Hello,

I've come to a point in my study of MASM where I can't work the examples for
disk I/O unless I have Windows 98 or less. I'm also wondering if I'll be
able to work the BIOS interrupt examples because I run WinXp as a
VirtualBox guest VM on a Linux host - I'm not sure how VirtualBox will
respond to BIOS interrupts.

At some point I want to attempt to write a defrag program in assembly, so
I'd really like to learn more about Disk I/O.

What are my best options?  FreeDos as a guest with VirtualBox, FreeDos and
dosemu on linux?  Download an illegal copy of win98 and use that as a VM
guest? (though I'm still not sure about win98/DOS BIOS interrupt in
VirtualBox).

I'm slowly moving away from windows - I've caught the linux bug!  So, I'm
wondering if I should just start learning NASM and linux interrupts - I
know "next to nothing" about linux programming now.  If this is the best
route to go, can I find NASM include files and libraries for linux and/or
windows (like masm32 provides for masm/windows api).

Regarding NASM, what book or online source can help bring me up to speed
quickly - to where I am presently with masm assembly... ie, skip all the
explanations of adc,mul,sbb,mov,movzx,lea, register/memory explanations,
etc?

On Fri, 22 Feb 2008 10:10:25 -0500
Pop Tart  <spamtrap@crayne.org> wrote:

> I've caught the linux bug!

Welcome to the club!

> Regarding NASM, what book or online source can help bring me up to
> speed quickly

The NASM manual would be a good start. It has a section entitled "Quick
Start for MASM Users", as well as chapters on writing for different
operating system environments. And, of course, it doesn't hurt that
several of the NASM developers participate in this news group.

-- 
Chuck 
http://www.pacificsites.com/~ccrayne/charles.html

Pop Tart  <spamtrap@crayne.org> wrote:
>
>I've come to a point in my study of MASM where I can't work the examples for
>disk I/O unless I have Windows 98 or less. ...
>
>At some point I want to attempt to write a defrag program in assembly, so
>I'd really like to learn more about Disk I/O.

Well, remember that defragging requires more than just disk I/O: it also
requires a detailed understanding of file systems.  Windows XP, Windows 98,
MS-DOS, and Linux each use VERY different file systems, which require VERY
different defragment strategies.

>What are my best options?  FreeDos as a guest with VirtualBox, FreeDos and
>dosemu on linux?  Download an illegal copy of win98 and use that as a VM
>guest? (though I'm still not sure about win98/DOS BIOS interrupt in
>VirtualBox).

If you really want to dink with disk I/O at the BIOS interrupt level, then
I would have to think you'd want a DOS guest in a VM.  Even with Win98, you
have to worry about sharing access to the drive with the rest of the
operating system.
-- 
Tim Roberts, timr@probo.com
Providenza & Boekelheide, Inc.

Pop Tart wrote:
> Hello,
> 
> I've come to a point in my study of MASM where I can't work the examples for
> disk I/O unless I have Windows 98 or less.

Is that a signed "less" or an unsigned "less"? :)

> I'm also wondering if I'll be
> able to work the BIOS interrupt examples because I run WinXp as a
> VirtualBox guest VM on a Linux host - I'm not sure how VirtualBox will
> respond to BIOS interrupts.

I don't know. I've run dosemu... not enough to really know what it does, 
but enough to know that it isn't "exactly like dos".

> At some point I want to attempt to write a defrag program in assembly, so
> I'd really like to learn more about Disk I/O.

Okay... There's more to "defrag" than disk I/O, and there's disk I/O 
outside of bios interrupts. Not running dos doesn't rule it out entirely.

> What are my best options?  FreeDos as a guest with VirtualBox, FreeDos and
> dosemu on linux?  Download an illegal copy of win98 and use that as a VM
> guest? (though I'm still not sure about win98/DOS BIOS interrupt in
> VirtualBox).

I'm not sure why you'd want Win98. I know it comes with "real dos", but 
in a "dos box", the bios interrupts are "fake" (I mean "emulated"). I 
think you'd want "real bios interrupts", if you're going to study them. 
I hope you've got a (working) floppy drive. That'll make it easy to boot 
to "real dos", or to your own boot sector, if you want to get even 
"lower level" than dos. It'll also give you a drive to "practice" on. 
Writing to a hard drive can get you into a world of pain if you screw 
up! You'll want to have a "rescue disk" on hand - a "live CD" would 
probably be okay, or a floppy, or a spare hard drive that's *not* in the 
machine while you're experimenting/learning.

> I'm slowly moving away from windows - I've caught the linux bug!

Windows is winning the "popularity contest" by a wide margin, but I find 
Linux quite "interesting" to work with. (when you get to the "defrag" 
level, though, there are a lot of file systems to deal with, which won't 
make it easy). You're a little farther from the hardware, but you *can* 
get to it, with some mumbo-jumbo.

> So, I'm
> wondering if I should just start learning NASM and linux interrupts - I
> know "next to nothing" about linux programming now.

We all start out equal in that regard. There's Linux Programming, and 
there's Assembly Linux Programming. The latter is an oxymoron - Linux is 
portable, asm is not. You're not "supposed" to do it in asm, and there 
are good reasons for this. But we *can*, and it's "educational", if 
nothing else. If you should come up with something "useful", it will be 
useful to a large number of people, even if not "all Linux users". A 
majority of a minority, if you will...

If you wanted to make this transition "stepwise", you could start using 
Nasm for some dos projects that you've already got working in Masmese, 
and make the transition to Linux when you're comfortable with Nasm. But 
the jump from dos in Masm to Linux in Nasm isn't *that* great. Linux is 
*not* "dos on steroids", but at a beginner level, it's a lot like dos 
(or can be). As you get into something more "advanced" (which I haven't, 
much) it gets more complicated - but this would also be true of Windows, 
or even dos...

>  If this is the best
> route to go, can I find NASM include files and libraries for linux and/or
> windows (like masm32 provides for masm/windows api).

Sure... or even if it isn't the "best" route... The MASM32 project is a 
bit more mature than anything for Nasm - kudos to Hutch and helpers for 
keeping that up! You may not find the Nasm stuff quite as "complete" 
(I'm not too familiar with either). But it'll include stuff like 
"invoke" and "proc" (maybe "if" and all), which Nasm doesn't have "built 
in". The "equates" are the most important part (?), and that should be 
pretty much the same. For Windows:

http://www.asmcommunity.net/projects/nasmx/

The Nasmx Project - formerly called NASM32 - is the "latest and 
greatest"... I ASSume they include all the "good stuff" from other Nasm 
include files - "nagoa+.inc", and all. If not... I'm sure they welcome 
contributions, and/or you can "mix and match" the "good parts" according 
to your taste. (but beware the "Tower of Babel"!).

There's a Linux section to Nasmx, too (and Xbox!), but the Linux section 
is not (yet) as extensive (the Xwindows .inc is better) as the include 
files available as the "asmutils" package. Part of 
http://linuxassembly.org a.k.a. http://asm.sf.net - the asmutils package 
is for more than Linux - BSD and other *nixish systems are included. 
This makes the package a bit of a PITA to use - "system.inc" includes 
other files, depending on what OS (etc.) is defined. So you need 
"-d__LINUX__ -d__ELF__ ..." defined - on the command line, or in an 
environment variable, or in the source (if you don't mind locking 
yourself out of the flexibility), or in a Makefile. Easily done, but if 
it isn't, perfectly good source will refuse to assemble... This has got 
a good selection of "equates" (and structure definitions) - any missing 
ones can be found in the appropriate C header file(s) (somewhat 
scattered around). There's a utility available to convert .h files to 
Nasm-syntax .inc files (with pretty good success). Great thanks to 
Johannes Kroll for that one! Works for Windows, too:

http://home.comcast.net/~fbkotler/h2incn-0.5.tar.gz

So include files can be improved/updated as needed...

> Regarding NASM, what book or online source can help bring me up to speed
> quickly - to where I am presently with masm assembly... ie, skip all the
> explanations of adc,mul,sbb,mov,movzx,lea, register/memory explanations,
> etc?

Well... the Nasm manual has conveniently removed the "adc" section for 
ya. :)  That, and a few examples, should get you through the Masm-Nasm 
transition.

Getting up to speed with Linux is a longer-term proposition. As I said, 
it's "easy as dos" at first... "The" book on Unix programming, I 
understand (haven't read it), is W. Richard Stevens' "Advanced 
Programming in the Unix Environment". This is C, not asm. Another C one 
- available online - "Advanced Linux Programming" by Mark Mitchell, 
Jeffrey Oldham, and Alex Samuel, of CodeSourcery LLC:

http://www.advancedlinuxprogramming.com/

Despite C, they're relevant, I think. Getting to assembly, Richard 
Blum's "Professional Assembly Language":

<http://www.wrox.com/WileyCDA/WroxTitle/productCd-0764579010,descCd-tableOfContents.html>

Or Bob Neveln's "Linux Assembly Language Programming":

<http://books.google.com/books?hl=en&id=pbB8Z1ewwEgC&dq=neveln&printsec=frontcover&source=web&ots=pqVw4xiug7&sig=QTxCp1UFnnH3wxebElKrfvxqVDE#PPR7,M1>

Exist... I haven't read any of 'em. I've looked at (and liked) Jonathan 
Bartlett's "Programming From the Ground Up" - asm for Linux, but Gas 
rather than Nasm (a greater transition than Masm<->Nasm, I think, but 
not impossible):

http://www.cafepress.com/bartlettpublish.8640017

Or the "free, as in freeloader" version:

http://savannah.nongnu.org/projects/pgubook/

The examples from that book have been translated to Nasm syntax, and are 
available here:

http://mysite.verizon.net/fbkotler/nasm-pgu-examples.tar.bz2

if you'd care to attempt to read the book about Gas and use Nasm...

Paul Carter's tut is for Windows or Linux or BSD (achieved by using a 
lot of C):

http://www.drpaulcarter.com/pcasm/

Might be too much on the "adc, registers and memory" side for you - I 
wouldn't write that stuff off completely yet - there may be subtleties 
you haven't mastered - but the Intel/AMD manuals are probably the best 
source for that...

Here's another Linux assembly tutorial I haven't looked at for a while - 
looks pretty good:

http://docs.cs.up.ac.za/programming/asm/derick_tut/

I should mention Jeff Owens' AsmIde:

http://members.save-net.com/jko%40save-net.com/asm/

I'm a "command line guy", and don't use the IDE, but the "AsmRef" 
section is *great* - rare Linux documentation aimed at asm!

That's probably enough to get you started... if you want to pursue this, 
we can probably come up with examples... something reading "infile" and 
writing to "outfile" to start with, and maybe get to some "lower level" 
disk I/O... there are a couple of things I've been meaning to try... 
Bios ints in Linux *are* possible - there's a "Linux Real Mode 
Interface" library - http://lrmi.sf.net - but definitely not a beginner 
project!

Best,
Frank

Frank Kotler  <spamtrap@crayne.org> writes:
> Pop Tart wrote:
> > Hello,
> > I've come to a point in my study of MASM where I can't work the
> > examples for
> > disk I/O unless I have Windows 98 or less.
> 
> Is that a signed "less" or an unsigned "less"? :)

Signed or unsigned? Signed or unsigned? We don't even know
if it's an integral type or floating point type yet!

However, I don't see any point in "windows 98", so it 
must be an integer and we're back to your question again.

Phil 
-- 
Dear aunt, let's set so double the killer delete select all.
-- Microsoft voice recognition live demonstration

On Sun, 24 Feb 2008 22:28:45 GMT, Frank Kotler  <spamtrap@crayne.org>
wrote:

>Pop Tart wrote:
>> Hello,
>> 
>> I've come to a point in my study of MASM where I can't work the examples for
>> disk I/O unless I have Windows 98 or less.
>
<snip>

>I'm not sure why you'd want Win98. I know it comes with "real dos", but 
>in a "dos box", the bios interrupts are "fake" (I mean "emulated"). I 
>think you'd want "real bios interrupts",

I assume the OP said "Win98 or less" because you can get to
real-mode DOS.  It's a little more cumbersome than just using the 
"DOS box" you get when you click on the MS-DOS option.  You
can get to real-mode via "Start - Shutdown - Restart the computer 
in MS-DOS mode", or you can create a desktop icon that automatically
goes to real-mode, runs your app, and then restarts Windows.  You
can't do either of those in later Windows versions (except WinMe,
with some trickery).

Best regards,


 
Bob Masta
 
              DAQARTA  v3.50
   Data AcQuisition And Real-Time Analysis
             www.daqarta.com
Scope, Spectrum, Spectrogram, FREE Signal Generator
        Science with your sound card!

On Fri, 22 Feb 2008 10:10:25 -0500, Pop Tart  <spamtrap@crayne.org>
wrote:

>I've come to a point in my study of MASM where I can't work the examples for
>disk I/O unless I have Windows 98 or less.

WinXP changed the segment defaults, which certainly affected some
32 bit protected mode programs. For us, anything that has failed
in the XP DOS box has always worked with DOSbox from
  http://www.dosbox.com/

Stephen


-- 
Stephen Pelc, stephenXXX@mpeforth.com
MicroProcessor Engineering Ltd - More Real, Less Time
133 Hill Lane, Southampton SO15 5AF, England
tel: +44 (0)23 8063 1441, fax: +44 (0)23 8033 9691
web: http://www.mpeforth.com - free VFX Forth downloads

Bob Masta wrote:
> On Sun, 24 Feb 2008 22:28:45 GMT, Frank Kotler  <spamtrap@crayne.org>
> wrote:
> 
> 
>>Pop Tart wrote:
>>
>>>Hello,
>>>
>>>I've come to a point in my study of MASM where I can't work the examples for
>>>disk I/O unless I have Windows 98 or less.
>>
> <snip>
> 
>>I'm not sure why you'd want Win98. I know it comes with "real dos", but 
>>in a "dos box", the bios interrupts are "fake" (I mean "emulated"). I 
>>think you'd want "real bios interrupts",
> 
> 
> I assume the OP said "Win98 or less" because you can get to
> real-mode DOS.  It's a little more cumbersome than just using the 
> "DOS box" you get when you click on the MS-DOS option.  You
> can get to real-mode via "Start - Shutdown - Restart the computer 
> in MS-DOS mode", or you can create a desktop icon that automatically
> goes to real-mode, runs your app, and then restarts Windows.  You
> can't do either of those in later Windows versions (except WinMe,
> with some trickery).

Yeah... I'm not so sure about "restart in dos mode". Type "exit", and 
IIRC you wind up back in Windows. Not what I call "real dos", but it 
probably is "real" enough for bios interrupts (not "real" enough for 
Flat Real Mode, IIRC).

You can hit some function key... F5? F8? during bootup, or what i did - 
edit msdos.sys. It's just a text file - didn't used to be, in the good 
old days. If you change "GUI=1" to "GUI=0" - "real dos". If you want the 
GUI, type "win". Before you can edit msdos.sys, you've gotta remove some 
"attributes" from it - system, hidden and readonly. "attrib -s -h -r 
msdos.sys" - put 'em back when you're done when you're done, if you feel 
like it. This used to be "everybody knows"... I suppose it's in danger 
of becoming "forgotten lore"...

If we're stuck with "greater than" Win98, where the dos box "doesn't 
smell right", and if we've got a floppy drive, we might be interested in 
this:

http://mikegonta.com/dosboots

(opendos, or similar, ought to be fine, too)

Maybe could make a bootable CD out of it, too(?).

Best,
Frank

On Mon, 25 Feb 2008 14:19:21 GMT, Frank Kotler  <spamtrap@crayne.org>
wrote:

>Bob Masta wrote:
>> On Sun, 24 Feb 2008 22:28:45 GMT, Frank Kotler  <spamtrap@crayne.org>
>> wrote:
>> 
>> 
>>>Pop Tart wrote:
>>>
>>>>Hello,
>>>>
>>>>I've come to a point in my study of MASM where I can't work the examples for
>>>>disk I/O unless I have Windows 98 or less.
>>>
>> <snip>
>> 
>>>I'm not sure why you'd want Win98. I know it comes with "real dos", but 
>>>in a "dos box", the bios interrupts are "fake" (I mean "emulated"). I 
>>>think you'd want "real bios interrupts",
>> 
>> 
>> I assume the OP said "Win98 or less" because you can get to
>> real-mode DOS.  It's a little more cumbersome than just using the 
>> "DOS box" you get when you click on the MS-DOS option.  You
>> can get to real-mode via "Start - Shutdown - Restart the computer 
>> in MS-DOS mode", or you can create a desktop icon that automatically
>> goes to real-mode, runs your app, and then restarts Windows.  You
>> can't do either of those in later Windows versions (except WinMe,
>> with some trickery).
>
>Yeah... I'm not so sure about "restart in dos mode". Type "exit", and 
>IIRC you wind up back in Windows. Not what I call "real dos", but it 
>probably is "real" enough for bios interrupts (not "real" enough for 
>Flat Real Mode, IIRC).

The fact that "exit" gets you back to Windows means nothing...
it does a complete reboot to get back there.  My old Daqarta
for DOS is very picky about real mode, and it runs just fine.
It tests the MSW for the PE flag, and won't even start if set.
I haven't tried Flat Real Mode, but I'd be surprised if it didn't
work.  I have never found anything, nor read anything, to 
indicate that this is anything other than normal Real Mode.

Best regards,





Bob Masta
 
              DAQARTA  v3.50
   Data AcQuisition And Real-Time Analysis
             www.daqarta.com
Scope, Spectrum, Spectrogram, FREE Signal Generator
        Science with your sound card!

On Feb 26, 7:05�am, NoS...@daqarta.com (Bob Masta) wrote:
> On Mon, 25 Feb 2008 14:19:21 GMT, Frank Kotler �<spamt...@crayne.org>
> wrote:
>
>
>
>
>
> >Bob Masta wrote:
> >> On Sun, 24 Feb 2008 22:28:45 GMT, Frank Kotler �<spamt...@crayne.org>
> >> wrote:
>
> >>>Pop Tart wrote:
>
> >>>>Hello,
>
> >>>>I've come to a point in my study of MASM where I can't work the examples for
> >>>>disk I/O unless I have Windows 98 or less.
>
> >> <snip>
>
> >>>I'm not sure why you'd want Win98. I know it comes with "real dos", but
> >>>in a "dos box", the bios interrupts are "fake" (I mean "emulated"). I
> >>>think you'd want "real bios interrupts",
>
> >> I assume the OP said "Win98 or less" because you can get to
> >> real-mode DOS. �It's a little more cumbersome than just using the
> >> "DOS box" you get when you click on the MS-DOS option. �You
> >> can get to real-mode via "Start - Shutdown - Restart the computer
> >> in MS-DOS mode", or you can create a desktop icon that automatically
> >> goes to real-mode, runs your app, and then restarts Windows. �You
> >> can't do either of those in later Windows versions (except WinMe,
> >> with some trickery).
>
How is this done: 'you can create a desktop icon that automatically
goes to real-mode, runs your app, and then restarts Windows' for real
mode enviornment on WIN98?  I would like to know how to do this
properly.

For checking, I used debug:
-f 100 L 2ff 90
-a
  mov  ax,1687
  int  2f
-p =100 3

-in debug to test for support of DPMI v0.9
(RET AH.0 if successful, AH.8Ah if unsuccessful, AH.1687h if no
support.)
For the [Start][Programs] MS-DOS Prompt, it _is_ supported indicating
VM mode and creating a shortcut for sending to the desktop is the
same.

For via "Start - Shutdown - Restart the computer in MS-DOS mode", it
is _not_ supported, rets: AH,1687h.

However, in debug, if
  pushf
  pop  bx
RETS:
BX.7202h  indicating IOPL == 3. _for__both_
Real 'Real Mode' returns IOPL == 0, if done from floppy boot sector
code.

Steve

> >Yeah... I'm not so sure about "restart in dos mode". Type "exit", and
> >IIRC you wind up back in Windows. Not what I call "real dos", but it
> >probably is "real" enough for bios interrupts (not "real" enough for
> >Flat Real Mode, IIRC).
>
> The fact that "exit" gets you back to Windows means nothing...
> it does a complete reboot to get back there. �My old Daqarta
> for DOS is very picky about real mode, and it runs just fine.
> It tests the MSW for the PE flag, and won't even start if set.
> I haven't tried Flat Real Mode, but I'd be surprised if it didn't
> work. �I have never found anything, nor read anything, to
> indicate that this is anything other than normal Real Mode.
>
> Best regards,
>
> Bob Masta
>
> � � � � � � � DAQARTA �v3.50
> � �Data AcQuisition And Real-Time Analysis
> � � � � � � �www.daqarta.com
> Scope, Spectrum, Spectrogram, FREE Signal Generator
> � � � � Science with your sound card!- Hide quoted text -
>
> - Show quoted text -

Bob Masta wrote:

....
>>Yeah... I'm not so sure about "restart in dos mode". Type "exit", and 
>>IIRC you wind up back in Windows. Not what I call "real dos", but it 
>>probably is "real" enough for bios interrupts (not "real" enough for 
>>Flat Real Mode, IIRC).
> 
> 
> The fact that "exit" gets you back to Windows means nothing...

It isn't what I'd consider "normal behavior" for "real dos".

> it does a complete reboot to get back there.

Mmmm? My machine was set up with "GUI=0", so a "complete reboot" would 
have brought me back to "real dos".

> My old Daqarta
> for DOS is very picky about real mode, and it runs just fine.
> It tests the MSW for the PE flag, and won't even start if set.

Well, that proves something...

> I haven't tried Flat Real Mode, but I'd be surprised if it didn't
> work.

I'm quite sure it didn't - for me. It may be that I didn't have emm386 
out of the picture, or something. "loadlin" worked, IIRC...

> I have never found anything, nor read anything, to 
> indicate that this is anything other than normal Real Mode.

Perhaps my memory is faulty. I didn't spend too much time in "that dos", 
as I recall, because it *didn't* seem "normal" to me. Hmmm, I think I 
know where I might be able to find a drive with a working Win98 on it. 
(I've been saying I "can't" run Windows... actually, I just "won't"... 
but maybe I "will"...) If I do, I'll try to investigate what I didn't 
like about "restart in dos mode" compared to "GUI=0". Your experience 
probably trumps my memory, but "that's not the way I remember it".

Best,
Frank

This puppy, run in dosemu, is blowin' my mind. Claims real mode. I guess 
dosemu must be providing me with a fake cr0(?). Jeez, I *hope* I'm not 
changing the real cr0!!!

org 100h

mov dx, realmsg

mov eax, cr0   ; either of these...
;smsw ax

and eax, 0FFFFFFFEh

mov cr0, eax  ; unexpectedly, no complaint
; lmsw ax     ; GPF kills dosemu

; smsw ax     ; reports v86 mode
mov eax, cr0  ; claims real real mode!!!

test al, 1
jz real
mov dx, v86msg

real:
mov ah, 9
int 21h
ret

realmsg db "real real mode$"
v86msg db "v86 mode$"

On Tue, 26 Feb 2008 09:30:40 -0800 (PST), s_dubrovich
<spamtrap@crayne.org> wrote:

>On Feb 26, 7:05�am, NoS...@daqarta.com (Bob Masta) wrote:

>> >> I assume the OP said "Win98 or less" because you can get to
>> >> real-mode DOS. �It's a little more cumbersome than just using the
>> >> "DOS box" you get when you click on the MS-DOS option. �You
>> >> can get to real-mode via "Start - Shutdown - Restart the computer
>> >> in MS-DOS mode", or you can create a desktop icon that automatically
>> >> goes to real-mode, runs your app, and then restarts Windows. �You
>> >> can't do either of those in later Windows versions (except WinMe,
>> >> with some trickery).
>>
>How is this done: 'you can create a desktop icon that automatically
>goes to real-mode, runs your app, and then restarts Windows' for real
>mode enviornment on WIN98?  I would like to know how to do this
>properly.
>

Go through the normal shortcut creation process: Right-click
where you want the icon, select New, then Shortcut, and enter
the complete path for the command line.  Proceed to select
a shortcut name and icon, then hit Finish to get an icon on
the desktop

Then right-click that icon, go to Properties - Advanced, and
select the "MS-DOS mode" checkbox.  That's it!

More detailed instructions for Daqarta for DOS are at:

    http://www.daqarta.com/readme.htm#win95

Best regards,





Bob Masta
 
              DAQARTA  v3.50
   Data AcQuisition And Real-Time Analysis
             www.daqarta.com
Scope, Spectrum, Spectrogram, FREE Signal Generator
        Science with your sound card!