I know that when I receive 2 float number it automatically converts it to IEEE Format.

I want to split, sign, mantisa, and exponent in 3 different memory or register positions.

Any advice on how to make this.

Thanks in advance
Posted on 2005-04-13 16:00:36 by luise.valencia
Perhaps you should look for the IEEE document explaining the floating-point format? Or the notes given to you by your teacher?
Posted on 2005-04-13 16:51:16 by f0dder
I know that. I am asking is for the assembly sintax or hints about it. I already  know the theory
Posted on 2005-04-13 18:37:57 by luisvalencia
If the IEEE format number is not in memory then you'll have to store it at an address:

fst FLOAT PTR

Once the number is in memory, you load it into a general purpose register:

mov eax, DWORD PTR

All you have to do now is shift the number around and store the bits - it is only a couple of instructions required. I hate to be so curt, but if you don't have the basic knowledge to complete this task then maybe reading the x86 manual is needed. Shifting is done with SHR/SHL and you might want to use AND to mask the parts you desire.

Please, note I have used "FLOAT" because I don't know which size IEEE value you desire.
Posted on 2005-04-13 23:15:02 by bitRAKE
the user is going to input two single numbers.
Posted on 2005-04-14 11:10:42 by luise.valencia