Questions about assembly for people that can program in it?

Hi,

Next time, please don't write anything in one paragraph, it's hardly readable. Also, know that assembly language is a family of languages, you need to tell us which on you are talking about; it's platform-specific (e.g., x86 assembly is different from ARM assembly). The ABI is sometimes also relevant---judging by your question, I take it you're talking about x86 assembly under DOS.

1. It's 0x100 (i.e., hexadecimal) and you don't; that's specific to the .COM executable format. The reason behind this is that the .COM format has its traces in an older operating system called CP/M. CP/M loaded the code at offset 0x100h due to the fact that the first 256 (or 0x100) bytes were used for the zero page, a structure containing all sorts of information that the OS passed to the program. Due to backwards compatibility, MS-DOS also provides a similar structure called the PSP (Program Segment Prefix), which can be read and used by programs running in the DOS environment.

2. Registers aren't really variables: you have a limited number of them and they are stored inside the CPU itself, not inside separate memory chips. DX is the name Intel gave to one of their 16-bit GPRs (General-Purpose Registers). Here's a list of all the x86 16-bit GPRs and what they stand for:

AX = accumulator

BX = base

CX = count

DX = data

SI = source index

DI = destination index

BP = base pointer

SP = stack pointer

So what does the X in the first 4 registers stand for? It doesn't stand for anything; it means both the high-order byte and the low-order byte are used. Their high- and low-order bytes can be accessed independently, as 8-bit registers (e.g., AH, AL, BH, BL, etc.). The others can only be accessed as 16-bit registers. Note that 32-bit x86 CPUs add 32-bit registers (e.g., EAX, EBX, where the "E" stands for "extended"), and x86-64 CPUs add 64-bit registers (e.g., RAX, RBX, where the "R" stands for "REX," which you shouldn't worry about for the time being).

3. I don't have enough room to do it here (there's a limit on how long an answer on Yahoo! Answers can be), and you have not mentioned what assembler you are using (they all have slightly different syntaxes, as assembly isn't a standardized language)---I suppose you're using MASM. You're in luck, one of the best x86 assembly language books is written for MASM and has also been made freely available for download by its author.

Book: http://homepage.mac.com/randyhyde/webster.cs.ucr.e...

Lab manual: http://homepage.mac.com/randyhyde/webster.cs.ucr.e...

Companion library: http://webster.cs.ucr.edu/AsmTools/MASM/stdlib/std... (you will need this to test some of the examples in the book)

This should answer everything. For now, know that assembly is nothing like C++, or C#. It's an entirely different programming paradigm and you need to use a different different mindset when programming in it; there are no variables, no types, no objects, no classes, no templates, no memory allocation or garbage collectors, etc. You only have the building blocks upon which those things are built. Other languages have yet other paradigms; if you try programming in Lisp, Prolog, or Smalltalk, you will see that they are nothing like what you are probably used to.

INT 21h is how you make a system call in DOS; this is how you access the DOS API for things like file manipulation or printing strings. It can do different things based on the contents of your registers. System calls in DOS use a software interrupt mechanism (hence the "INT") and the handler used is 21h. You will get more familiar with this when the time is right.

4. You don't have variables. What you do have, is access to memory. You put some data at a specific address and then you can read it or write to it later. Unlike other languages, such as C#, you can store a single byte and read it as part of a 4-byte value (i.e., including the bytes surrounding it), or you can store 4 bytes and only read 1 of them. Each byte is considered its own separate thing and you can interpret them however you like. The CPU doesn't know whether you're data is a string or an integer or even more code (since the latter is also stored as bytes), it depends on how your code treats it.

I advise you that _after_ you read the book I linked you to, you start using the official x86 documentation.

Intel CPUs: http://www.intel.com/content/www/us/en/processors/...

AMD CPUs: http://developer.amd.com/documentation/guides/Page...

Cheers,

Bogdan