Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
tree: 69964ea4c7
Fetching contributors…

Octocat-spinner-32-eaf2f5

Cannot retrieve contributors at this time

file 114 lines (87 sloc) 3.11 kb
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
CPU load
--------

Linux exports various bits of information via `/proc/stat' and
`/proc/uptime' that userland tools, such as top(1), use to calculate
the average time system spent in a particular state, for example:

    $ iostat
    Linux 2.6.18.3-exp (linmac) 02/20/2007

    avg-cpu: %user %nice %system %iowait %steal %idle
              10.01 0.00 2.92 5.44 0.00 81.63

    ...

Here the system thinks that over the default sampling period the
system spent 10.01% of the time doing work in user space, 2.92% in the
kernel, and was overall 81.63% of the time idle.

In most cases the `/proc/stat' information reflects the reality quite
closely, however due to the nature of how/when the kernel collects
this data sometimes it can not be trusted at all.

So how is this information collected? Whenever timer interrupt is
signalled the kernel looks what kind of task was running at this
moment and increments the counter that corresponds to this tasks
kind/state. The problem with this is that the system could have
switched between various states multiple times between two timer
interrupts yet the counter is incremented only for the last state.


Example
-------

If we imagine the system with one task that periodically burns cycles
in the following manner:

 time line between two timer interrupts
|--------------------------------------|
 ^ ^
 |_ something begins working |
                                      |_ something goes to sleep
                                     (only to be awaken quite soon)

In the above situation the system will be 0% loaded according to the
`/proc/stat' (since the timer interrupt will always happen when the
system is executing the idle handler), but in reality the load is
closer to 99%.

One can imagine many more situations where this behavior of the kernel
will lead to quite erratic information inside `/proc/stat'.


/* gcc -o hog smallhog.c */
#include <time.h>
#include <limits.h>
#include <signal.h>
#include <sys/time.h>
#define HIST 10

static volatile sig_atomic_t stop;

static void sighandler (int signr)
{
     (void) signr;
     stop = 1;
}
static unsigned long hog (unsigned long niters)
{
     stop = 0;
     while (!stop && --niters);
     return niters;
}
int main (void)
{
     int i;
     struct itimerval it = { .it_interval = { .tv_sec = 0, .tv_usec = 1 },
                             .it_value = { .tv_sec = 0, .tv_usec = 1 } };
     sigset_t set;
     unsigned long v[HIST];
     double tmp = 0.0;
     unsigned long n;
     signal (SIGALRM, &sighandler);
     setitimer (ITIMER_REAL, &it, NULL);

     hog (ULONG_MAX);
     for (i = 0; i < HIST; ++i) v[i] = ULONG_MAX - hog (ULONG_MAX);
     for (i = 0; i < HIST; ++i) tmp += v[i];
     tmp /= HIST;
     n = tmp - (tmp / 3.0);

     sigemptyset (&set);
     sigaddset (&set, SIGALRM);

     for (;;) {
         hog (n);
         sigwait (&set, &i);
     }
     return 0;
}


References
----------

http://lkml.org/lkml/2007/2/12/6
Documentation/filesystems/proc.txt (1.8)


Thanks
------

Con Kolivas, Pavel Machek
Something went wrong with that request. Please try again.