This code isn't working.
mov [eax], FFhEAX contains the address of the destination of the MOV operation. While "mov eax,SOME_DATA_SOURCE_or_CONSTANT_DATA_VALUE" would move 4 bytes from the data source (or constant value) into the EAX register, the square brackets around EAS like [EAS] is supposed to completely change the meaning of it. [EAX] is supposed to mean "use the the value stored in EAX register as a memory address". Therefore, "mov [eax],SOME_DATA_SOURCE_or_CONSTANT_DATA_VALUE" is supposed to mean "copy SOME_DATA_SOURCE_or_CONSTANT_DATA_VALUE to memory at the address stored in the EAX register". But it doesn't work. Every time I compile, I get this error.
Quoteerror A2070: invalid instruction operands
Why am I getting this error?
mov [eax],dl
mov [eax],dx
mov [eax],edx
in the above lines, the assembler knows the size of the operand (byte, word, dword)
mov [eax],0FFh
the assembler doesn't know whether you want to write a byte, word, or dword sized operand
mov byte ptr [eax],0FFh
mov word ptr [eax],0FFh
mov dword ptr [eax],0FFh
now, the assembler knows what you want
Quote from: Ben321 on March 23, 2015, 07:25:08 PM
... would move 4 bytes from the data source (or constant value) into the EAX register
mov dword ptr [eax], 0FFh
Quote from: dedndave on March 23, 2015, 09:45:33 PM
mov [eax],dl
mov [eax],dx
mov [eax],edx
in the above lines, the assembler knows the size of the operand (byte, word, dword)
mov [eax],0FFh
the assembler doesn't know whether you want to write a byte, word, or dword sized operand
mov byte ptr [eax],0FFh
mov word ptr [eax],0FFh
mov dword ptr [eax],0FFh
now, the assembler knows what you want
What does the "ptr" keyword do?
Hi Ben,
Quote from: Ben321 on March 24, 2015, 04:22:29 AM
What does the "ptr" keyword do?
not much. It's a syntax question for Masm/jWasm. There are other assemblers available and the
ptr stuff isn't necessary for those. But again: this syntax is necessary for Masm and jWasm.
Gunther
I assume ptr means "pointer" so as to say "use the value stored in eax register as a pointer to the desired memory destination", but I thought that was already covered by putting "eax" in brackets like "[eax]". Without brackets eax by itself means "use the eax register as the destination". So using "ptr" seems quite redundant. However without "ptr" that line of code fails. Why?
Quote from: Ben321 on March 24, 2015, 05:10:03 AM
However without "ptr" that line of code fails. Why?
The reason is the syntax (see post #4).
Gunther
[eax]
the value in EAX is an address
for 32-bit code, addresses are always 32-bits wide
the data operand, however, may be byte, word, dword, in some cases, oword
when working with the FPU, it might be real4 ptr, real8 ptr, or real10 ptr
"ptr" does mean pointer
if you want to translate, something like, "the address in EAX is a pointer to a ____"
the PTR operator may be used in a few different ways, though :P
Quote from: dedndave on March 24, 2015, 06:40:12 AM
when working with the FPU, it might be real4 ptr, real8 ptr, or real10 ptr
Don't forget word ptr on the FPU ;-)
include \masm32\MasmBasic\MasmBasic.inc ; download (http://masm32.com/board/index.php?topic=94.0)
Init
push 1234+65536*5678
fild word ptr [esp]
fimul word ptr [esp+2]
fistp dword ptr [esp]
pop eax
Inkey Str$("1234*5678=%i", eax)
Exit
end start
jj,
as long as we're at it let's not forget dword and qword ptr on the FPU :)
include \masm32\include\masm32rt.inc
.data?
dword1 dd ?
dword2 dd ?
.code
start:
; standard, sensible way to use dword ptr on the FPU ---------
push 1234
push 5678
fild dword ptr [esp]
fimul dword ptr [esp+4]
fistp dword ptr [esp+4]
pop eax
pop eax
printf("1234*5678=%i\n", eax)
; and normal, obvious way to use qword ptr on the FPU ---------
push 5678
push 1234
fild qword ptr [esp]
fstp qword ptr [esp]
pop dword1
pop dword2
printf("1234+65536^2*5678=%.19g\n", qword ptr dword1)
ret
end start
Quote from: rrr314159 on March 24, 2015, 01:33:34 PM
; and normal, obvious way to use qword ptr on the FPU ---------
*** WARNING - mathematician detected :greensml: