Floating Point examples

Index > Main > Floating Point examples

Author

Thread

me239

Joined: 06 Jan 2011
Posts: 200

me239

Hey guys, I've been looking all over for some SIMPLE examples of floating point math done in assembly and printing it. Preferably 16 bit, thanks!

13 Oct 2011, 22:35

Fred

Joined: 22 Oct 2010
Posts: 39

Fred

Floating point math in SSE is very easy to do.

Code:

movd  xmm0,[esp+4] ;move some float from the stack to xmm0[0-31]
movd  xmm1,[esp+8] ;^ but to xmm1[0-31]
addss xmm0,xmm1    ;add scalar single (xmm[0-31] + xmm[0-31])

That's about as simple of an example you'll get Very Happy

I don't think there's any instructions for 16 bit floats... yet. I'm pretty sure some new processor was going to support 32 <-> 16 bit float conversion though.

Printing floats would require some windows calls(if you're on windows anyway) and that's unfortunately not my area of expertise. Razz

14 Oct 2011, 06:25

macgub

Joined: 11 Jan 2006
Posts: 283
Location: Poland

macgub

The same code done on FPU:

Code:

fninit
fld     dword[esp+4]   ; single precision, 32 bits.
fadd  dword[esp+8]   ; st0 - result

18 Oct 2011, 10:05

Akujin

Joined: 11 Oct 2011
Posts: 26
Location: Exile

Akujin

Internals of FPU and MA$M examples.
http://www.ray.masmcode.com/downloads/SimplyFPU1-6.zip

_________________
CLI
HLT

18 Oct 2011, 19:25

edfed

Joined: 20 Feb 2006
Posts: 4267
Location: Now

edfed

simple example:

compute ebx=eax*1.32456

Code:

example1:
push dword 1.32456 eax ;preload in stack
finit     ;init the fpu
fld [ss:esp+4] ;load float 1.32456
fimul [ss:esp] ;mul by eax value
fistp [ss:esp+4] ;store in second pop
pop eax eax ;pop two times, free the stack
ret ;quit

compute eax=eax*sin(eax*pi)

Code:

example2:
push eax
finit
fldpi ;load pi value
fimul [ss:esp] ;mul by eax
fsin ;sinus of st0 in st0
fimul [ss:esp]
fistp [ss:esp]
pop eax
ret

the fpu is simple in fact.

18 Oct 2011, 21:08

me239

Joined: 06 Jan 2011
Posts: 200

me239

edfed wrote:

simple example:

compute ebx=eax*1.32456

Code:

example1:
push dword 1.32456 eax ;preload in stack
finit     ;init the fpu
fld [ss:esp+4] ;load float 1.32456
fimul [ss:esp] ;mul by eax value
fistp [ss:esp+4] ;store in second pop
pop eax eax ;pop two times, free the stack
ret ;quit

compute eax=eax*sin(eax*pi)

Code:

example2:
push eax
finit
fldpi ;load pi value
fimul [ss:esp] ;mul by eax
fsin ;sinus of st0 in st0
fimul [ss:esp]
fistp [ss:esp]
pop eax
ret

the fpu is simple in fact.

now how would I print one of those values to the prompt?

19 Oct 2011, 03:51

emc

Joined: 20 Aug 2011
Posts: 90
Location: France

emc

You can use printf with a formatted string (with %f) to print floating values.

22 Oct 2011, 06:49

me239

Joined: 06 Jan 2011
Posts: 200

me239

emc wrote:

You can use printf with a formatted string (with %f) to print floating values.

I'm using DOS

22 Oct 2011, 23:16

< 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