Edit

Interrupt Time

  • Article

Interrupt time is the amount of time since the system was last started, in 100-nanosecond intervals. The interrupt-time count begins at zero when the system starts and is incremented at each clock interrupt by the length of a clock tick. The exact length of a clock tick depends on underlying hardware and can vary between systems.

Unlike system time, the interrupt-time count is not subject to adjustments by users or the Windows time service, making it a better choice for measuring short durations. Applications that require greater precision than the interrupt-time count should use a high-resolution timer. Use the QueryPerformanceFrequency function to retrieve the frequency of the high-resolution timer and the QueryPerformanceCounter function to retrieve the counter's value.

The QueryInterruptTime, QueryInterruptTimePrecise, QueryUnbiasedInterruptTime, and QueryUnbiasedInterruptTimePrecise functions can be used to retrieve the interrupt-time count. Unbiased interrupt-time means that only time that the system is in the working state is counted—therefore, the interrupt-time count is not "biased" by time the system spends in sleep or hibernation.

Windows Server 2008, Windows Vista, Windows Server 2003 and Windows XP/2000: The QueryUnbiasedInterruptTime function is available starting with Windows 7 and Windows Server 2008 R2.

The timer resolution set by the timeBeginPeriod and timeEndPeriod functions affects the resolution of the QueryInterruptTime and QueryUnbiasedInterruptTime functions. However, increasing the timer resolution is not recommended because it can reduce overall system performance and increase power consumption by preventing the processor from entering power-saving states. Instead, applications should use a high-resolution timer.

Note

The QueryInterruptTime, QueryInterruptTimePrecise, QueryUnbiasedInterruptTime, and QueryUnbiasedInterruptTimePrecise functions produces different results on debug ("checked") builds of Windows, because the interrupt-time count and tick count are advanced by approximately 49 days. This helps to identify bugs that might not occur until the system has been running for a long time. The checked build is available to MSDN subscribers through the Microsoft Developer Network (MSDN) Web site.

 

The following example shows how to retrieve the interrupt-time count by calling the QueryInterruptTime, QueryInterruptTimePrecise, QueryUnbiasedInterruptTime, and QueryUnbiasedInterruptTimePrecise functions. Link to the OneCore.lib library when you build a console application that calls these functions.

#include "stdafx.h"
#include <windows.h>
#include <realtimeapiset.h>

void InterruptTimeTest()
{
    ULONGLONG InterruptTime;
    ULONGLONG PreciseInterruptTime;
    ULONGLONG UnbiasedInterruptTime;
    ULONGLONG PreciseUnbiasedInterruptTime;

        // The interrupt time that QueryInterruptTime reports is based on the 
        // latest tick of the system clock timer. The system clock timer is 
        // the hardware timer that periodically generates interrupts for the 
        // system clock. The uniform period between system clock timer 
        // interrupts is referred to as a system clock tick, and is typically 
        // in the range of 0.5 milliseconds to 15.625 milliseconds, depending 
        // on the hardware platform. The interrupt time value retrieved by 
        // QueryInterruptTime is accurate within a system clock tick.

    QueryInterruptTime(&InterruptTime);
    printf("Interrupt time: %.7f seconds\n", 
            (double)InterruptTime/(double)10000000);

        // Precise interrupt time is more precise than the interrupt time that
        // QueryInterruptTime reports because the functions that report  
        // precise interrupt time read the timer hardware directly.

    QueryInterruptTimePrecise(&PreciseInterruptTime);
    printf("Precise interrupt time: %.7f seconds\n", 
            (double)PreciseInterruptTime/(double)10000000);

        // Unbiased interrupt time means that only time that the system is in 
        // the working state is counted. Therefore, the interrupt-time count 
        // is not biased by time the system spends in sleep or hibernation.

    QueryUnbiasedInterruptTime(&UnbiasedInterruptTime);
    printf("Unbiased interrupt time: %.7f seconds\n", 
            (double)UnbiasedInterruptTime/(double)10000000);

        // QueryUnbiasedInterruptTimePrecise gets an interrupt-time count
        // that is both unbiased and precise, as defined in the comments
        // included earlier in this example.

    QueryUnbiasedInterruptTimePrecise(&PreciseUnbiasedInterruptTime);
    printf("Precise unbiased interrupt time: %.7f seconds\n", 
            (double)PreciseUnbiasedInterruptTime/(double)10000000);

}

int main(void)
{
    void InterruptTimeTime();

    InterruptTimeTest();
    return(0);
}

To retrieve elapsed time that accounts for sleep or hibernation, use the GetTickCount or GetTickCount64 function, or use the System Up Time performance counter. This performance counter can be retrieved from the performance data in the registry key HKEY_PERFORMANCE_DATA. The value returned is an 8-byte value. For more information, see Performance Counters.

QueryInterruptTime

QueryInterruptTimePrecise

QueryUnbiasedInterruptTime

QueryUnbiasedInterruptTimePrecise

timeBeginPeriod

timeEndPeriod