Fasm 1 macro fill data.

Index > Macroinstructions > Fasm 1 macro fill data.

Author

Thread

Roman

Joined: 21 Apr 2012
Posts: 1847

Roman 20 Dec 2019, 10:39

I want macro Fill.

Code:

Fill buf1, 1,7,3,44
Fill buf2, 4,5

Fill gen asm code:

Code:

Mov eax, buf1
Mov [eax], 1
Mov [eax+4], 7
Mov [eax+8], 3
Mov [eax+12], 44

20 Dec 2019, 10:39

revolution
When all else fails, read the source

Joined: 24 Aug 2004
Posts: 20448
Location: In your JS exploiting you and your system

revolution 20 Dec 2019, 11:16

Perhaps something like this:

Code:

macro Fill address,[value] {
 common
  local i
  i = 0
  Mov eax, address
 forward
  Mov dword[eax + i], value
  i = i + 4
}

Untested, so see if it works for you.

20 Dec 2019, 11:16

Roman

Joined: 21 Apr 2012
Posts: 1847

Roman 20 Dec 2019, 11:26

Super puper Smile

Thanks.

20 Dec 2019, 11:26

ProMiNick

Joined: 24 Mar 2012
Posts: 804
Location: Russian Federation, Sochi

ProMiNick 20 Dec 2019, 13:58

above variant not able to fill values from memory (only imm or reg) because of absence mov mem to mem instruction.
but push memory exists

macro FillByPush address, size, [value] {
common
mov eax, address
add eax,size-4
xchg eax, esp
reverse
push value
common
xchg esp, eax ; if restoring eax significant| or just mov esp, eax
add eax, 4 ; if restoring eax significant
}

but it require to precede its values by size
in large structures my variant is significantly shorter & faster, and structures size is something predefined.

20 Dec 2019, 13:58

revolution
When all else fails, read the source

Joined: 24 Aug 2004
Posts: 20448
Location: In your JS exploiting you and your system

revolution 20 Dec 2019, 14:13

ProMiNick wrote:

above variant not able to fill values from memory (only imm or reg) because of absence mov mem to mem instruction.
but push memory exists

macro FillByPush address, size, [value] {
common
mov eax, address
add eax,size-4
xchg eax, esp
reverse
push value
common
xchg esp, eax ; if restoring eax significant| or just mov esp, eax
add eax, 4 ; if restoring eax significant
}

but it require to precede its values by size
in large structures my variant is significantly shorter & faster, and structures size is something predefined.

You can also use:

Code:

mov eax, address + size - 4

But also note that this code is only suitable for an OS running in protected mode with multiple stacks implemented, or if you know interrupts are already disabled. Otherwise an intervening interrupt will get a bad stack.

20 Dec 2019, 14:13

DimonSoft

Joined: 03 Mar 2010
Posts: 1228
Location: Belarus

DimonSoft 20 Dec 2019, 18:47

Changing ESP to point to arbitrary places gets program into trouble almost definitely: Why do we even need to define a red zone? Can’t I just use my stack for anything?

20 Dec 2019, 18:47

ProMiNick

Joined: 24 Mar 2012
Posts: 804
Location: Russian Federation, Sochi

ProMiNick 20 Dec 2019, 22:18

Using stack for struct initialization (taken by me from reversing OOPs hardcoded internals) is always simpler, shorter and faster, but yes it is applicable only for protected mode with multiple stacks implemented.
But I test it even more limited - OOP use pushes to initialize object located in heap, not in stack. I test this technic only in heap too!

20 Dec 2019, 22:18

< Last Thread | Next Thread >

Forum Rules:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum