summaryrefslogtreecommitdiffstats
path: root/tools/power
diff options
context:
space:
mode:
authorThomas Renninger <trenn@suse.de>2012-11-27 13:17:48 +0100
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>2012-11-27 23:07:20 +0100
commitc8cfc3c6bf404b0f110631d5bba234982e6ad24f (patch)
treedd6e71768799b81cca6e78fb0a26a45526a5af85 /tools/power
parentcpupower tools: Fix warning and a bug with the cpu package count (diff)
downloadlinux-c8cfc3c6bf404b0f110631d5bba234982e6ad24f.tar.xz
linux-c8cfc3c6bf404b0f110631d5bba234982e6ad24f.zip
cpupower: Provide -c param for cpupower monitor to schedule process on all cores
If an MSR based monitor is run in parallel this is not needed. This is the default case on all/most Intel machines. But when only sysfs info is read via cpupower monitor -m Idle_Stats (typically the case for non root users) or when other monitors are PCI based (AMD), Idle_Stats, read from sysfs can be totally bogus: cpupower monitor -m Idle_Stats PKG |CORE|CPU | POLL | C1-N | C3-N | C6-N 0| 0| 0| 0.00| 0.00| 0.24| 99.81 0| 0| 32| 0.00| 0.00| 0.00| 100.7 ... 0| 17| 20| 0.00| 0.00| 0.00| 173.1 0| 17| 52| 0.00| 0.00| 0.07| 173.0 0| 18| 68| 0.00| 0.00| 0.00| 0.00 0| 18| 76| 0.00| 0.00| 0.00| 0.00 ... With the -c option all cores are woken up and the kernel did update cpuidle statistics before reading out sysfs. This causes some overhead. Therefore avoid if possible, use if needed: cpupower monitor -c -m Idle_Stats PKG |CORE|CPU | POLL | C1-N | C3-N | C6-N 0| 0| 0| 0.00| 0.00| 0.00| 100.2 0| 0| 32| 0.00| 0.00| 0.00| 100.2 ... 0| 8| 8| 0.00| 0.00| 0.00| 99.82 0| 8| 40| 0.00| 0.00| 0.00| 99.81 0| 9| 24| 0.00| 0.00| 0.00| 100.3 0| 9| 56| 0.00| 0.00| 0.00| 100.2 0| 16| 4| 0.00| 0.00| 0.00| 99.75 0| 16| 36| 0.00| 0.00| 0.00| 99.38 ... Signed-off-by: Thomas Renninger <trenn@suse.de> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Diffstat (limited to 'tools/power')
-rw-r--r--tools/power/cpupower/man/cpupower-monitor.115
-rw-r--r--tools/power/cpupower/utils/helpers/helpers.h1
-rw-r--r--tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c18
-rw-r--r--tools/power/cpupower/utils/idle_monitor/cpupower-monitor.h17
4 files changed, 48 insertions, 3 deletions
diff --git a/tools/power/cpupower/man/cpupower-monitor.1 b/tools/power/cpupower/man/cpupower-monitor.1
index 1141c2073719..e01c35d13b6e 100644
--- a/tools/power/cpupower/man/cpupower-monitor.1
+++ b/tools/power/cpupower/man/cpupower-monitor.1
@@ -7,11 +7,11 @@ cpupower\-monitor \- Report processor frequency and idle statistics
.RB "\-l"
.B cpupower monitor
-.RB [ "\-m <mon1>," [ "<mon2>,..." ] ]
+.RB [ -c ] [ "\-m <mon1>," [ "<mon2>,..." ] ]
.RB [ "\-i seconds" ]
.br
.B cpupower monitor
-.RB [ "\-m <mon1>," [ "<mon2>,..." ] ]
+.RB [ -c ][ "\-m <mon1>," [ "<mon2>,..." ] ]
.RB command
.br
.SH DESCRIPTION
@@ -64,6 +64,17 @@ Only display specific monitors. Use the monitor string(s) provided by \-l option
Measure intervall.
.RE
.PP
+\-c
+.RS 4
+Schedule the process on every core before starting and ending measuring.
+This could be needed for the Idle_Stats monitor when no other MSR based
+monitor (has to be run on the core that is measured) is run in parallel.
+This is to wake up the processors from deeper sleep states and let the
+kernel re
+-account its cpuidle (C-state) information before reading the
+cpuidle timings from sysfs.
+.RE
+.PP
command
.RS 4
Measure idle and frequency characteristics of an arbitrary command/workload.
diff --git a/tools/power/cpupower/utils/helpers/helpers.h b/tools/power/cpupower/utils/helpers/helpers.h
index f84985f630e2..aa9e95486a2d 100644
--- a/tools/power/cpupower/utils/helpers/helpers.h
+++ b/tools/power/cpupower/utils/helpers/helpers.h
@@ -114,6 +114,7 @@ struct cpupower_topology {
extern int get_cpu_topology(struct cpupower_topology *cpu_top);
extern void cpu_topology_release(struct cpupower_topology cpu_top);
+
/* CPU topology/hierarchy parsing ******************/
/* X86 ONLY ****************************************/
diff --git a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
index 7a657f3da23b..c4bae9203a69 100644
--- a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
+++ b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.c
@@ -39,6 +39,7 @@ static int mode;
static int interval = 1;
static char *show_monitors_param;
static struct cpupower_topology cpu_top;
+static unsigned int wake_cpus;
/* ToDo: Document this in the manpage */
static char range_abbr[RANGE_MAX] = { 'T', 'C', 'P', 'M', };
@@ -314,16 +315,28 @@ int fork_it(char **argv)
int do_interval_measure(int i)
{
unsigned int num;
+ int cpu;
+
+ if (wake_cpus)
+ for (cpu = 0; cpu < cpu_count; cpu++)
+ bind_cpu(cpu);
for (num = 0; num < avail_monitors; num++) {
dprint("HW C-state residency monitor: %s - States: %d\n",
monitors[num]->name, monitors[num]->hw_states_num);
monitors[num]->start();
}
+
sleep(i);
+
+ if (wake_cpus)
+ for (cpu = 0; cpu < cpu_count; cpu++)
+ bind_cpu(cpu);
+
for (num = 0; num < avail_monitors; num++)
monitors[num]->stop();
+
return 0;
}
@@ -332,7 +345,7 @@ static void cmdline(int argc, char *argv[])
int opt;
progname = basename(argv[0]);
- while ((opt = getopt(argc, argv, "+li:m:")) != -1) {
+ while ((opt = getopt(argc, argv, "+lci:m:")) != -1) {
switch (opt) {
case 'l':
if (mode)
@@ -351,6 +364,9 @@ static void cmdline(int argc, char *argv[])
mode = show;
show_monitors_param = optarg;
break;
+ case 'c':
+ wake_cpus = 1;
+ break;
default:
print_wrong_arg_exit();
}
diff --git a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.h b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.h
index 9312ee1f2dbc..9e43f3371fbc 100644
--- a/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.h
+++ b/tools/power/cpupower/utils/idle_monitor/cpupower-monitor.h
@@ -65,4 +65,21 @@ extern long long timespec_diff_us(struct timespec start, struct timespec end);
"could be inaccurate\n"), mes, ov); \
}
+
+/* Taken over from x86info project sources -> return 0 on success */
+#include <sched.h>
+#include <sys/types.h>
+#include <unistd.h>
+static inline int bind_cpu(int cpu)
+{
+ cpu_set_t set;
+
+ if (sched_getaffinity(getpid(), sizeof(set), &set) == 0) {
+ CPU_ZERO(&set);
+ CPU_SET(cpu, &set);
+ return sched_setaffinity(getpid(), sizeof(set), &set);
+ }
+ return 1;
+}
+
#endif /* __CPUIDLE_INFO_HW__ */