Actually, it's even more complicated. Modern versions of Windows sometimes use rtdtc internally to the KeQueryPerformanceCounter function (with workarounds to account for the changing CPU frequency)
The results that you are looking at are 7 years old. I don't believe they would still apply now (but I might be wrong). GV -------------------------------------------------- From: "Guy Harris" <[email protected]> Sent: Tuesday, September 28, 2010 9:30 AM To: <[email protected]> Subject: Re: [Winpcap-users] About the packets loss ,what is the bottleneck ? > > On Sep 28, 2010, at 4:22 AM, Helmut Vaupotitsch wrote: > >> 1) According to Section 5.3 and Figures 6,9 in >> http://www.winpcap.org/docs/WinPcap-SBAC03.pdf (small packets), >> timestamp optimization is switched off in the standard distribution, >> which costs *a lot* of cpu processing. > > According to section 5.3, "This high performance counter is incremented at > every processor clock cycle, so its precision is equivalent to the CPU > frequency"; as far as I know, on a lot of machines there is no guarantee > that the CPU frequency is a constant, at least with some operating > systems, so it might be Really Hard to turn the TSC value into a > reasonable time stamp on those machines. > > _______________________________________________ > Winpcap-users mailing list > [email protected] > https://www.winpcap.org/mailman/listinfo/winpcap-users _______________________________________________ Winpcap-users mailing list [email protected] https://www.winpcap.org/mailman/listinfo/winpcap-users
