Microsoft MVP성태의 닷넷 이야기
글쓴 사람
정성태 (techsharer at outlook.com)
홈페이지
첨부 파일
 
(연관된 글이 2개 있습니다.)

Windows 10부터 바뀐 QueryPerformanceFrequency, QueryPerformanceCounter

예전 글에서,

윈도우 운영체제의 시간 함수 (5) - TSC(Time Stamp Counter)와 QueryPerformanceCounter
; https://www.sysnet.pe.kr/2/0/11068

제가 QueryPerformanceFrequency는 초당 틱 수를 반환한다고 했고, QueryPerformanceCounter는 그 틱의 단위가 적용된다고 했습니다.

재미있는 것은, 그 당시 QueryPerformanceFrequency는 (시스템마다 다를 수 있는) 3,328,129라는 값을 반환했는데요, 최근에 다시 소스 코드를 돌려 보니,

using System;
using System.Runtime.InteropServices;

internal class Program
{
    [DllImport("Kernel32.dll")]
    static extern bool QueryPerformanceFrequency(out long lpFrequency);

    static void Main(string[] args)
    {
        long lpFrequency;
        QueryPerformanceFrequency(out lpFrequency);
        Console.WriteLine($"QueryPerformanceFrequency == {lpFrequency}");
    }
}

고정적으로 10,000,000 값이 나옵니다. ^^; 검색 결과 이런 글이 나오는군요.

QueryPerformanceFrequency returns 10mhz on Windows 10 Build 1809+, causing performance loss/increased jitter/latency
; https://answers.microsoft.com/en-us/windows/forum/all/queryperformancefrequency-returns-10mhz-on-windows/44946807-5355-4b36-ba3e-43aa86ce30c0?auth=1

그러니까, Windows 10 build 1809부터 고정적으로 저 값이 나오도록 바뀐 것입니다. 아마도 frequency의 값을 운영체제 차원에서 정규화를 시도해 보정한 값이 나오는 듯합니다.

따라서, 이제는 QueryPerformanceFrequency에서 고정적으로 10,000,000 값을 반환해 주기 때문에 QueryPerformanceCounter의 1차이는,

T = 1 / 10,000,000 = 0.0000001초
                   = 0.0001밀리초
                   = 0.1마이크로초
                   = 100나노초

100나노초 단위가 된 것입니다. 이런 의미에서 봤을 때 다음의 Win32 API들이,

윈도우 운영체제의 시간 함수 (3) - QueryInterruptTimePrecise, QueryInterruptTime 함수
; https://www.sysnet.pe.kr/2/0/11066

문서에서 100나노초라고 일관되게 정한 것이 이해가 갑니다. ^^

lpInterruptTimePrecise [out]

A pointer to a ULONGLONG in which to receive the interrupt-time count in system time units of 100 nanoseconds. Divide by ten million, or 1e7, to get seconds (there are 1e9 nanoseconds in a second, so there are 1e7 100-nanoseconds in a second).


개인적으로는, 이런 변화는 나쁘지 않다고 봅니다. 향후에는 더 이상 QueryPerformanceFrequency 값을 구할 필요 없이 10,000,000으로 고정해도 되고, QueryPerformanceCounter의 단위를 100나노초라고 가정하고 계산할 수 있기 때문입니다.

단지, 한동안은 Windows 10 build 1809 미만의 컴퓨터도 고려를 해야 한다면 QueryPerformanceFrequency에 따른 계산을 병행할 필요는 있겠습니다. ^^




그런데, 실제로 테스트를 해보면 Windows 10뿐만 아니라 (VM에서 테스트했지만) Windows 7 x64, Windows Server 2012 R2에서도 동일하게 QueryPerformanceFrequency == 10,000,000 값이 나옵니다.

굳이 짜 맞춰 본다면, 저 당시에 Windows 10을 필두로 하위 운영체제에도 점차로 적용된 것이 아닌가... 예상하는데요, 그렇다고는 하지만 이것을 100%라고 가정할 수는 없으므로 쉽사리 10,000,000 값을 고정적으로 쓰기에는 다소 불안한 면이 있습니다. ^^




그러고 보니, 최근에 썼던 마이크로초 단위의 sleep 기능을 설명한 코드가,

C# - 윈도우 환경에서 usleep을 호출하는 방법
; https://www.sysnet.pe.kr/2/0/12980

이랬는데,

static internal class NativeMethods
{
    [DllImport("Kernel32.dll")]
    static extern bool QueryPerformanceCounter(out long lpPerformanceCount);

    public static void uSleep(long waitTime) 
    {
        long time1 = 0, time2 = 0;

        QueryPerformanceCounter(out time1);

        do
        {
            QueryPerformanceCounter(out time2);
        } while ((time2 - time1) < waitTime);
    }
}

왜 제가 QueryPerformanceCounter의 반환값을 마이크로초 단위라고 가정했을까요? ^^; 이것을 범용적으로 만들려면 이렇게 작성해야 하고,

static internal class NativeMethods
{
    [DllImport("Kernel32.dll")]
    static extern bool QueryPerformanceCounter(out long lpPerformanceCount);

    [DllImport("Kernel32.dll")]
    static extern bool QueryPerformanceFrequency(out long ticksPerSecond);

    static long _ticksPerMicroSeconds;

    static NativeMethods()
    {
        long ticksPerSecond = 0;
        QueryPerformanceFrequency(out ticksPerSecond);
        _ticksPerMicroSeconds = ticksPerSecond / 1000 / 1000;
    }

    public static void uSleep(long waitTime)
    {
        long utime1 = 0, utime2 = 0;

        QueryPerformanceCounter(out utime1);
        utime1 = utime1 / _ticksPerMicroSeconds;

        do
        {
            QueryPerformanceCounter(out utime2);
            utime2 = utime2 / _ticksPerMicroSeconds;
        } while ((utime2 - utime1) < waitTime);
    }
}

Windows 10 이상의 PC에서 동작하는 것을 가정한다면 다음과 같이 만들어도 됩니다.

static internal class NativeMethods
{
    [DllImport("Kernel32.dll")]
    static extern bool QueryPerformanceCounter(out long lpPerformanceCount);

    public static void uSleep(long waitTime)
    {
        waitTime = waitTime * 10; // 100나노초 단위를 갖는 QueryPerformanceCounter 값과 비교하기 위해!
        long time1 = 0, time2 = 0;

        QueryPerformanceCounter(out time1);

        do
        {
            QueryPerformanceCounter(out time2);
        } while ((time2 - time1) < waitTime);
    }
}




[이 글에 대해서 여러분들과 의견을 공유하고 싶습니다. 틀리거나 미흡한 부분 또는 의문 사항이 있으시면 언제든 댓글 남겨주십시오.]

[연관 글]






[최초 등록일: ]
[최종 수정일: 4/22/2022]

Creative Commons License
이 저작물은 크리에이티브 커먼즈 코리아 저작자표시-비영리-변경금지 2.0 대한민국 라이센스에 따라 이용하실 수 있습니다.
by SeongTae Jeong, mailto:techsharer at outlook.com

비밀번호

댓글 작성자
 




1  [2]  3  4  5  6  7  8  9  10  11  12  13  14  15  ...
NoWriterDateCnt.TitleFile(s)
13580정성태3/15/20241165닷넷: 2231. C# - ReceiveTimeout, SendTimeout이 적용되지 않는 Socket await 비동기 호출파일 다운로드1
13579정성태3/13/20241495오류 유형: 899. HTTP Error 500.32 - ANCM Failed to Load dll
13578정성태3/11/20241635닷넷: 2230. C# - 덮어쓰기 가능한 환형 큐 (Circular queue)파일 다운로드1
13577정성태3/9/20241889닷넷: 2229. C# - 닷넷을 위한 난독화 도구 소개 (예: ConfuserEx)
13576정성태3/8/20241545닷넷: 2228. .NET Profiler - IMetaDataEmit2::DefineMethodSpec 사용법
13575정성태3/7/20241682닷넷: 2227. 최신 C# 문법을 .NET Framework 프로젝트에 쓸 수 있을까요?
13574정성태3/6/20241562닷넷: 2226. C# - "Docker Desktop for Windows" Container 환경에서의 IPv6 DualMode 소켓
13573정성태3/5/20241570닷넷: 2225. Windbg - dumasync로 분석하는 async/await 호출
13572정성태3/4/20241648닷넷: 2224. C# - WPF의 Dispatcher Queue로 알아보는 await 호출의 hang 현상파일 다운로드1
13571정성태3/1/20241630닷넷: 2223. C# - await 호출과 WPF의 Dispatcher Queue 동작 확인파일 다운로드1
13570정성태2/29/20241638닷넷: 2222. C# - WPF의 Dispatcher Queue 동작 확인파일 다운로드1
13569정성태2/28/20241546닷넷: 2221. C# - LoadContext, LoadFromContext 그리고 GAC파일 다운로드1
13568정성태2/27/20241611닷넷: 2220. C# - .NET Framework 프로세스의 LoaderOptimization 설정을 확인하는 방법파일 다운로드1
13567정성태2/27/20241620오류 유형: 898. .NET Framework 3.5 이하에서 mscoree.tlb 참조 시 System.BadImageFormatException파일 다운로드1
13566정성태2/27/20241632오류 유형: 897. Windows 7 SDK 설치 시 ".NET Development" 옵션이 비활성으로 선택이 안 되는 경우
13565정성태2/23/20241480닷넷: 2219. .NET CLR2 보안 모델에서의 개별 System.Security.Permissions 제어
13564정성태2/22/20241617Windows: 259. Hyper-V Generation 1 유형의 VM을 Generation 2 유형으로 바꾸는 방법
13563정성태2/21/20241705디버깅 기술: 196. windbg - async/await 비동기인 경우 메모리 덤프 분석의 어려움
13562정성태2/21/20241700오류 유형: 896. ASP.NET - .NET Framework 기본 예제에서 System.Web에 대한 System.IO.FileNotFoundException 예외 발생
13561정성태2/20/20241931닷넷: 2218. C# - (예를 들어, Socket) 비동기 I/O에 대한 await 호출 시 CancellationToken을 이용한 취소파일 다운로드1
13560정성태2/19/20241934디버깅 기술: 195. windbg 분석 사례 - Semaphore 잠금으로 인한 Hang 현상 (닷넷)
13559정성태2/19/20242809오류 유형: 895. ASP.NET - System.Security.SecurityException: 'Requested registry access is not allowed.'
13558정성태2/18/20242010닷넷: 2217. C# - 최댓값이 1인 SemaphoreSlim 보다 Mutex 또는 lock(obj)를 선택하는 것이 나은 이유
13557정성태2/18/20241765Windows: 258. Task Scheduler의 Author 속성 값을 변경하는 방법
13556정성태2/17/20241825Windows: 257. Windows - Symbolic (hard/soft) Link 및 Junction 차이점
13555정성태2/15/20241965닷넷: 2216. C# - SemaphoreSlim 사용 시 주의점
1  [2]  3  4  5  6  7  8  9  10  11  12  13  14  15  ...