, "SLM BCLK: %.1f Mhz\n", freq); return freq; } @@ -3264,7 +3861,7 @@ int set_temperature_target(struct thread_data *t, struct core_data *c, struct pk target_c_local = (msr >> 16) & 0xFF; - if (debug) + if (!quiet) fprintf(outf, "cpu%d: MSR_IA32_TEMPERATURE_TARGET: 0x%08llx (%d C)\n", cpu, msr, target_c_local); @@ -3299,13 +3896,30 @@ void decode_misc_enable_msr(void) unsigned long long msr; if (!get_msr(base_cpu, MSR_IA32_MISC_ENABLE, &msr)) - fprintf(outf, "cpu%d: MSR_IA32_MISC_ENABLE: 0x%08llx (%s %s %s)\n", + fprintf(outf, "cpu%d: MSR_IA32_MISC_ENABLE: 0x%08llx (%sTCC %sEIST %sMWAIT %sPREFETCH %sTURBO)\n", base_cpu, msr, - msr & (1 << 3) ? "TCC" : "", - msr & (1 << 16) ? "EIST" : "", - msr & (1 << 18) ? "MONITOR" : ""); + msr & MSR_IA32_MISC_ENABLE_TM1 ? "" : "No-", + msr & MSR_IA32_MISC_ENABLE_ENHANCED_SPEEDSTEP ? "" : "No-", + msr & MSR_IA32_MISC_ENABLE_MWAIT ? "No-" : "", + msr & MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE ? "No-" : "", + msr & MSR_IA32_MISC_ENABLE_TURBO_DISABLE ? "No-" : ""); } +void decode_misc_feature_control(void) +{ + unsigned long long msr; + + if (!has_misc_feature_control) + return; + + if (!get_msr(base_cpu, MSR_MISC_FEATURE_CONTROL, &msr)) + fprintf(outf, "cpu%d: MSR_MISC_FEATURE_CONTROL: 0x%08llx (%sL2-Prefetch %sL2-Prefetch-pair %sL1-Prefetch %sL1-IP-Prefetch)\n", + base_cpu, msr, + msr & (0 << 0) ? "No-" : "", + msr & (1 << 0) ? "No-" : "", + msr & (2 << 0) ? "No-" : "", + msr & (3 << 0) ? "No-" : ""); +} /* * Decode MSR_MISC_PWR_MGMT * @@ -3320,6 +3934,9 @@ void decode_misc_pwr_mgmt_msr(void) if (!do_nhm_platform_info) return; + if (no_MSR_MISC_PWR_MGMT) + return; + if (!get_msr(base_cpu, MSR_MISC_PWR_MGMT, &msr)) fprintf(outf, "cpu%d: MSR_MISC_PWR_MGMT: 0x%08llx (%sable-EIST_Coordination %sable-EPB %sable-OOB)\n", base_cpu, msr, @@ -3327,11 +3944,30 @@ void decode_misc_pwr_mgmt_msr(void) msr & (1 << 1) ? "EN" : "DIS", msr & (1 << 8) ? "EN" : "DIS"); } +/* + * Decode MSR_CC6_DEMOTION_POLICY_CONFIG, MSR_MC6_DEMOTION_POLICY_CONFIG + * + * This MSRs are present on Silvermont processors, + * Intel Atom processor E3000 series (Baytrail), and friends. + */ +void decode_c6_demotion_policy_msr(void) +{ + unsigned long long msr; + + if (!get_msr(base_cpu, MSR_CC6_DEMOTION_POLICY_CONFIG, &msr)) + fprintf(outf, "cpu%d: MSR_CC6_DEMOTION_POLICY_CONFIG: 0x%08llx (%sable-CC6-Demotion)\n", + base_cpu, msr, msr & (1 << 0) ? "EN" : "DIS"); + + if (!get_msr(base_cpu, MSR_MC6_DEMOTION_POLICY_CONFIG, &msr)) + fprintf(outf, "cpu%d: MSR_MC6_DEMOTION_POLICY_CONFIG: 0x%08llx (%sable-MC6-Demotion)\n", + base_cpu, msr, msr & (1 << 0) ? "EN" : "DIS"); +} void process_cpuid() { unsigned int eax, ebx, ecx, edx, max_level, max_extended_level; unsigned int fms, family, model, stepping; + unsigned int has_turbo; eax = ebx = ecx = edx = 0; @@ -3340,7 +3976,7 @@ void process_cpuid() if (ebx == 0x756e6547 && edx == 0x49656e69 && ecx == 0x6c65746e) genuine_intel = 1; - if (debug) + if (!quiet) fprintf(outf, "CPUID(0): %.4s%.4s%.4s ", (char *)&ebx, (char *)&edx, (char *)&ecx); @@ -3351,7 +3987,7 @@ void process_cpuid() if (family == 6 || family == 0xf) model += ((fms >> 16) & 0xf) << 4; - if (debug) { + if (!quiet) { fprintf(outf, "%d CPUID levels; family:model:stepping 0x%x:%x:%x (%d:%d:%d)\n", max_level, family, model, stepping, family, model, stepping); fprintf(outf, "CPUID(1): %s %s %s %s %s %s %s %s %s\n", @@ -3394,8 +4030,18 @@ void process_cpuid() __cpuid(0x6, eax, ebx, ecx, edx); has_aperf = ecx & (1 << 0); + if (has_aperf) { + BIC_PRESENT(BIC_Avg_MHz); + BIC_PRESENT(BIC_Busy); + BIC_PRESENT(BIC_Bzy_MHz); + } do_dts = eax & (1 << 0); + if (do_dts) + BIC_PRESENT(BIC_CoreTmp); + has_turbo = eax & (1 << 1); do_ptm = eax & (1 << 6); + if (do_ptm) + BIC_PRESENT(BIC_PkgTmp); has_hwp = eax & (1 << 7); has_hwp_notify = eax & (1 << 8); has_hwp_activity_window = eax & (1 << 9); @@ -3403,10 +4049,11 @@ void process_cpuid() has_hwp_pkg = eax & (1 << 11); has_epb = ecx & (1 << 3); - if (debug) - fprintf(outf, "CPUID(6): %sAPERF, %sDTS, %sPTM, %sHWP, " + if (!quiet) + fprintf(outf, "CPUID(6): %sAPERF, %sTURBO, %sDTS, %sPTM, %sHWP, " "%sHWPnotify, %sHWPwindow, %sHWPepp, %sHWPpkg, %sEPB\n", has_aperf ? "" : "No-", + has_turbo ? "" : "No-", do_dts ? "" : "No-", do_ptm ? "" : "No-", has_hwp ? "" : "No-", @@ -3416,10 +4063,11 @@ void process_cpuid() has_hwp_pkg ? "" : "No-", has_epb ? "" : "No-"); - if (debug) + if (!quiet) decode_misc_enable_msr(); - if (max_level >= 0x7 && debug) { + + if (max_level >= 0x7 && !quiet) { int has_sgx; ecx = 0; @@ -3445,7 +4093,7 @@ void process_cpuid() if (ebx_tsc != 0) { - if (debug && (ebx != 0)) + if (!quiet && (ebx != 0)) fprintf(outf, "CPUID(0x15): eax_crystal: %d ebx_tsc: %d ecx_crystal_hz: %d\n", eax_crystal, ebx_tsc, crystal_hz); @@ -3462,6 +4110,7 @@ void process_cpuid() crystal_hz = 25000000; /* 25.0 MHz */ break; case INTEL_FAM6_ATOM_GOLDMONT: /* BXT */ + case INTEL_FAM6_ATOM_GEMINI_LAKE: crystal_hz = 19200000; /* 19.2 MHz */ break; default: @@ -3470,7 +4119,7 @@ void process_cpuid() if (crystal_hz) { tsc_hz = (unsigned long long) crystal_hz * ebx_tsc / eax_crystal; - if (debug) + if (!quiet) fprintf(outf, "TSC: %lld MHz (%d Hz * %d / %d / 1000000)\n", tsc_hz / 1000000, crystal_hz, ebx_tsc, eax_crystal); } @@ -3485,7 +4134,7 @@ void process_cpuid() base_mhz = max_mhz = bus_mhz = edx = 0; __cpuid(0x16, base_mhz, max_mhz, bus_mhz, edx); - if (debug) + if (!quiet) fprintf(outf, "CPUID(0x16): base_mhz: %d max_mhz: %d bus_mhz: %d\n", base_mhz, max_mhz, bus_mhz); } @@ -3493,56 +4142,96 @@ void process_cpuid() if (has_aperf) aperf_mperf_multiplier = get_aperf_mperf_multiplier(family, model); - do_nhm_platform_info = do_nhm_cstates = do_smi = probe_nhm_msrs(family, model); + BIC_PRESENT(BIC_IRQ); + BIC_PRESENT(BIC_TSC_MHz); + + if (probe_nhm_msrs(family, model)) { + do_nhm_platform_info = 1; + BIC_PRESENT(BIC_CPU_c1); + BIC_PRESENT(BIC_CPU_c3); + BIC_PRESENT(BIC_CPU_c6); + BIC_PRESENT(BIC_SMI); + } do_snb_cstates = has_snb_msrs(family, model); + + if (do_snb_cstates) + BIC_PRESENT(BIC_CPU_c7); + do_irtl_snb = has_snb_msrs(family, model); - do_pc2 = do_snb_cstates && (pkg_cstate_limit >= PCL__2); - do_pc3 = (pkg_cstate_limit >= PCL__3); - do_pc6 = (pkg_cstate_limit >= PCL__6); - do_pc7 = do_snb_cstates && (pkg_cstate_limit >= PCL__7); - do_c8_c9_c10 = has_hsw_msrs(family, model); + if (do_snb_cstates && (pkg_cstate_limit >= PCL__2)) + BIC_PRESENT(BIC_Pkgpc2); + if (pkg_cstate_limit >= PCL__3) + BIC_PRESENT(BIC_Pkgpc3); + if (pkg_cstate_limit >= PCL__6) + BIC_PRESENT(BIC_Pkgpc6); + if (do_snb_cstates && (pkg_cstate_limit >= PCL__7)) + BIC_PRESENT(BIC_Pkgpc7); + if (has_slv_msrs(family, model)) { + BIC_NOT_PRESENT(BIC_Pkgpc2); + BIC_NOT_PRESENT(BIC_Pkgpc3); + BIC_PRESENT(BIC_Pkgpc6); + BIC_NOT_PRESENT(BIC_Pkgpc7); + BIC_PRESENT(BIC_Mod_c6); + use_c1_residency_msr = 1; + } + if (is_dnv(family, model)) { + BIC_PRESENT(BIC_CPU_c1); + BIC_NOT_PRESENT(BIC_CPU_c3); + BIC_NOT_PRESENT(BIC_Pkgpc3); + BIC_NOT_PRESENT(BIC_CPU_c7); + BIC_NOT_PRESENT(BIC_Pkgpc7); + use_c1_residency_msr = 1; + } + if (is_skx(family, model)) { + BIC_NOT_PRESENT(BIC_CPU_c3); + BIC_NOT_PRESENT(BIC_Pkgpc3); + BIC_NOT_PRESENT(BIC_CPU_c7); + BIC_NOT_PRESENT(BIC_Pkgpc7); + } + if (is_bdx(family, model)) { + BIC_NOT_PRESENT(BIC_CPU_c7); + BIC_NOT_PRESENT(BIC_Pkgpc7); + } + if (has_hsw_msrs(family, model)) { + BIC_PRESENT(BIC_Pkgpc8); + BIC_PRESENT(BIC_Pkgpc9); + BIC_PRESENT(BIC_Pkgpc10); + } do_irtl_hsw = has_hsw_msrs(family, model); do_skl_residency = has_skl_msrs(family, model); do_slm_cstates = is_slm(family, model); do_knl_cstates = is_knl(family, model); - if (debug) + if (!quiet) decode_misc_pwr_mgmt_msr(); + if (!quiet && has_slv_msrs(family, model)) + decode_c6_demotion_policy_msr(); + rapl_probe(family, model); perf_limit_reasons_probe(family, model); - if (debug) + if (!quiet) dump_cstate_pstate_config_info(family, model); + if (!quiet) + dump_sysfs_cstate_config(); + if (!quiet) + dump_sysfs_pstate_config(); + if (has_skl_msrs(family, model)) calculate_tsc_tweak(); - do_gfx_rc6_ms = !access("/sys/class/drm/card0/power/rc6_residency_ms", R_OK); + if (!access("/sys/class/drm/card0/power/rc6_residency_ms", R_OK)) + BIC_PRESENT(BIC_GFX_rc6); - do_gfx_mhz = !access("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", R_OK); + if (!access("/sys/class/graphics/fb0/device/drm/card0/gt_cur_freq_mhz", R_OK)) + BIC_PRESENT(BIC_GFXMHz); - return; -} + if (!quiet) + decode_misc_feature_control(); -void help() -{ - fprintf(outf, - "Usage: turbostat [OPTIONS][(--interval seconds) | COMMAND ...]\n" - "\n" - "Turbostat forks the specified COMMAND and prints statistics\n" - "when COMMAND completes.\n" - "If no COMMAND is specified, turbostat wakes every 5-seconds\n" - "to print statistics, until interrupted.\n" - "--add add a counter\n" - " eg. --add msr0x10,u64,cpu,delta,MY_TSC\n" - "--debug run in \"debug\" mode\n" - "--interval sec Override default 5-second measurement interval\n" - "--help print this help message\n" - "--out file create or truncate \"file\" for all output\n" - "--version print version information\n" - "\n" - "For more help, run \"man turbostat\"\n"); + return; } @@ -3579,7 +4268,7 @@ void topology_probe() topo.max_cpu_num = 0; for_all_proc_cpus(count_cpus); if (!summary_only && topo.num_cpus > 1) - show_cpu = 1; + BIC_PRESENT(BIC_CPU); if (debug > 1) fprintf(outf, "num_cpus %d max_cpu_num %d\n", topo.num_cpus, topo.max_cpu_num); @@ -3599,6 +4288,15 @@ void topology_probe() for_all_proc_cpus(mark_cpu_present); /* + * Validate that all cpus in cpu_subset are also in cpu_present_set + */ + for (i = 0; i < CPU_SUBSET_MAXCPUS; ++i) { + if (CPU_ISSET_S(i, cpu_subset_size, cpu_subset)) + if (!CPU_ISSET_S(i, cpu_present_setsize, cpu_present_set)) + err(1, "cpu%d not present", i); + } + + /* * Allocate and initialize cpu_affinity_set */ cpu_affinity_set = CPU_ALLOC((topo.max_cpu_num + 1)); @@ -3639,15 +4337,15 @@ void topology_probe() if (debug > 1) fprintf(outf, "max_core_id %d, sizing for %d cores per package\n", max_core_id, topo.num_cores_per_pkg); - if (debug && !summary_only && topo.num_cores_per_pkg > 1) - show_core = 1; + if (!summary_only && topo.num_cores_per_pkg > 1) + BIC_PRESENT(BIC_Core); topo.num_packages = max_package_id + 1; if (debug > 1) fprintf(outf, "max_package_id %d, sizing for %d packages\n", max_package_id, topo.num_packages); - if (debug && !summary_only && topo.num_packages > 1) - show_pkg = 1; + if (!summary_only && topo.num_packages > 1) + BIC_PRESENT(BIC_Package); topo.num_threads_per_core = max_siblings; if (debug > 1) @@ -3662,7 +4360,7 @@ allocate_counters(struct thread_data **t, struct core_data **c, struct pkg_data int i; *t = calloc(topo.num_threads_per_core * topo.num_cores_per_pkg * - topo.num_packages, sizeof(struct thread_data) + sys.thread_counter_bytes); + topo.num_packages, sizeof(struct thread_data)); if (*t == NULL) goto error; @@ -3671,14 +4369,14 @@ allocate_counters(struct thread_data **t, struct core_data **c, struct pkg_data (*t)[i].cpu_id = -1; *c = calloc(topo.num_cores_per_pkg * topo.num_packages, - sizeof(struct core_data) + sys.core_counter_bytes); + sizeof(struct core_data)); if (*c == NULL) goto error; for (i = 0; i < topo.num_cores_per_pkg * topo.num_packages; i++) (*c)[i].core_id = -1; - *p = calloc(topo.num_packages, sizeof(struct pkg_data) + sys.package_counter_bytes); + *p = calloc(topo.num_packages, sizeof(struct pkg_data)); if (*p == NULL) goto error; @@ -3789,24 +4487,24 @@ void turbostat_init() process_cpuid(); - if (debug) + if (!quiet) for_all_cpus(print_hwp, ODD_COUNTERS); - if (debug) + if (!quiet) for_all_cpus(print_epb, ODD_COUNTERS); - if (debug) + if (!quiet) for_all_cpus(print_perf_limit, ODD_COUNTERS); - if (debug) + if (!quiet) for_all_cpus(print_rapl, ODD_COUNTERS); for_all_cpus(set_temperature_target, ODD_COUNTERS); - if (debug) + if (!quiet) for_all_cpus(print_thermal, ODD_COUNTERS); - if (debug && do_irtl_snb) + if (!quiet && do_irtl_snb) print_irtl(); } @@ -3815,6 +4513,7 @@ int fork_it(char **argv) pid_t child_pid; int status; + snapshot_proc_sysfs_files(); status = for_all_cpus(get_counters, EVEN_COUNTERS); if (status) exit(status); @@ -3826,6 +4525,7 @@ int fork_it(char **argv) if (!child_pid) { /* child */ execvp(argv[0], argv); + err(errno, "exec %s", argv[0]); } else { /* parent */ @@ -3841,6 +4541,7 @@ int fork_it(char **argv) * n.b. fork_it() does not check for errors from for_all_cpus() * because re-starting is problematic when forking */ + snapshot_proc_sysfs_files(); for_all_cpus(get_counters, ODD_COUNTERS); gettimeofday(&tv_odd, (struct timezone *)NULL); timersub(&tv_odd, &tv_even, &tv_delta); @@ -3862,6 +4563,7 @@ int get_and_dump_counters(void) { int status; + snapshot_proc_sysfs_files(); status = for_all_cpus(get_counters, ODD_COUNTERS); if (status) return status; @@ -3876,13 +4578,13 @@ int get_and_dump_counters(void) } void print_version() { - fprintf(outf, "turbostat version 4.16 24 Dec 2016" + fprintf(outf, "turbostat version 17.04.12" " - Len Brown <lenb@kernel.org>\n"); } -int add_counter(unsigned int msr_num, char *name, unsigned int width, - enum counter_scope scope, enum counter_type type, - enum counter_format format) +int add_counter(unsigned int msr_num, char *path, char *name, + unsigned int width, enum counter_scope scope, + enum counter_type type, enum counter_format format, int flags) { struct msr_counter *msrp; @@ -3894,31 +4596,46 @@ int add_counter(unsigned int msr_num, char *name, unsigned int width, msrp->msr_num = msr_num; strncpy(msrp->name, name, NAME_BYTES); + if (path) + strncpy(msrp->path, path, PATH_BYTES); msrp->width = width; msrp->type = type; msrp->format = format; + msrp->flags = flags; switch (scope) { case SCOPE_CPU: - sys.thread_counter_bytes += 64; msrp->next = sys.tp; sys.tp = msrp; - sys.thread_counter_bytes += sizeof(unsigned long long); + sys.added_thread_counters++; + if (sys.added_thread_counters > MAX_ADDED_COUNTERS) { + fprintf(stderr, "exceeded max %d added thread counters\n", + MAX_ADDED_COUNTERS); + exit(-1); + } break; case SCOPE_CORE: - sys.core_counter_bytes += 64; msrp->next = sys.cp; sys.cp = msrp; - sys.core_counter_bytes += sizeof(unsigned long long); + sys.added_core_counters++; + if (sys.added_core_counters > MAX_ADDED_COUNTERS) { + fprintf(stderr, "exceeded max %d added core counters\n", + MAX_ADDED_COUNTERS); + exit(-1); + } break; case SCOPE_PACKAGE: - sys.package_counter_bytes += 64; msrp->next = sys.pp; sys.pp = msrp; - sys.package_counter_bytes += sizeof(unsigned long long); + sys.added_package_counters++; + if (sys.added_package_counters > MAX_ADDED_COUNTERS) { + fprintf(stderr, "exceeded max %d added package counters\n", + MAX_ADDED_COUNTERS); + exit(-1); + } break; } @@ -3928,7 +4645,8 @@ int add_counter(unsigned int msr_num, char *name, unsigned int width, void parse_add_command(char *add_command) { int msr_num = 0; - char name_buffer[NAME_BYTES]; + char *path = NULL; + char name_buffer[NAME_BYTES] = ""; int width = 64; int fail = 0; enum counter_scope scope = SCOPE_CPU; @@ -3943,6 +4661,11 @@ void parse_add_command(char *add_command) if (sscanf(add_command, "msr%d", &msr_num) == 1) goto next; + if (*add_command == '/') { + path = add_command; + goto next; + } + if (sscanf(add_command, "u%d", &width) == 1) { if ((width == 32) || (width == 64)) goto next; @@ -3968,6 +4691,10 @@ void parse_add_command(char *add_command) type = COUNTER_SECONDS; goto next; } + if (!strncmp(add_command, "usec", strlen("usec"))) { + type = COUNTER_USEC; + goto next; + } if (!strncmp(add_command, "raw", strlen("raw"))) { format = FORMAT_RAW; goto next; @@ -3992,36 +4719,26 @@ void parse_add_command(char *add_command) next: add_command = strchr(add_command, ','); - if (add_command) + if (add_command) { + *add_command = '\0'; add_command++; + } } - if (msr_num == 0) { - fprintf(stderr, "--add: (msrDDD | msr0xXXX) required\n"); + if ((msr_num == 0) && (path == NULL)) { + fprintf(stderr, "--add: (msrDDD | msr0xXXX | /path_to_counter ) required\n"); fail++; } /* generate default column header */ if (*name_buffer == '\0') { - if (format == FORMAT_RAW) { - if (width == 32) - sprintf(name_buffer, "msr%d", msr_num); - else - sprintf(name_buffer, "MSR%d", msr_num); - } else if (format == FORMAT_DELTA) { - if (width == 32) - sprintf(name_buffer, "cnt%d", msr_num); - else - sprintf(name_buffer, "CNT%d", msr_num); - } else if (format == FORMAT_PERCENT) { - if (width == 32) - sprintf(name_buffer, "msr%d%%", msr_num); - else - sprintf(name_buffer, "MSR%d%%", msr_num); - } + if (width == 32) + sprintf(name_buffer, "M0x%x%s", msr_num, format == FORMAT_PERCENT ? "%" : ""); + else + sprintf(name_buffer, "M0X%x%s", msr_num, format == FORMAT_PERCENT ? "%" : ""); } - if (add_counter(msr_num, name_buffer, width, scope, type, format)) + if (add_counter(msr_num, path, name_buffer, width, scope, type, format, 0)) fail++; if (fail) { @@ -4029,20 +4746,214 @@ next: exit(1); } } + +int is_deferred_skip(char *name) +{ + int i; + + for (i = 0; i < deferred_skip_index; ++i) + if (!strcmp(name, deferred_skip_names[i])) + return 1; + return 0; +} + +void probe_sysfs(void) +{ + char path[64]; + char name_buf[16]; + FILE *input; + int state; + char *sp; + + if (!DO_BIC(BIC_sysfs)) + return; + + for (state = 10; state > 0; --state) { + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", + base_cpu, state); + input = fopen(path, "r"); + if (input == NULL) + continue; + fgets(name_buf, sizeof(name_buf), input); + + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + sp = strchr(name_buf, '-'); + if (!sp) + sp = strchrnul(name_buf, '\n'); + *sp = '%'; + *(sp + 1) = '\0'; + + fclose(input); + + sprintf(path, "cpuidle/state%d/time", state); + + if (is_deferred_skip(name_buf)) + continue; + + add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_USEC, + FORMAT_PERCENT, SYSFS_PERCPU); + } + + for (state = 10; state > 0; --state) { + + sprintf(path, "/sys/devices/system/cpu/cpu%d/cpuidle/state%d/name", + base_cpu, state); + input = fopen(path, "r"); + if (input == NULL) + continue; + fgets(name_buf, sizeof(name_buf), input); + /* truncate "C1-HSW\n" to "C1", or truncate "C1\n" to "C1" */ + sp = strchr(name_buf, '-'); + if (!sp) + sp = strchrnul(name_buf, '\n'); + *sp = '\0'; + fclose(input); + + sprintf(path, "cpuidle/state%d/usage", state); + + if (is_deferred_skip(name_buf)) + continue; + + add_counter(0, path, name_buf, 64, SCOPE_CPU, COUNTER_ITEMS, + FORMAT_DELTA, SYSFS_PERCPU); + } + +} + + +/* + * parse cpuset with following syntax + * 1,2,4..6,8-10 and set bits in cpu_subset + */ +void parse_cpu_command(char *optarg) +{ + unsigned int start, end; + char *next; + + if (!strcmp(optarg, "core")) { + if (cpu_subset) + goto error; + show_core_only++; + return; + } + if (!strcmp(optarg, "package")) { + if (cpu_subset) + goto error; + show_pkg_only++; + return; + } + if (show_core_only || show_pkg_only) + goto error; + + cpu_subset = CPU_ALLOC(CPU_SUBSET_MAXCPUS); + if (cpu_subset == NULL) + err(3, "CPU_ALLOC"); + cpu_subset_size = CPU_ALLOC_SIZE(CPU_SUBSET_MAXCPUS); + + CPU_ZERO_S(cpu_subset_size, cpu_subset); + + next = optarg; + + while (next && *next) { + + if (*next == '-') /* no negative cpu numbers */ + goto error; + + start = strtoul(next, &next, 10); + + if (start >= CPU_SUBSET_MAXCPUS) + goto error; + CPU_SET_S(start, cpu_subset_size, cpu_subset); + + if (*next == '\0') + break; + + if (*next == ',') { + next += 1; + continue; + } + + if (*next == '-') { + next += 1; /* start range */ + } else if (*next == '.') { + next += 1; + if (*next == '.') + next += 1; /* start range */ + else + goto error; + } + + end = strtoul(next, &next, 10); + if (end <= start) + goto error; + + while (++start <= end) { + if (start >= CPU_SUBSET_MAXCPUS) + goto error; + CPU_SET_S(start, cpu_subset_size, cpu_subset); + } + + if (*next == ',') + next += 1; + else if (*next != '\0') + goto error; + } + + return; + +error: + fprintf(stderr, "\"--cpu %s\" malformed\n", optarg); + help(); + exit(-1); +} + +int shown; +/* + * parse_show_hide() - process cmdline to set default counter action + */ +void parse_show_hide(char *optarg, enum show_hide_mode new_mode) +{ + /* + * --show: show only those specified + * The 1st invocation will clear and replace the enabled mask + * subsequent invocations can add to it. + */ + if (new_mode == SHOW_LIST) { + if (shown == 0) + bic_enabled = bic_lookup(optarg, new_mode); + else + bic_enabled |= bic_lookup(optarg, new_mode); + shown = 1; + + return; + } + + /* + * --hide: do not show those specified + * multiple invocations simply clear more bits in enabled mask + */ + bic_enabled &= ~bic_lookup(optarg, new_mode); + +} + void cmdline(int argc, char **argv) { int opt; int option_index = 0; static struct option long_options[] = { {"add", required_argument, 0, 'a'}, + {"cpu", required_argument, 0, 'c'}, {"Dump", no_argument, 0, 'D'}, - {"debug", no_argument, 0, 'd'}, + {"debug", no_argument, 0, 'd'}, /* internal, not documented */ {"interval", required_argument, 0, 'i'}, {"help", no_argument, 0, 'h'}, + {"hide", required_argument, 0, 'H'}, // meh, -h taken by --help {"Joules", no_argument, 0, 'J'}, + {"list", no_argument, 0, 'l'}, {"out", required_argument, 0, 'o'}, - {"Package", no_argument, 0, 'p'}, - {"processor", no_argument, 0, 'p'}, + {"quiet", no_argument, 0, 'q'}, + {"show", required_argument, 0, 's'}, {"Summary", no_argument, 0, 'S'}, {"TCC", required_argument, 0, 'T'}, {"version", no_argument, 0, 'v' }, @@ -4051,18 +4962,24 @@ void cmdline(int argc, char **argv) progname = argv[0]; - while ((opt = getopt_long_only(argc, argv, "+C:c:Ddhi:JM:m:o:PpST:v", + while ((opt = getopt_long_only(argc, argv, "+C:c:Ddhi:JM:m:o:qST:v", long_options, &option_index)) != -1) { switch (opt) { case 'a': parse_add_command(optarg); break; + case 'c': + parse_cpu_command(optarg); + break; case 'D': dump_only++; break; case 'd': debug++; break; + case 'H': + parse_show_hide(optarg, HIDE_LIST); + break; case 'h': default: help(); @@ -4084,14 +5001,18 @@ void cmdline(int argc, char **argv) case 'J': rapl_joules++; break; + case 'l': + list_header_only++; + quiet++; + break; case 'o': outf = fopen_or_die(optarg, "w"); break; - case 'P': - show_pkg_only++; + case 'q': + quiet = 1; break; - case 'p': - show_core_only++; + case 's': + parse_show_hide(optarg, SHOW_LIST); break; case 'S': summary_only++; @@ -4113,15 +5034,24 @@ int main(int argc, char **argv) cmdline(argc, argv); - if (debug) + if (!quiet) print_version(); + probe_sysfs(); + turbostat_init(); /* dump counters and exit */ if (dump_only) return get_and_dump_counters(); + /* list header and exit */ + if (list_header_only) { + print_header(","); + flush_output_stdout(); + return 0; + } + /* * if any params left, it must be a command to fork */ |