[TRLog] Program crashing on 6x86 remains!

[ok1rr@gw.ok0ncg.ampr.org]

From: ok1rr@ok0ncg.#boh.cze.eu 
To: trlog@contesting.com

Hi there,
some month ago I referred about program crashing on machine 
using 6x86 processor by IBM/Cyrix. Unfortunately, some OMs 
advised me to look for RFI into computer. I spent some tens 
of hours trying to fix this problem which still remains in 
evidence.

I am using VT5099A chipset Pentium system board with the
6x86L-PR200+ CPU, Pentium BIOS REV:3.0 (Plug & Play) and
Award Modular BIOS v4.51PG. System has the S3 video card
using S3 86C325 Video BIOS version 1.00-10 by S3 Inc., 1996.
Finally, my system uses Windows 95 version 4.00.950B (OSR2)
without any other MS-DOS (no previous version of MS-DOS),
64 MB of RAM and two IDE HDD (1GB/16bit FAT and 4 GB/32bit 
FAT, both LBA) controlled by on-board IDE controller. TR log
is located on non bootable drive D, 1GB/16 bit FAT.

I tried to check or change some BIOS options and inspected
some wait states etc. I also tried to disable on-board ports
but the problem remains - program crashes after some QSOs
(the number varies from 6 to 60). Of course I carefully 
watched the memory indicator which stayed any time over 100000.
I inspected both CONFIG.SYS and AUTOEXEC.BAT and tried both 
also as empty files with zero length. Finally, I prepared
a very small TRMASTER.DTA of about 41k length with some 500
stations inside - the problem remains. First tests made with
my previous configuration under MS-DOS 6.22 but the problem
was still here.

Tests were made with TR Log versions 6.04 to 6.18, all results
were the same - program crashed. There was no error code,
my computer simply freezed (Ctrl-Alt-Del did nothing) or
gone and restarted itself. This occurs if some data are
typed in - I am trying to type in call K1AR, K1A appears
and computer became stoned.

Another tests were made using TR DEBUG mode - no problem!
Program was able to "work" over 700 QSO! What difference
is between normal and simulator/debug modes - disabled
keying port? (I am using LPT1 port because of possibility
to connect a paddle).

Finally, I tried to run the same version/configuration of TR
on 8 other machines (486 or Pentiums 90 to 166) without any 
problem. TR ran fine even I used a BIG log file from Pantelleria 
(IH9) expedition with about 15000 QSOs and SAY HI ENABLE was TRUE 
with an enormous TRMASTER.DTA file containing detailed data 
of over 28000 stations.

It should be noticed that I made these tests WITHOUT any
keyed radio, so NO RF WAS PRESENT at all. I want to tell
the readers that I did really a big effort within over 50
hours before I decided to write this message. So do not
write me, please, that I should watch free memory, remove 
all TSRs or my computer is exposed to excessive RF...
I hope that bad configuration or improper setup of the 
machine may be excluded.

It seems that TR Log has some kind of alergy to the Cyrix
processor (Win95 recognizes it as CyrixInstead). Maybe some
part of compiler output code does something strange to the 
IBM/Cyrix processor? Or is it a problem of the S3 video card?

I will be hardly interested if someone has the same CPU
and S3 video card and runs TR without any problem.

73 Martin, OK1RR

public@brailcom.cz
 or
Martin.K@pemail.net
 or
ok1rr@ok0ncg.epe.cz


                                                 73 Martin, OK1RR

--
FAQ on WWW:               http://www.contesting.com/trlogfaq.html
Submissions:              trlog@contesting.com
Administrative requests:  trlog-REQUEST@contesting.com
Problems:                 owner-trlog@contesting.com
Feature Wishlist:	  http://web.jzap.com/n6tr/trwish.html