I am using the first method, listed here.
I want to take the time measurement of a construction of a tree, which contains about 2841482 nodes (inner and leaves).
Here it is:
which gives me always a negative value, like -13481628 ns.Code:struct timespec start, end; clock_gettime(CLOCK_MONOTONIC, &start); Tree t(...); clock_gettime(CLOCK_MONOTONIC, &end); int64_t time; time = timespecDiff(&end, &start); std::cout<<"Time: " << time << " ns\n";
Code - functions and small libraries I use
It’s 2014 and I still use printf() for debugging.
"Programs must be written for people to read, and only incidentally for machines to execute. " —Harold Abelson
The first thing would be to check the return result of clock_gettime()
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
I do not know exactly what native type is used for timeA_p->tv_sec, but timeA_p->tv_nsec is long and I do not see a cast to int_64 inside
So I suppose there is mathematical overflow... - you need to add cast to int64_t inside this expressionCode:int64_t timespecDiff(struct timespec *timeA_p, struct timespec *timeB_p) { return ((timeA_p->tv_sec * 1000000000) + timeA_p->tv_nsec) - ((timeB_p->tv_sec * 1000000000) + timeB_p->tv_nsec); }
All problems in computer science can be solved by another level of indirection,
except for the problem of too many layers of indirection.
– David J. Wheeler
It might overflow.Code:timeA_p->tv_sec * 1000000000
Next time include all functions inside the post, so that others don't have to see irrelevant content like that.
Are struct timespec members declared as type int64_t? If not, then the code in timespecDiff() could be the issue.
update - previous posts already mentioned this.
Already tried that Salem, but I got this
Vart, where exactly do you suggest me adding int64_t to?Code:main.cpp:54: error: no match for 'operator>>' in '(+std::operator<< [with _Traits = std::char_traits<char>](((std::basic_ostream<char, std::char_traits<char> >&)(+(&std::cout)->std::basic_ostream<_CharT, _Traits>::operator<< [with _CharT = char, _Traits = std::char_traits<char>](end.timespec::tv_sec))), ((const char*)" ")))->std::basic_ostream<_CharT, _Traits>::operator<< [with _CharT = char, _Traits = std::char_traits<char>](start.timespec::tv_sec) >> " \n"'
Code - functions and small libraries I use
It’s 2014 and I still use printf() for debugging.
"Programs must be written for people to read, and only incidentally for machines to execute. " —Harold Abelson
Shouldn't your code be:
Code:int64_t timespecDiff(struct timespec *timeA_p, struct timespec *timeB_p) { return (((int64_t)timeA_p->tv_sec * 1000000000) + timeA_p->tv_nsec) - (((int64_t)timeB_p->tv_sec * 1000000000) + timeB_p->tv_nsec); }
The cost of software maintenance increases with the square of the programmer's creativity. - Robert D. Bliss
Not sure, but what is sizeof( time_t ) on your system? If it's say 4 you'd get an integer overflow in timespecDiff.Originally Posted by std10093
Code:#include <cmath> #include <complex> bool euler_flip(bool value) { return std::pow ( std::complex<float>(std::exp(1.0)), std::complex<float>(0, 1) * std::complex<float>(std::atan(1.0) *(1 << (value + 2))) ).real() < 0; }
That was the case! Thank you!!! Poland guy you are right about the link.
Code - functions and small libraries I use
It’s 2014 and I still use printf() for debugging.
"Programs must be written for people to read, and only incidentally for machines to execute. " —Harold Abelson
It should be noted that there are instances on multi-core CPUs where time calculations can return negative values. This usually happens with the QueryPerformanceCounter API calls. The simple answer is to clamp the result at 0 even though this does not result in an accurate timer.
