# time ./a.out test.data

the output:

real 0m49.172s
user 0m5.638s
sys 0m4.850s

Why real time and user time have big difference? I've tried to change
a code: for example disable print() function, the real time and user
time closer. How to improve the real time in a program?

In article <f7db0716-0e06-4caa-95fc-284130408b44
@l32g2000hse.googlegroups.com>, cplusplusquestion@gmail.com says...
> # time ./a.out test.data
>
> the output:
>
> real 0m49.172s
> user 0m5.638s
> sys 0m4.850s
>
> Why real time and user time have big difference? I've tried to change
> a code: for example disable print() function, the real time and user
> time closer. How to improve the real time in a program?

This basically means your program is I/O bound. Without knowing more
about how that's occurring it's impossible to suggest improvements.
Sometimes, it's just the nature of that particular program (e.g. cat
will almost always be heavily I/O bound).

--
Later,
Jerry.

The universe is a figment of its own imagination.

On Jan 14, 6:18 pm, cplusplusquest...@gmail.com wrote:
> # time ./a.out test.data
>
> the output:
>
> real 0m49.172s
> user 0m5.638s
> sys 0m4.850s
>
> Why real time and user time have big difference? I've tried to change
> a code: for example disable print() function, the real time and user
> time closer. How to improve the real time in a program?

http://www.google.com/search?hl=en&q...=Google+Search

The first result would probably be a good place to start.

I wouldn't do anything without a profiler though.

On Jan 15, 1:44 pm, jonnyot...@gmail.com wrote:
> On Jan 14, 6:18 pm, cplusplusquest...@gmail.com wrote:
>
> > # time ./a.out test.data
>
> > the output:
>
> > real 0m49.172s
> > user 0m5.638s
> > sys 0m4.850s
>
> > Why real time and user time have big difference? I've tried to change
> > a code: for example disable print() function, the real time and user
> > time closer. How to improve the real time in a program?
>
> http://www.google.com/search?hl=en&q...2B%2B+code&btn...
>
> The first result would probably be a good place to start.
>
> I wouldn't do anything without a profiler though.

Thanks. It's really ful.