
                   comp.os.os2.programmer.tools     (Usenet)

                  Sunday, 26-Sep-1999 to Friday, 01-Oct-1999

+----------------------------------------------------------------------------+

From: hei@hatespam.norman.no                            01-Oct-99 12:29:23
  To: All                                               01-Oct-99 12:19:04
Subj: ICAT problems

From: hei@hatespam.norman.no (Harald Eilertsen)

Greetings

I have some problems debugging an asm function in a driver with ICAT. 
Somehow ICAT seems to think the function is located in a 16bit code 
segment. Probably due to this, the asm instructions it shows are 
completely wrong. However, singlestepping through the code and 
watching the registers view, it is apparent that the right 
instructions are executed. (That is, the instructions executed are not
the same as is shown in the window.)

The asm function is built with ALP with the -Sv:M510 switch and 
codeview debug info. (I've tried without debug info as well.) It's 
called from a 32bit segment compiled with Watcom 10.6. Everything at 
ring 0. I use plain 386 instructions only, no fancy stuff.

Anybody got a clue what's going on?

Take Care!
--
Harald Eilertsen
Norman Data Defence Systems
http://www.norman.no/

--- WtrGate+ v0.93.p7 sn 165
 * Origin: Usenet: Norman Data Defence Systems (1:109/42)

+----------------------------------------------------------------------------+

From: uno@40th.com                                      01-Oct-99 13:00:17
  To: All                                               01-Oct-99 12:19:04
Subj: Re: ICAT problems

From: uno@40th.com (uno@40th.com)

Harald Eilertsen? (hei@hatespam.norman.no?) wrote (1 Oct 1999 12:29:47 GMT):
>I have some problems debugging an asm function in a driver with ICAT. 
>Somehow ICAT seems to think the function is located in a 16bit code

That's the usual case for OS/2 drivers, at least the PDD kind.

>called from a 32bit segment compiled with Watcom 10.6. Everything at 
>ring 0. I use plain 386 instructions only, no fancy stuff.

32-bit segments in an OS/2 PDD sound pretty fancy to me, but I suppose
you do know what you're doing, right?  Also, you said M510 switch --
MASM 510 didn't do 32-bit segments.  Does ALP do them anyway?

 '`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'
 Corne1 Huth  -  http://40th.com/
 Bullet database engines/servers 3.1  Win32-WinCE-OS2-Linux+

--- WtrGate+ v0.93.p7 sn 165
 * Origin: Usenet: Yanaguana (1:109/42)

+----------------------------------------------------------------------------+

From: veit@simi.gmd.de                                  01-Oct-99 14:12:16
  To: All                                               01-Oct-99 14:23:08
Subj: Re: ICAT problems

From: veit@simi.gmd.de (Holger Veit)

On 1 Oct 1999 13:00:35 GMT, uno@40th.com <uno@40th.com> wrote:
>
>Harald Eilertsen? (hei@hatespam.norman.no?) wrote (1 Oct 1999 12:29:47 GMT):
>>I have some problems debugging an asm function in a driver with ICAT. 
>>Somehow ICAT seems to think the function is located in a 16bit code
>
>That's the usual case for OS/2 drivers, at least the PDD kind.
>
>>called from a 32bit segment compiled with Watcom 10.6. Everything at 
>>ring 0. I use plain 386 instructions only, no fancy stuff.
>
>32-bit segments in an OS/2 PDD sound pretty fancy to me, but I suppose

They aren't fancy. They work, besides a lot of thunking and a small 16 bit 
stub with the device header and the strategy/interrupt entry points (that
immediately jmp into the flat segment).

>you do know what you're doing, right?  Also, you said M510 switch --
>MASM 510 didn't do 32-bit segments.  Does ALP do them anyway?

TEXT32 SEGMENT PUBLIC DWORD USE32 'CODE'

Holger

-- 
Signature fault - code dumbed

--- WtrGate+ v0.93.p7 sn 165
 * Origin: Usenet: GMD-AiS (1:109/42)

+----------------------------------------------------------------------------+

From: hei@hatespam.norman.no                            01-Oct-99 14:21:15
  To: All                                               01-Oct-99 14:23:08
Subj: Re: ICAT problems

From: hei@hatespam.norman.no (Harald Eilertsen)

On Fri, 1 Oct 1999 13:00:35, uno@40th.com (uno@40th.com) wrote:

> Harald Eilertsen? (hei@hatespam.norman.no?) wrote (1 Oct 1999 12:29:47 GMT):
> >called from a 32bit segment compiled with Watcom 10.6. Everything at 
> >ring 0. I use plain 386 instructions only, no fancy stuff.
> 
> 32-bit segments in an OS/2 PDD sound pretty fancy to me, but I suppose
> you do know what you're doing, right?  

I hope so... :) the rest of it works, and there's other asm functions 
that both work and show up right in the debugger. Everything is 
compiled with the same assembler and switches. It seems to be ICAT 
that's having problems, not OS/2, ALP or the C compiler.

> Also, you said M510 switch --
> MASM 510 didn't do 32-bit segments.  Does ALP do them anyway?

Obviously. The Extra DevHlps I've added to what's already in the 
mwdd32.sys support driver works well. The disassembling with watcoms 
disassembler or ida pro also gives the expected results. That is, 
segments thunks and code looks the way they should.

Thanx for the reply, though :)

Take Care!
--
Harald Eilertsen
Norman Data Defence Systems
http://www.norman.no/

--- WtrGate+ v0.93.p7 sn 165
 * Origin: Usenet: Norman Data Defence Systems (1:109/42)

+----------------------------------------------------------------------------+

From: uno@40th.com                                      01-Oct-99 16:51:19
  To: All                                               01-Oct-99 17:06:16
Subj: Re: ICAT problems

From: uno@40th.com (uno@40th.com)

Harald Eilertsen? (hei@hatespam.norman.no?) wrote (1 Oct 1999 14:21:31 GMT):
>Obviously.

To you, obviously (never used ALP since 6.11d worked just fine).  Long
story short: better get any support requests in to the ICAT team because
come the 22nd of never ICAT runs out of money and it too swirls and
swirls and then finally swooooosh! -- it's gone.


 '`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'`'
 Corne1 Huth  -  http://40th.com/
 Bullet database engines/servers 3.1  Win32-WinCE-OS2-Linux+

--- WtrGate+ v0.93.p7 sn 165
 * Origin: Usenet: Yanaguana (1:109/42)

+----------------------------------------------------------------------------+

From: coltail@multimania.com                            01-Oct-99 18:51:07
  To: All                                               01-Oct-99 17:06:17
Subj: PERSO!!!!!!!!!!

From: "coltail" <coltail@multimania.com>

Hallo! Hello! Salut!

Viens voir mon site perso.avec plein d'images sympas.......
Come to see my personal site with a lot of funny pictures.....
Hier gibt es ein personal site mit lustiger Bilder.....

http://www.multimania.com/coltail


--- WtrGate+ v0.93.p7 sn 165
 * Origin: Origin Line 1 Goes Here (1:109/42)

+----------------------------------------------------------------------------+

+============================================================================+
