dp in data segment

[gelatine1]

Hello, this is a very stupid question but I just can't figure it out. How should I use the dp in the data segment ? I tried like this but it gave an error when I tried to compile.

Code Select Expand

rect dd 0 dp(4)
I just want 4 duplicates of this 0 value.

Thanks in advance
Jannes

[Gunther]

Jannes,

Hello, this is a very stupid question but I just can't figure it out. How should I use the dp in the data segment ? I tried like this but it gave an error when I tried to compile.

Code Select Expand

rect dd 0 dp(4)
I just want 4 duplicates of this 0 value.

Thanks in advance
Jannes

your question isn't clear for me. Do you mean the dup operator? If so, it's a syntax question.

Gunther

[gelatine1]

Ohhh Yes I think I meant dup and yes indeed it's a syntax question. I'm not sure how to use it.

[jj2007]

Here is an example. Launch the exe with Olly and see what it does...

include \masm32\include\masm32rt.inc

.data?
rc   RECT 5 dup(<>)

.code
start:
   mov eax, offset rc
   mov ecx, offset rc.top
   mov edx, rc[0*RECT].right
   mov edx, rc[1*RECT].right
   mov edx, rc[2*RECT].right
   exit

end start

[gelatine1]

Oh yes I see :)  If I want to declare 4 dwords containing 0 then I should use

Code Select Expand

rect dd 4 dup(0) instead of

Code Select Expand

rect dd 0 dup(4).
Thanks for the help guys

[dedndave]

that will work
however, for rectangles, we generally use a RECT structure

http://msdn.microsoft.com/en-us/library/dd162897%28v=vs.85%29.aspx

most of the MS structures are defined for you in windows.inc

Code Select Expand

RECT STRUCT
  left    dd      ?
  top     dd      ?
  right   dd      ?
  bottom  dd      ?
RECT ENDS

now, you can do something similar to what Jochen showed you
however - his code creates 5 RECT structures   
you probably only need one

this will create an uninitialized RECT
we usually place it in the .DATA or .DATA? section

Code Select Expand

rc RECT <>

if you would like to, you may initialize the structure (.DATA section)
it has 4 members, so....

Code Select Expand

rc RECT <10,25,30,35>

you may access the members as

Code Select Expand

    mov     eax,rc.right

or, use a register for the address and

Code Select Expand

    mov     edx,offset rc
    mov     eax,[edx].RECT.right

[jj2007]

or, use a register for the address and

Code Select Expand

    mov     edx,offset rc
    mov     eax,[edx].RECT.right

This is indeed a good solution because its encoding is much smaller than for a global rc.right (for example). Typically, you would use a non-volatile register such as esi, edi, ebx.

[Gunther]

That's all not bad but back to the question of DUP. One can use, for example:

Code Select Expand


     db     128  dup (?)

That statement would reserve 128 bytes with arbitrary content.

Gunther

[MichaelW]

Gunther,

I'm not sure how this is so, because the listing shows that the reserved bytes are in the .data section, but in my tests under Windows the reserved bytes were zeroed.

[nidud]

deleted

[Gunther]

I'm not sure how this is so, because the listing shows that the reserved bytes are in the .data section, but in my tests under Windows the reserved bytes were zeroed.

Either a design decision or a feature of Windows. An error is unthinkable. But joke apart, it's interesting to know.

Gunther