flat assembler
Message board for the users of flat assembler.

Index > Heap > Java for DOS (32bit, 64bit)

Author
Thread Post new topic Reply to topic
sleepsleep



Joined: 05 Oct 2006
Posts: 8898
Location: ˛                             ⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣Posts: 334455
sleepsleep
i saw dexu thread about he currently coding a 32bit DOS

then i ponder if there exists JAVA jre and jdk or (open source java) for DOS ...

or is there 64bit DOS ? i think it would be really COOL!!!!...

DOS as the core OS, then Java sit on top of it deal with all the user interactions.

then DOS 32/64bit driver???

then just wanna ask

is that possible to use 64bit register as if two 32bits that run concurrent as if they separate entity...

ok, what i mean is like below..

if let say i got some billion billion summation that their total i already knew before hand will within 32 bit range.

eg.
1. 5 + 45
2. 8 + 283

what i mean is, could we use the register together (since we will deal with ADD instruction in this example)

so, i put the value in correct 32 bit column

so 64bit = 0xFFFFFFFF | 0xFFFFFFFF
so i put them in below format

RAX = 00 00 00 05 00 00 00 2D
ADD = 00 00 00 08 00 00 01 1B
total = 00 00 00 0D 00 00 01 48

so it is the software part to draw the result out from the total

or, people already coding in such a way??
Post 05 Jan 2011, 19:06
View user's profile Send private message Reply with quote
b1528932



Joined: 21 May 2010
Posts: 287
b1528932
your question has no sense.
when you put something inno eax, 32bit high part of rax will be zero extended.
Post 05 Jan 2011, 19:17
View user's profile Send private message Reply with quote
sleepsleep



Joined: 05 Oct 2006
Posts: 8898
Location: ˛                             ⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣Posts: 334455
sleepsleep
maybe my diagram a bit weird... Surprised

but basically, i mean, i want to count 2 summation in 1 time ADD instruction.

so, if 5 + 8

and 45 + 283 ... i would craft my RAX in above so i could get the result in 1 time ADD... instead of 2.

so if i will count 2 billion or more summation, i will always use 1/2 of it.

hmmm, i don't know if my question makes sense or not... but i just wanna know if people already code like this or it just wasting time to code like this.
Post 05 Jan 2011, 19:31
View user's profile Send private message Reply with quote
neville



Joined: 13 Jul 2008
Posts: 507
Location: New Zealand
neville
I think I understand what you mean sleepsleep and I think it works so long as there are no half-carries from lower 32-bits to upper 32-bits - if results fit within 32 bits as you have said.

But since you can't directly address upper 32 bits of long registers I think any time saved in arithmetic instructions would be taken by masking and shifting data to the right place.

Btw, I thought in long mode if you address EAX the upper 32-bits of RAX are not touched i.e. not zero extended as suggested by b152... ??

_________________
FAMOS - the first memory operating system
Post 06 Jan 2011, 01:45
View user's profile Send private message Visit poster's website Reply with quote
asmhack



Joined: 01 Feb 2008
Posts: 431
asmhack
Results of 32-bit operations are implicitly zero extended to 64-bit values.
This differs from 16 and 8 bit operations, that don't affect the upper part of registers.
Post 06 Jan 2011, 02:07
View user's profile Send private message Reply with quote
sleepsleep



Joined: 05 Oct 2006
Posts: 8898
Location: ˛                             ⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣Posts: 334455
sleepsleep
Quote:

I think I understand what you mean sleepsleep and I think it works so long as there are no half-carries from lower 32-bits to upper 32-bits - if results fit within 32 bits as you have said.

yeah,

Quote:

I think any time saved in arithmetic instructions would be taken by masking and shifting data to the right place.

i was thinking about this,, but wonder if there if people who coded like this? (eg. crafted their data in such a way for the processor to calculate)

because under normal (daily) use,,, hmm, eg. accounting, or point of sales or general human financial record, very few who actually reach the 32bit register limit.... most people deal with number far less than 32bit register limit.
Post 06 Jan 2011, 06:04
View user's profile Send private message Reply with quote
edfed



Joined: 20 Feb 2006
Posts: 4237
Location: 2018
edfed
i still have made some lists about jvm for fasm.

now you just have to implement every opcodes, let say, it represent approximatelly 1ppm of the job, but will be needed if any body wants to dev something about JVM using fasm.

let me know if it have any usefullness.


Description: all basic structures needed to interpret the byte code.
Download
Filename: jvmasm.zip
Filesize: 5.92 KB
Downloaded: 73 Time(s)

Post 06 Jan 2011, 08:50
View user's profile Send private message Visit poster's website Reply with quote
sleepsleep



Joined: 05 Oct 2006
Posts: 8898
Location: ˛                             ⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣Posts: 334455
sleepsleep
/me was thinking how to use edfed code...
Post 06 Jan 2011, 15:56
View user's profile Send private message 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 can attach files in this forum
You can download files in this forum


Copyright © 1999-2020, Tomasz Grysztar.

Powered by rwasa.