Sender: nate AT cartsys DOT com
Message-ID: <375B21AD.CA46FE83@cartsys.com>
Date: Sun, 06 Jun 1999 18:34:37 -0700
From: Nate Eldredge <nate AT cartsys DOT com>
X-Mailer: Mozilla 4.08 [en] (X11; I; Linux 2.2.5 i586)
MIME-Version: 1.0
To: djgpp AT delorie DOT com
Subject: Re: profiling question
References: <199906061955 DOT PAA03897 AT delorie DOT com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp AT delorie DOT com

Batchex wrote:
> 
> I'm trying to to profile my program with gprof, but it always said
> that my functions ran for 0.0 ms. The sample counts is 0.05556 sec. I
> then tried rawtime() function, which  if I'm not mistaken, can track
> up to 54.945 ms. It also said that my function ran for 0 ms. So, I
> added delay(200) function at the end of my function. Now I got some
> numbers (165 ms) with rawtime(). But, the problem is gprof still
> report 0.0 ms for my function. First, I've suspected the problem was
> that because I'm calling external assembly functions, but as I tried
> with different source code in which no linking to external assembly,
> this is not the case. Furthermore, gprof doesn't report the time
> taken for calling any C library functions (there are no entries for
> that), only reports the time taken for calling my functions, which is
> all 0.0 ms. Does anybody know why this is happening ? I programmed
> under pure MS-DOS and have added -pg option when compiling.

There's a bug in DJGPP v2.02 with respect to profiling.  I've sent you
the fix under separate cover.
-- 

Nate Eldredge
nate AT cartsys DOT com