sse big float value to integer 32bits

flat assembler
Message board for the users of flat assembler.

Index > Main > sse big float value to integer 32bits

Author

Thread

Roman

Joined: 21 Apr 2012
Posts: 1969

Roman 30 May 2025, 07:05

for example

Code:

;in 32bits program
mov eax,-95000000000.0
movd xmm0,eax
cvtss2si ecx,xmm0 ;I get not int -95000000000. I get -17563648

How get correct int value ?
I expected two regs. eax(lower part) and edx(height part)

30 May 2025, 07:05

Jessé

Joined: 03 May 2025
Posts: 59
Location: Brazil

Jessé 30 May 2025, 08:53

If I understand correctly, there are 2 problems: using single precision for such a big number may corrupt the result; you're trying to fit a conversion result number larger than a 32 bit number into a 32 bit register.
For this, under 32 bit, the best approach might be the FPU:

Code:

; ...
    sub   esp, 8
    finit
    fld    qword [bigfloat]
    fistp  qword [esp]
    pop    eax
    pop    edx

bigfloat    dq -95000000000.0

30 May 2025, 08:53

Roman

Joined: 21 Apr 2012
Posts: 1969

Roman 31 May 2025, 05:15

Thanks.
I forgot about fpu.
Long time using sse,avx

31 May 2025, 05:15

< 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