Sender: nate AT cartsys DOT com
Message-ID: <369BAB42.7FE4F14A@cartsys.com>
Date: Tue, 12 Jan 1999 12:06:26 -0800
From: Nate Eldredge <nate AT cartsys DOT com>
X-Mailer: Mozilla 4.08 [en] (X11; I; Linux 2.0.36 i586)
MIME-Version: 1.0
To: djgpp AT delorie DOT com
CC: George Lugovoy <glugovoy AT uic DOT rnd DOT runnet DOT ru>
Subject: Re: gprof needs help
References: <Pine DOT SUN DOT 3 DOT 91 DOT 990111154803 DOT 25036G-100000 AT is>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp AT delorie DOT com

Eli Zaretskii wrote:
> 
> On Mon, 11 Jan 1999, George Lugovoy wrote:
> 
> > Each sample counts as 0.055556 seconds.
> >   %   cumulative   self              self     total
> >  time   seconds   seconds    calls  Ts/call  Ts/call  name
> > 100.00      0.28     0.28
> > __get_current_directory
> >   0.00      0.28     0.00    27113     0.00     0.00  ForceExt
> >   0.00      0.28     0.00    15148     0.00     0.00  FL_Open
> >   0.00      0.28     0.00    14039     0.00     0.00  Long2Str
> 
> Was this program linked with DJGPP v2.02?
> 
> If so, then the profiling code there has a bug that manifests itself in
> the above way: the first function is shown to take 100% of the time, and
> the rest take 0%.  This info is bogus, don't make any conclusions based
> on it.
> 
> I can send you a patched source of a library module which is responsible
> for this.  You will need to compile it, put it into libc.a, and then
> rebuild your program and run it again.  *Then* the profile should make
> sense.

Sheesh!  We're not doing very well with the profiling support, are we? 
First the _mono_printf in v2.01, and now this...

Does this warrant an update of the DJDEV zips?
-- 

Nate Eldredge
nate AT cartsys DOT com