Xref: news-dnh.mv.net comp.os.msdos.djgpp:2685
Newsgroups: comp.os.msdos.djgpp
Path: news-dnh.mv.net!mv!news.sprintlink.net!newsfeed.internetmci.com!EU.net!sun4nl!phcoms4.seri.philips.nl!newssvr!kunst
From: kunst AT natlab DOT research DOT philips DOT com (Pieter Kunst)
Subject: Re: clock() unit ?
Sender: news AT natlab DOT research DOT philips DOT com (USENET News System)
Organization: Philips Research Laboratories, Eindhoven, The Netherlands
References: <19951012170909 DOT olly AT mantis DOT co DOT uk> <DGE0sx DOT Aq6 AT natlab DOT research DOT philips DOT com> <19951013163056 DOT olly AT mantis DOT co DOT uk>
Date: Mon, 16 Oct 1995 08:50:28 GMT
Lines: 32
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Dj-Gateway: from newsgroup comp.os.msdos.djgpp

In article <19951013163056 DOT olly AT mantis DOT co DOT uk> Olly Betts writes:
> >>I *think* that CLOCKS_PER_SEC is an old alternative name for CLK_TCK,so they
> >>should be the same if both are defined.  I may be wrong though.
> >
> >No, CLOCKS_PER_SEC is an ANSI C constant, from <time.h> (See K&R,2nd, p.255).
> >clock()/CLOCKS_PER_SEC returns the processor time used by the program,
> >in seconds, since the beginning of execution.
> >
> >CLK_TCK is PC only. You'd better avoid using it.
>
> We've a draft of the ANSI standard here (October 31, 1988) which says
> CLK_TCK is what you divide clock() by, though it's quite possible they
> changed it in the issued version of the standard (which I don't have access
> to currently).
>
> It's not as simple as "CLK_TCK is PC only" though.

max AT alcyone DOT darkside DOT com (Erik Max Francis) writes:
> > It's not as simple as "CLK_TCK is PC only" though.
> My copy of the ANSI C 1990 standard doesn't even mention CLK_TCK.


Perhaps we'd better take the upcoming ANSI C++ standard as a reference ?!?
You can browse through the 'April Working Paper' of the ANSI C++ standard at:

  http://www.cygnus.com/~mrs/wp-draft/index.html

Here you'll find references to CLOCKS_PER_SEC in section 18.7, Table 5
[lib.support.runtime] and in annex C.4, Table 1 [diff.library].
CLK_TCK is not mentioned in this paper.

Pieter.