site_archiver@lists.apple.com Delivered-To: Darwin-dev@lists.apple.com conversion = info.denom / 1e9 * info.numer; and Steve's said On May 1, 2008, at 3:55 PM, Steve Checkoway wrote: Kristopher Matthews wrote:

conversion = 1e9 * info.numer / info.denom;

#include <mach/mach_time.h> #include <stdint.h> #include <unistd.h> #include <stdio.h> #include <stdlib.h> static double mach_elapsed_time( uint64_t start, uint64_t end ) { static double conversion = 0.; if( conversion == 0. ) { mach_timebase_info_data_t info; mach_timebase_info( &info ); conversion = info.numer / (1e9 * info.denom); } return (end - start)*conversion; } int main( int argc, const char *argv[] ) { uint64_t start = mach_absolute_time(); uint64_t end; int sleepTime = argc == 2? atoi(argv[1]):3; sleep( sleepTime ); end = mach_absolute_time(); printf( "%f seconds\n", mach_elapsed_time(start, end) ); return 0; } -- Steve Checkoway _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/kmmx%40mac.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... This email sent to site_archiver@lists.apple.com Okay, this is the last time I'm saying anything on this. But Steve is correct, my posted code is wrong. My code said: conversion = info.numer / (1e9 * info.denom); Mine only works by chance because info.numer == info.denom == 1 (for my particular system anyway, YMMV). conversion = info.denom / 1e9 * info.numer; Now you're multiplying the numerator and the denominator and dividing by 1000000000. Before, you were multiplying the numerator by 1e9 and dividing by the denominator.

conversion = 1e-9 * (double) info.numer / (double)

info.denom; } Before that, you were dividing the numerator by (the denominator * 1e9). Think about what units you want here. Your mach_elapsed_time() is returning a value in seconds. mach_absolute_time() gives you a time in units of mt (mach time, a unit I just made up). You can consider info.numer as having units of ns--nanoseconds--and info.denom as having units of mt. Since you want seconds in the end, your conversion had better have units of s/mt. So let's consider the three cases above: 1. info.denom / 1e9 * info.numer has units of mt * ns / 1e9 = mt * s. 2. 1e9 * info.numer / info.denom has units of 1e9 * ns / mt = as / mt (that is attoseconds per mach time or 10^-18 seconds per mach time) 3. 1e-9 * info.numer / info.denom has units of 1e-9 * ns / mt = s/mt. Your original (i.e., number 3) was correct. I'm not sure exactly what Terry was talking about when he was saying you were making mistakes. The following code is a complete working example. The numbers I get are pretty close to being the number of seconds asleep. I just checked the link you gave and it almost exactly the same as this except that info.numer is being multiplied by 1e-9 before being divided by info.denom. This email sent to kmmx@mac.com smime.p7s