flat assembler
Message board for the users of flat assembler.

flat assembler > Main > undefined instructions

Author
Thread Post new topic Reply to topic
Ali.Z



Joined: 08 Jan 2018
Posts: 232
fasm1:

missing UD0 UD1, but it have UD2.
according to intel's manual ud0 and ud1 were mentioned as well as their opcodes.

_________________
Asm For Wise Humans
Post 22 Dec 2018, 10:51
View user's profile Send private message Reply with quote
Tomasz Grysztar
Assembly Artist


Joined: 16 Jun 2003
Posts: 7367
Location: Kraków, Poland
Post 22 Dec 2018, 12:46
View user's profile Send private message Visit poster's website Reply with quote
Ali.Z



Joined: 08 Jan 2018
Posts: 232
thanks for the info.

i noticed fasm have some undocumented instructions as well, is UMOV really replaced?

_________________
Asm For Wise Humans
Post 22 Dec 2018, 13:30
View user's profile Send private message Reply with quote
Tomasz Grysztar
Assembly Artist


Joined: 16 Jun 2003
Posts: 7367
Location: Kraków, Poland
As mentioned in the thread I linked, I had not been implementing instructions of third parties (non-Intel) other than AMD into fasm.

Also, currently available macro sets for fasmg include only Intel instruction sets (not even AMD-specific ones like 3DNow! or XOP). However, they include some undocumented instructions not implemented in fasm, like XBTS/IBTS in 80386.INC. The difference between instructions like XBTS/IBTS and ones like SALC is that the former were available only on 80386, while SALC continues to be supported by later CPUs, including modern ones. With fasmg you can select a later core using a header like P5.INC and then XBTS/IBTS become unavailable.

Nonetheless, with all instruction sets for fasmg being implemented simply in form of macros, it should be relatively easy for anyone to write additional ones for instruction specific to x86 variations from other vendors. You may find implementation of SCALL instruction of C&T Super386 CPU as a demonstration.

This is also something I might discuss in one of my streams - perhaps I could implement an few instruction encoders live.
Post 04 Jan 2019, 17:21
View user's profile Send private message Visit poster's website Reply with quote
Ali.Z



Joined: 08 Jan 2018
Posts: 232
its always good it include undocumented instructions for specific intel models, just for completeness sake.

true i did not use fasmg (didnt even bother downloading it) but i dislike the idea of including specific files for specific instruction set.

its always good to let the flat assembler to check whatever instruction you are using and assembles it without bothering yourself with inc files.

going off topic little bit:
arent macros slower to process? during assembling and encoding stage.

or they are just faster than dumping all instructions in one file, like tables.inc from fasm1.

_________________
Asm For Wise Humans
Post 04 Jan 2019, 22:42
View user's profile Send private message Reply with quote
Tomasz Grysztar
Assembly Artist


Joined: 16 Jun 2003
Posts: 7367
Location: Kraków, Poland
Ali.A wrote:
going off topic little bit:
arent macros slower to process? during assembling and encoding stage.

or they are just faster than dumping all instructions in one file, like tables.inc from fasm1.
Please read the first section of Introduction to fasmg to get a general idea. In short: yes, they are slower.
Post 04 Jan 2019, 23:06
View user's profile Send private message Visit poster's website Reply with quote
ProMiNick



Joined: 24 Mar 2012
Posts: 356
Location: Russian Federation, Sochi
in continue to offtopic:
processing of all instruction is in ASSEMBLE.INC (X86_64.INC and AVX.INC too).not in tables.inc.
tables.inc holds only pointers for processing handlers.

of course macroprocessing is slower than built in processing, because built in processing operates on in memory binary structures, but macros emulate such process operating only on text representation of such structures.

Dumping all instructions in one file isn`t enought. Coder should know not only specifics of current architecture but internal structure of flatassembler itself. Look at revolution work - there not only table of ARM instruction set, but much much work over it (from processing excluded sizing operators, prefixes, segmentation ... added to proccessing masked execution suffixes to instructions, instruction optimization there varied from cpu model(instruction set allowed) and so on ...),

In case of fasmg coder not needed to know fasmg internal architecture and able to realize any instruction set via macros. And it is not faster (but it is much simpler to end programer realize new architecture).
Post 05 Jan 2019, 00:04
View user's profile Send private message Send e-mail Reply with quote
Display posts from previous:
Post new topic Reply to topic

Jump to:  


< 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


Copyright © 1999-2019, Tomasz Grysztar.

Powered by rwasa.