I'm not new to assembly, but I am new to FASM. I found a really good tutorial under the Documentation section of this website, but quickly found out that it was for DOS machines. Any suggestions?

I'm sort of a newbie at linux asm as well. I don't know if any fasm tutorials for linux exist I found a nasm tutorial which was fairly helpful.

Thanks. Although I'd hate to betray FASM, I'll try it out.

Thanks again.
-Brandon

Don't consider it deserting FASM, but more like helping
TASM, NASM are closely related to FASM so you won't lose that much.

It wouldn't be impossible for someone (with Tomasz's permission, of course) to rewrite the DOS tutorial for linux. Wouldn't it be a matter of replacing int 21h examples with int 80h ones?

[content deleted]

if the same functions are mounted to the same numbers, it shouldn't be a problem. However, though i have never programmed for DOS, i don't see that as very likely. It's very difficult to start out with anything other than windows with fasm unless you're willing to use ld by command line and learn the order of params before you seriously code. Lots of amd64 examples, though that are readily assemblable in the examples section to help with using them. =) Just a matter of assembling your hello world then using man. If you've used fasm for anything else, not a problem. but, if you're new to fasm, i would recommend taking to programming for windows using wine until you get used to the fasm syntax if you're used to programming for windows as using a different assembler (why learn 2 new things at once if you can learn 2 things seperately?).

There are some good examples in the FreeBSD developer's handbook:

http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/x86.html

Hope that helps

By the way kohlrak is right. Fasm is not as great success on Linux (on other non windows systems even less) as it is on windows. It's really a pain to debug your code (even 64-bit) generated by fasm (absolutely no debug info). Its preprocessor capabilities have been designed with windows in mind. On Unix systems you mostly don't need to use as sophisticated macros (except if you want something fasm doesn't provide by default) as on windows. On Unices however there are infinite header files with infinite times infinite constants that you'd like to use without the necessity of tracking them down to their bare values that then you'd have to hard-code in your program (just look at the Linux examples, hard coded magic values everywhere, terrible). The other assemblers (nasm, yasm, etc.) may be better choice (even when it's about debugging), but I don't use them either. I simply use the gnu assembler (gas, as). It's part of the gnu compiler infrastructure that is defacto standard on the most popular Unix systems (Linux, BSDs, Mac), but it can also be found on the others (Windows, Solaris, AIX, etc.) as well. And it works with the c-preprocessor automatically and seamlessly. So if you are once used to it then you won't have any problems with other systems. You don't need to invoke gnu-ld or gnu-as directly, but of course you have to know some options, but it should not be a problem. As an example here the little hello world linux example for powerpc (I just compile it with the gnu powerpc cross-compiler and run it with qemu-ppc on my intel machine . The intel program looks almost exactly the same except the mnemonics between the labels _start and msg. And here the fasm code of the same powerpc program for the case you don't want to install the cross compiler (qemu you will still need):
It's maybe a little bit OFF and seems a gnu advocacy, but I just wanted to demonstrate you that on Unix systems you mostly need constants from headers but not macros overwhelming everything. And if you are used to an assembler (actually to a compiler infrastructure) then you're not lost on other systems/platforms (e.g. if you want to write some complicated calculations to your super computer consisted of some PS3s, or just want to learn another assembly. Qemu can emulate quite a few processors ).