From: g DOT smith AT chem DOT canterbury DOT ac DOT nz (Greg Smith)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: debugging?
Date: Thu, 24 Aug 2000 08:40:25 GMT
Organization: University of Canterbury
Lines: 56
Message-ID: <39a4dd3d.106648430@news.canterbury.ac.nz>
References: <39a34d59 DOT 4345588 AT news DOT canterbury DOT ac DOT nz> <8nvvk5$hst$1 AT nets3 DOT rz DOT RWTH-Aachen DOT DE>
NNTP-Posting-Host: 132.181.173.156
X-Newsreader: Forte Free Agent 1.21/32.243
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

On 23 Aug 2000 07:48:21 GMT, Hans-Bernhard Broeker
<broeker AT physik DOT rwth-aachen DOT de> wrote:

>This is by far not precise enough to give a helpful response.  We'll
>need compilable code (minimized to the bare necessities before
>posting, please) to be able to test this out.
>
I thought so... I wasn't sure if I should expect a helpful response :)
Glad to see I'm too cynical about such things...

>The debugger can be confused, sometimes, about the values of certain
>variables in highly optimized code which are currently kept in
>registers during execution.

This is good to know but the code isn't optimised at all.

typedef struct{
	double rho,hz,m,m0,kp,kpt,ftr,kbb,kt11,alpha,lambda,td,
		t0,trp,time,kttrp;
}model_data;

typedef struct{
	double h;
}simulation_data;

main(){

	int pulse = 0;
	model_data mdata;
	simulation_data sdata;
	double h;

        mdata.time = 0;
        mdata.t0 = 0.1;
        sdata.h = 0.01;
        mdata.td = 1;

   do{

		if(mdata.time >= mdata.t0*pulse && mdata.t0*pulse <
mdata.time+sdata.h){
			pulse++;
		}

	h=sdata.h;
	if(sdata.h >= mdata.t0*pulse - mdata.time)h = mdata.t0*pulse -
mdata.time;

   	mdata.time += h;

   }while(mdata.time < mdata.td);
	return 0;
}

On my system the program hangs just before pulse = 6 but the debugger
says the condition to execute 'pulse++' is satisfied. Any ideas?