Info-IBMPC@B.ISI.EDU (Info-IBMPC Digest) (10/11/86)
Info-IBMPC Digest Friday, October 10, 1986 Volume 5 : Issue 92
This Week's Editor: Phyllis O'Neil
Today's Topics:
Re: Hardware Debugger Query
Re: Microsoft C V4.0 Upgrade
INT 2e
IBM PC Terminal Emulator - Procomm 2.4 available
C Primers
SETENV.C
Using DEBUG for Low-level Format
Re: Floppy Head Step Rate Zap
Today's Queries:
Z-158
1.2 Meg Floppy on Leading Edge
Networking Question
Graphics Emulation Software
NEC PC
----------------------------------------------------------------------
To: lenoil@xx.lcs.mit.edu
Cc: info-ibmpc@usc-isib
Subject: Re: Hardware Debugger Query
Date: Mon, 06 Oct 86 12:57:41 -0500
From: Mark Colan <mtc@ATHENA.MIT.EDU>
The only "hardware assistance" in the Prof Debug Facility is a small
card with nothing but a push-button switch on it. It generates a
non-maskable interrupt.
While I am not familiar with the specific products such as ATRON or
PERISCOPE, I believe that hardware assistance generally means an
adaptor card for the debugger which monitors the CPU state directly
(one of them plugs into the math coproc slot for intimate contact
with the CPU). These cards do not come cheap, which is why I am
not directly familiar with them.
The advantages of this approach include:
1. The debugger can reside on its own card, in its own memory, in a
place where a runaway program cannot accidentally overwrite it.
Thus, your system may be trashed, but the debugger still runs to
do post-mortum debugging.
2. By monitoring the hardware directly, you can determine when a given
location is being stored to. For example, if the symptom of your bug
is that part of your program, or DOS itself, is overwritten, you can
have the debugger trap when that location is stored upon. Thus you
can find the error in your program, which is usually a bad pointer.
Such problems are VERY difficult to find without such a card.
Of course, if a protected-mode DOS ever comes out, it will be impossible
to overwrite DOS with a runaway pointer; such attempts would presumably
be trapped by the hardware. It would still be a useful tool for finding
where the program overwrites itself.
The PC Tech Journal frequently runs ads on these types of products.
Mark Colan
MIT Project Athena
------------------------------
To: LENOIL@xx.lcs.mit.edu
Cc: info-ibmpc@b.isi.edu
Subject: Re: Microsoft C V4.0 Upgrade
Date: Mon, 06 Oct 86 14:22:47 PDT
From: Brent Chapman <chapman%cory.Berkeley.EDU@BERKELEY.EDU>
Robert Lenoil (LENOIL@XX.LCS.MIT.EDU) writes:
> 4.0 has changed so much that they
> send you an entire new set of manuals, rather than a supplement.
I think this gives people the wrong idea; 4.0 isn't really different
from 3.0 in any incompatible way. There are some new functions, and
some new command line options, but all the old stuff is still there.
All of my source code compiled perfectly without any changes to the code
or the makefiles. There are now three manuals instead of two, but much
of the extra space is devoted to things that weren't a part of the 3.0
release, such as MS-MAKE and CodeView.
The three manuals in 4.0 are:
Run-Time Library Reference Manual -- contains all the man pages
for the function calls
User's Guide -- info on
compiling and compiler options, as well as instructions for
using LIB, LINK, and MAKE
Microsoft CodeView and C Language Reference Manual -- CodeView is
the new source-level debugger. The C Language Reference is
Microsoft's "official" C definition, which is apparently
common to its MS-DOS and its XENIX products.
FYI, the Ver 3.0 manuals were:
Library Reference Manual -- basicly the same as the Ver 4.0
Run-Time Library Reference Manual
User's Guide and Reference Manual -- basicly the same as the Ver 4.0
User's Guide, minus the MAKE info (Ver 3.0 didn't include MAKE),
plus the C Language Reference Manual.
I'm quite pleased with Ver 4. Compile time appears to be about the same
as for Ver 3, but optimization appears to be much better. Ver 4
executables are both smaller and faster than their Ver 3 equivalents (at
least for my applications).
CodeView is really a nice, useful debugger. Unfortunately, I can't
use it on my applications, because there isn't enough memory (on a
640K AT). Compiling and linking my programs to include CodeView
debugging information about doubles their size. (The average,
non-CodeView-ready binary for the programs I work on is about 120K.
Compiling and linking for CodeView increases this to over 200K.)
CodeView is absolutely fantastic for more reasonably sized stuff,
though.
Brent Chapman
chapman@cory.berkeley.edu or ucbvax!cory!chapman
------------------------------
Date: Mon, 6 Oct 86 14:06:51 PDT
From: Jim Anderson <bilbo.jta@LOCUS.UCLA.EDU>
To: info-ibmpc@b.isi.edu
Subject: INT 2e
Daniel Briggs, rather matter of factly, talks about INT 2e as though
everyone out there knows about it. It seems very interesting.
Investigation shows you can point DS:SI to a command string and call
COMMAND.COM by executing INT 2e. The command string consist of a
count byte (apparently ignored) and a CR terminated command. SS:SP
are clobbered across the call so it is necessary to save them, much
like an EXEC call.
Passing a SET command will set variables in COMMAND's copy of the
environment. Of course this does not effect the environment of the
current process.
As with all undocumented functions, Caveat Programmer.
------------------------------
Date: Mon, 6 Oct 1986 19:21 MDT
From: Keith Petersen <W8SDZ@SIMTEL20.ARPA>
To: INFO-HZ100@RADC-TOPS20.ARPA, INFO-IBMPC@B.ISI.EDU
Subject: IBM PC Terminal Emulator - Procomm 2.4 available
>Procomm version 2.4 is just out. I recommend it highly. It has
>significant improvements over 2.3, which was very good itself. One
>point of correction -- they INSIST on a $25 donation (not that you
>couldn't just download it and not pay; but they don't just "ask").
>Procomm has sliding window kermit implemented. This does a lot to
>overcome the problems with the small packetsize -- but does any other
>Kermit implementation support that yet?
Procomm 2.4 is now available via FTP from SIMTEL20:
Filename Type Bytes CRC
Directory PD:<MSDOS.MODEM-UTIL>
PROCOM24.ARC.1 BINARY 134528 4F96H
PRODOC24.ARC.1 BINARY 121856 13C2H
--Keith Petersen
Arpa: W8SDZ@SIMTEL20.ARPA
Uucp: {bellcore,decwrl,harvard,lll-crg,ucbvax,uw-beaver}!simtel20.arpa!w8sdz
GEnie Mail: W8SDZ
RCP/M Royal Oak: 313-759-6569 (300, 1200, 2400 bps)
------------------------------
Date: Mon 6 Oct 86 20:53:30-PDT
From: David Kristofferson <KRISTOFFERSON%BIONET@SUMEX-AIM.Stanford.EDU>
Subject: C Primers
To: info-ibmpc@USC-ISIB.ARPA
cc: KRISTOFFERSON%BIONET@SUMEX-AIM.Stanford.EDU
I recently used a book by Bryan Costales called "C from A to Z"
published by Prentice-Hall (1985). The book has plenty of examples and
appears to cover a wider range of subjects than most of the other
books that I've spent hours staring at in bookstores. There are no
exercises, however, which I think is a bit of a drawback. I also have
Purdham's "C Programming Guide" which I don't like as much. I found
"The C Programmer's Handbook" by Thom Hogan to be a useful reference
mainly because my cheap C compiler manual (EcoSoft) left much to be
desired. Finally, the "C Toolbox" by William Hunt and "Dr. Dobb's
Toolbook of C" are good books to move on to, not to mention, of
course, K&R. When you get right down to it, I've enjoyed the parts of
K&R that I've read! Most of these books should be available in your
standard B. Dalton or Waldenbooks, etc., bookstores.
Dave Kristofferson
BIONET
------------------------------
Date: 7 October 86 16:28-PST
From: DEP%SLACVM.BITNET@WISCVM.WISC.EDU
To: INFO-IBMPC@USC-ISIB.ARPA
Subject: SETENV.C
Interestingly enough, I noticed in my Microsoft C compiler version 4.0
package ... it may have been there in earlier versions ... there is
included a program called SETENV, which actually modifies COMMAND.COM
to provide for a new environment size at boot time.
------------------------------
Date: Tue 7 Oct 86 18:08:22-EDT
From: Paul G. Weiss <PGW@XX.LCS.MIT.EDU>
Subject: Using DEBUG for Low-level Format
To: info-ibmpc@B.ISI.EDU
I recently watched the hardware technician install a Seagate drive in
one of my computers. He did the low level format by doing:
A>debug
-g=c800:5
which is the address of the format routine in the controller's BIOS.
He then did regular DOS FORMAT, which will find bad tracks if any
exist.
------------------------------
Date: Wed, 8 Oct 86 05:15:11 -0300
From: Ariel Cohen <aco%taurus.BITNET@WISCVM.WISC.EDU>
To: INFO-IBMPC@USC-ISIB.ARPA
Subject: Re: Floppy Head Step Rate Zap
I tried Mr. Winkel's tip (#88) and saw no improvement. After playing
with other parameters in the table (which is pointed to by the vector
of int 1E) and changing the parameters of DOS 3.1 jo those of DOS 2.1
I got an improvement of about 35%. The first nibble of the byte at
offset 0 was changed to D Hex (this byte is now DF Hex). The byte at
offset 9 (which is supposed to be 19 Hex but strange enough is 01 in
the copy I patched) was changed to 0F (this is the parameter for the
head settle time), and the byte at offset 10 (decimal) was changed
from 04 to 02 (motor start-up time). I believe the best thing to do
is playing with these parameters, using DEBUG, and perform a timing
test (copying an entire diskette, for instance) to determine which are
the best parameters for your system.
------------------------------
From: germar@nprdc.arpa (Marcelo Germar)
Date: 6 October 1986 1308-PDT (Monday)
To: info-ibmpc@b.isi.edu
Subject: Company Address
Does anybody know the address of Peripherals Enterprise Co., a maker
of multi-function card for the pc?
Your response will be very much appreciated.
--marcelo
------------------------------
Date: 06 OCT 1986 14:17 EDT
To: <INFO-IBMPC@USC-ISIB.ARPA>
From: Kenneth R. Van Wyk <LUKEN%LEHICDC1.BITNET@WISCVM.WISC.EDU>
Subject: Z-158 and 80286 Co-processor
Does anyone out there have any experience in using any of the 80286
"co- processor" boards which are available in a Zenith Z-158? If so
(good and/or bad), I would really appreciate hearing from you. I have
a Z-158 in my office and I've been considering upgrading it with a
80286 board, but I don't want to waste any time and effort if it can't
be done on a Zenith due to any hardware/software incompatibilities.
Thanks.
Kenneth R. Van Wyk
Lehigh University
<LUKEN@LEHICDC1.BITNET>
------------------------------
Date: Tue, 7 Oct 86 13:06 EDT
From: slade.wbst@Xerox.COM
Subject: 1.2 Mb Floppy on Leading Edge Model D
To: info-ibmpc@b.isi.edu
There may have been something in the digest on this before; if so
could someone tell me when? Otherwise, has anyone tried this?
My Dad just got a Leading Edge Model D with 360k, 30Meg hard disk and
floppy controller on the motherboard (from what he tells me, the hard
disk controller is separate in a slot). He would like to add a 1.2Mb
1/2 floppy. the sparse instructions indicate that the Model D can
control up to four floppies but dont say how.
Has anyone added a 1.2MB floppy to this clone and what do you need to
do? He has seen that Priority One has been advertizing a Weltec 16T
1.2 Mb floppy-has anyone used this drive and if he can figure out how,
is it suitable for this application.
I'll summarize the responses, if I get any, for the digest.
Michael Slade
arpanet: slade.wbst@xerox.com
------------------------------
Date: 1986 Oct 7 15:56 EDT
From: Bob Babcock <PEPRBV%CFATA1.BITNET@WISCVM.WISC.EDU>
To: <Info-IBMPC@B.ISI.EDU>
Subject: Networking
Are there any communications programs available which would allow two
PC's to share files? The programs would need to run as servers in the
background so that either machine could ask the other for a file
without requiring the sender to interrupt the current application.
...
If not, any recomendations on how to set up a two machine network which
is not likely to expand? Cost is a major consideration.
Thanks for any responses.
------------------------------
Date: Wed, 8 Oct 86 08:15 IST
From: Chezy Gal <A45%TAUNIVM.BITNET@WISCVM.WISC.EDU>
Subject: Looking for a Good Color Graphics Emulation Program
To: <INFO-IBMPC@B.ISI.EDU>
Does anyone know of a good color graphics emulation program that can
work with SAS/GRAPH on an IBM 4381 running VM/SP (our PCs are
connected to the mainframe through an IBM 7171 controller). The
program must have a VT100 or TeleVideo 925 emulation (for the 7171). A
Tektronix 4105 emulation would be just right. Of course, I would
prefer a PD program, but any pointers to commercial products would
also be gratefully appreciated.
Chezy Gal, PC Advisor
Computation Center
Tel Aviv University
BITNET: A45@TAUNIVM
Acknowledge-To: Chezy Gal <A45@TAUNIVM>
------------------------------
Date: Wed, 8 Oct 86 18:21 AST
From: <IUS%DACTH51.BITNET@WISCVM.WISC.EDU> (Eberhard W. Lisse)
Subject: NEC Personal Computer
To: info-ibmpc@b.isi.edu
Hi,
I have read about the NEC personal computer. How compatible is it ?
Thanks,
el
------------------------------
End of Info-IBMPC Digest
************************
-------