X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f
Message-Id: <3.0.1.16.20040208120922.428f5b8a@earthlink.net>
X-Sender: ethros AT earthlink DOT net
X-Mailer: Windows Eudora Light Version 3.0.1 (16)
Date: Sun, 08 Feb 2004 12:09:22 -0500
To: Eli Zaretskii <eliz AT elta DOT co DOT il>, CBFalconer <cbfalconer AT yahoo DOT com>
From: Ethan Rosenberg <ethros AT earthlink DOT net>
Subject: Re: Cross Platform Incompatabilites?
Cc: djgpp AT delorie DOT com
In-Reply-To: <9743-Sat07Feb2004181714+0200-eliz@elta.co.il>
References: <3 DOT 0 DOT 1 DOT 16 DOT 20040204105918 DOT 33f75924 AT earthlink DOT net>
 <3 DOT 0 DOT 1 DOT 16 DOT 20040204105918 DOT 33f75924 AT earthlink DOT net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

At 06:17 PM 2/7/04 +0200, you wrote:
>> Date: Wed, 04 Feb 2004 10:59:18 -0500
>> From: Ethan Rosenberg <ethros AT earthlink DOT net>
>> 
>> I have DJGPP installed on my computer at home, which is running DOS 6.1.  I
>> have written C code which I run on two other platforms:
>> 
>> 1]Windows 98
>>     a) DOS shell from Windows
>>     b) Command prompt from boot 
>> 
>> 2] Windows 2000
>>      DOS shell from Windows
>> 
>> The results from 1(a) and 2 agree.  The results from 1(b) and DOS 6.1 do
>> not agree with each other or with the results from  1(a) and 2.   Both
>> 1(b) and DOS 6.1 have the DPMI routines that come with DJGPP.
>
>Details, please!  How exactly do the results ``not agree''?  It's hard
>to even try thinking of which of the myriad of possible differences
>between the two system configurations could cause this, without
>knowing what is the difference in the results of running the program.
>
>Also, please tell in detail how do you get to the ``command prompt
>from boot'' configuration.  It could be that we mean two different
>things when we talk about that.


The program is calculating a Fast Fourier Transform (FFT).  I have checked
it on test data, and know it works.  I have also compile with the -Wall
flag, to see if there was something subtle there.  No warnings of
incompatiblities were found.

The differences are *possibly* generated in a routine which calculates the
integral under the curve of FFT*FFT*2, using the trapezoidal rule.  The
error appears in only one integration of a calculation of probably a
hundred integrations.  The integration routine does not seem to have errors.  

This is the first time I have encountered any inconsistencies, in hundreds
of calculations.  The files I am using are VERY big.  The FFT vector
contains 16k complex variables, composed of two float type variables.  I
*might* be having addressing problems.

Command Prompt form Boot:  Hold the control key down when Win98 boots, and
it gives a number of options, one of which is "command prompt". 

I received the following answer from:  CBFalconer <cbfalconer AT yahoo DOT com>

>Salient differences:  DOS 6.1 and 1(b) do not handle long file
>names nor multitasking.  DOS 6.1 cannot access FAT32 partitions.
>W2000 does peculiar things associated with hardware access, from
>the DOS viewpoint.

I do not use long filenames, or perform multi-tasking.  This is a pure DOS
program that is running in a shell off windows, of from the command prompt.
 I do not think I am accessing FAT32.

My index of suspicion goes to my config.sys file which is:

DEVICE=C:\DOS\HIMEM.SYS
DEVICE=C:\DOS\EMM386.EXE NOEMS HIGHSCAN I=B000-B7FF
BUFFERS=40,0
FILES=50
DOS=UMB
LASTDRIVE=Z
FCBS=40,0
INSTALL=C:\DOS\SHARE.EXE /l:500 /f:5100
DEVICEHIGH /L:2,41104 =C:\BPCDROM\BPCDDRV.SYS /D:BPCDDRV$  /D:BPCDDRV$
DEVICEHIGH /L:2,12048 =C:\DOS\SETVER.EXE
DEVICEHIGH /L:2,9072 =C:\DOS\ANSI.SYS
DOS=HIGH
SHELL=C:\DOS\COMMAND.COM /P /E:4096
STACKS=9,256
BREAK=ON


Does anyone see anything wrong??

Much thanks in advance.

Ethan Rosenberg