Hi Privalov,

I'm just wondering what's on the to-do-list of fasm for the future??? (if you
have any)... ...sorry my English...

Tommy

There are a lots of them. The most important currently are:

1. The programming tutorial - this is really an essential if we want to attract new people to use flat assembler. This project is already started and you can see the actual results on this forum.
   
2. x86-64 implementation - not started yet, altough I've already made some architectural changes to prepare the ground for it. I'll start working on it when I finally get the printed manuals from AMD (hopefully soon). Actually I'm planning to start 2.x versioning when the x86-64 output is fully supported.
   
3. Source-level debugging - for this one all changes that were needed inside the core of flat assembler are already done, and were tested with my experimental listing-capable release. Now it's only a matter of writing debugger capable of getting that info from the fasm's core during the assembly and using it later for the debugging process. It would be also good to equip fasm with debug info generation for object formats - I'm considering to start this project soon, but yet I'm still concentrating mainly on the documentation/tutorials projects.
   
4. Finally trace out and fix all the bugs in AsmEdit/fasmw - it keeps crashing every now and then, and this is the reason why interface version of fasmw is still 0.90. It would be a great help if people for whom it's crashing provided some dumps of memory made on crash, since it nastily doesn't want to crash for me (probably it is related a bit to the style of using fasmw's editor).
   

Cool Privalov!

1) I do totally agree!
2) and 4) are the most important ones in my opinion...
3) Also important

Well, to sum all together: The plans seems to be perfect. Exactly what we need! (as always! )

Keep up the good work!

Regards,
Tommy

I couldn't agree more, what a great product.

> "AsmEdit/fasmw bugs" >>> Personal use ways.

I also have this terrific problem. Unfortunatly, it seems this is no other solution than waiting for users reports. There is always somebody doing something in a way you would never imagine or expect...

Once you will have Source Level Debugging, the reports will be more helpfull because most users will not only report the problem, but also... the fix...


Betov.

Glad all,

i think we need,

5.linker, loader
6.c compiler and c libraries written in fasm
7.python or Ruby style syntax + fasm

this is simply low priority, just opinion.

--

Assembly Language Debugger

Understanding the detailed Architecture of AMD's 64 bit Core

Original message
Hi there!

its been a long time since I posted ont his board,
as a matter of fact im still working on my YATF
project, i'll try to update it soon.

I'm sure nasm style listing output would be easy to
generate, but I don't have time to familiarize myself
with the source, I'm sure its a quick hack that
someone could produce:

address compiled bytes code

I know source level debugging is on the way,
but still this simple feature is very useful and
is absent from the flat assembler.

Another thing, when I push a label (push Routine)
in use32 mode`, it compiles the push imm8 instruction!
by default should be the dword, simple logic ;o)
once again, I dont know the source, that's my
lame excuse for not making it myself =)

Adam

Written after response below
Thanks Privalov for the thread and the basic
information (!!), to give back to the community
I'll try to finish my project promptly and
help in the fasm tutorial the best I can.

First: you have probably missed this thread: http://board.flatassembler.net/topic.php?t=213
Second: push imm8 in the 32-bit mode is just the short form of 32-bit instruction - it pushes the double word anyway - there's not such instruction as "push byte" really.

>># The programming tutorial -

i agree. an asm greenie like myself could really use programming turorials. icezellion's are great for masm, and i see imagineer and you working on translating them (good work and thanks!). i like imagineer's version since yours seems to be out of date concerning fasmw (i needed to fiddle around with the includes a bit which might throw off a complete newbie in programming).

also, your own "tutorial" in fasm is looking like a professionally writtin instruction book rather than a tutorial.

after i decided to dump java i thought about converting to hlm, read it a bit, than masm (they have many useful tools built in + tutorials). but i can not explain why after i downloaded the fasmw, i am drawn to it more than any other.

>># Source-level debugging

i played around with a cute toy for an hour : emu8080 v.247
http://www.ziplib.com/emu8086/Emu8086Setup.exe

it has disassemble, memory dump, live register display, single step code execution. is there any plan to implement such a monster into fasm? it would imo make a complete integrated system.

>># Finally trace out and fix all the bugs in AsmEdit/fasmw

the editor is probably the weakest link in fasm. i managed to hook up radasm with console fasm, but did not like that setup and came back to fasmw. i have not experienced any crash, but so far i work on small code.

if i am asking for too much, just tell me to shut up

best parts of fasm are no make, no link, and dos optional. thanks very much for those.

And what about IA-64? I believe we are not far from it...

IA-64 is something completely different, and in no way compatible with x86.

Do you mean that IA-64 FASM is not on the to-do-list at all? Another question: What do you think - how far we are from using this platform as we use IA-32 compatibile platform in these days?

The IA-64 (iirc intel's Itanium and Xenon CPUs) are pure server CPUs, the AMD64 (Opteron and Athlon64) are both targeted at server (opteron) and "pro"/workstation (Athlon64). I hope the AMD64 architecture will become as natural as the 80386 (from 16 to 32 bits, but now 32 to 64 ) (and that intel will follow, so that AMD don't get monopol on it, even if they're, IMO, the best some competition keeps the prices low and development up)
So I'd say we're pretty far away form the IA-64 (unles you're working on servers using it), and since it's incompatible with the IA32 (x86-32) I don't think well endup using it. I hope the compabillity will be as important as it were when A20 was added (the 20th addresbit), else many might hold on to the IA32 since they've spend so much time coding for it.

ehm, Hi..

the WAR(AMD vs Intel) was happened, Intel won..
(regardless of x86-compatibililty)

on the other side, i heard IA64 is more better than x86 architecture.
i think intel have enough power to kill AMD,
the lower price, the 90nm technology, SOI, more!!


--
PS: until some revolutional CPU-computer come, intel owned earth
i want hardware that make another hardware at low cost.

64bit is going to be forced onto people if they like it or not anyway...

eventually there will be no more 32bit hardware be sold and that software will migrate to 64bit..

also since microsoft is going to try forcing people to use the TC software and hardware wich is only going to be in there 64bit OS you wont have much of a choice..

http://www.cl.cam.ac.uk/~rja14/tcpa-faq.html

I don't hope so, because IMHO IA-64 is much more stronger than AMD x86-64. Intel have set up absolutely new architecture without direct x86 legacy - that's why is it strong. AMD haven't.

It's not incompatibile . Itanium cover IA-32, as-is in these days, absolutely. It has new special instructions to switch between IA-32 and IA-64 code.

It's compatible.

IMHO that's true.