getting elapsed time in milliseconds

[Thomas Baier]

Hi there,

I need to know how to get the elapsed time in milliseconds. I tried
time_t start,end;
time(start);
do_something();
time(end);

double dif = difftime(end, start);

but this only returns the elapsed time in seconds.
So how can I get it in milliseconds?


Thanks for hlep

Thomas

[WW]

Re: getting elapsed time in milliseconds

Thomas Baier wrote:[color=blue]
> I need to know how to get the elapsed time in milliseconds. I tried[/color]
[SNIP][color=blue]
> So how can I get it in milliseconds?[/color]

By a high resolution timer facility provided by your operating system, if it
provides one. Not possible to do only using standard C++ language features.
Please post your question to a newsgroup dedicated to yiur
compiler/platform.

--
WW aka Attila

[Thomas Baier]

Re: getting elapsed time in milliseconds

Thomas Baier wrote:

I'm working under Suse Linux 8.2 with g++. Maybe anybody in this newgroup
could answer my question now. Or could anybody tell me a good newsgroup to
post?


Thomas

[Alf P. Steinbach]

Re: getting elapsed time in milliseconds

On Sat, 20 Sep 2003 19:26:50 +0200, Thomas Baier <thomas@tho-bai.de> wrote:
[color=blue]
>Thomas Baier wrote:
>
>I'm working under Suse Linux 8.2 with g++. Maybe anybody in this newgroup
>could answer my question now. Or could anybody tell me a good newsgroup to
>post?[/color]

This group is not "please find a newsgroup for me".

Please don't post any more such questions (read the FAQ etc. about
topicality, in other words, RTFM).

That said, try [comp.os.linux.d evelopment.*].

[J. Campbell]

Re: getting elapsed time in milliseconds

Thomas Baier <thomas@tho-bai.de> wrote in message news:<3f6c881d$ 0$23106$9b4e6d9 3@newsread2.arc or-online.net>...[color=blue]
> Hi there,
>
> I need to know how to get the elapsed time in milliseconds. I tried
> time_t start,end;
> time(start);
> do_something();
> time(end);
>
> double dif = difftime(end, start);
>
> but this only returns the elapsed time in seconds.
> So how can I get it in milliseconds?
>
>
> Thanks for hlep
>
> Thomas[/color]

The following will do what you want:


int main(){
int clo = clock();
//do stuff
cout << (clock() - clo) << endl;
return 0;
}

[Kevin Goodsell]

Re: getting elapsed time in milliseconds

J. Campbell wrote:[color=blue]
> Thomas Baier <thomas@tho-bai.de> wrote in message news:<3f6c881d$ 0$23106$9b4e6d9 3@newsread2.arc or-online.net>...
>[color=green]
>>So how can I get it in milliseconds?
>>[/color]
>
> The following will do what you want:
>
>
> int main(){
> int clo = clock();
> //do stuff
> cout << (clock() - clo) << endl;
> return 0;
> }[/color]

clock() is not a function for retrieving elapsed (real) time, and it's
result is not measured in milliseconds. clock() gives *processor time*,
and the precision it uses depends on the implementation.

-Kevin
--
My email address is valid, but changes periodically.
To contact me please use the address from a recent posting.

[J. Campbell]

Re: getting elapsed time in milliseconds

Kevin Goodsell <usenet1.spamfr ee.fusion@never box.com> wrote in message news:<cdmbb.111 3$vS.633@newsre ad3.news.pas.ea rthlink.net>...[color=blue]
> J. Campbell wrote:[color=green]
> > Thomas Baier <thomas@tho-bai.de> wrote in message news:<3f6c881d$ 0$23106$9b4e6d9 3@newsread2.arc or-online.net>...
> >[color=darkred]
> >>So how can I get it in milliseconds?
> >>[/color]
> >
> > The following will do what you want:
> >
> >
> > int main(){
> > int clo = clock();
> > //do stuff
> > cout << (clock() - clo) << endl;
> > return 0;
> > }[/color]
>
> clock() is not a function for retrieving elapsed (real) time, and it's
> result is not measured in milliseconds. clock() gives *processor time*,
> and the precision it uses depends on the implementation.
>
> -Kevin[/color]

Thanks for the heads-up. On my implimentation, clock() returns
milliseconds with circa 15 ms precision, and I'd assumed (always
imprudent) that it worked the same across the board. I dug a little
and learned of the CLOCKS_PER_SEC constant, which will help a bit for
gross measurements. Thanks for keeping me on the
straight-and-narrow...