MSP430/Assembly Calculations

Ben Trigoboff said:

I'm having problems with moving -7 into a register in the first place, it keeps showing up as 0x00F9 = 249. I don't know if this is correct or not because our Professor informed us to simply enter a negative integer as such: #-7.

That's correct, it's due to the fact that negative values are stored in two's complement format on MSP430 (and many other CPUs too). 

EDIT: I think this page does a better job of explaining negative number representation than the main two's complement page: http://en.wikipedia.org/wiki/Signed_number_representations

>moving -7 into a register in the first place, it keeps showing up as 0x00F9 = 249

That is correct, if you use signed instructions on these numbers everything works out correct in the end.
signed 8bit numbers are from -128 to 127,
But if you look at them in a purely 8bit representation they will of course be in the range of 0-255
 
Mixing signed 8bit in to signed 16bit result can you have to make sure that you do things correct.

Comparison of signed integers:
JGE: Jump if greater or equal
JL: Jump if less
JN: Jump if negative

SXT: Extend Sign
RRA: Rotate right arithmetically
RLA: Rotate left arithmetically

The first part should result in: -55 (0xC9 in memory)

(-55)*(-55)= 3025

This is because to square a number just means to multiply it by itself.
For example, (-2) squared is (-2)(-2) = 4.
This is positive because when you multiply two negative numbers you get a positive result.

>add.w   R13, R13                ;perform squaring operation
looks more like multiply by 2 to me

Ben Trigoboff said:

create another loop and add 55 to itself 55 times, or is there a more efficient algorithm in assembly that I am unaware of?

 At first is this MSP430 related? I think no.

 To do multiplication which device are you planning to use? Has this multiplication hardware or not? Can move multiply terms to hardware register then read the result or call a software multiply, if you need do as exercise I think basic shift and add is better than add loop.. Programming zen teach not to do algorithm increasing time in exponential way or time dependent on value of operand.

My old post where I do 8bit *8bit in assembly pretty fast

e2e.ti.com/.../340445

It don't work for signed bytes, but as both are negative just prepare data:
inv.b R13 
inc.b R13
mov.b R13.,R12
call #multiply

software multiplication is a general term and is not very specific to the MSP430, so that is why he recommend to google it.

>Looks good, but slightly beyond what I'm familiar to, as I don't quite understand what rotate through carry does

http://www.physics.mcmaster.ca/phys3b06/MSP430/Instruction_Set.pdf

Swap switches bytes in a words, also useful for a shortcut for: byte*256=word or byte<<8

This maybe make it little easier to understand what it does (Parallax assembler)
it's using lower 8bit shifted out in R13 to know when to add R12 to upper bits of R13,
but as R13 is the same register that is shr, its true value in the end emerge.

                shl     R12, #7        ' prepare R12 so it adds to the upper bits of R13 
                shr     R13, #1 wc     ' bit 0
        if_c    add     R13, R12       ' add R12 to > R13 (in pasm source is last)
                shr     R13, #1 wc     ' bit 1
        if_c    add     R13, R12
                shr     R13, #1 wc     ' bit 2
        if_c    add     R13, R12
                shr     R13, #1 wc     ' bit 3
        if_c    add     R13, R12

Ben Trigoboff said:

. without the same knowledge. Good learning experience though as our whole class is basically researching and utilizing alternate resources to figure these things out.

 That's called "home work" so I am pay'd to do lesson to my pupils not to give free lesson, sorry this is MSP430 related forum and if something is relative to specific problem welcome here, if this has to switch on a free lesson to somewhere professor take money and we lose .. this is not a good idea for us we are here to share experience not to teaching job, none here pay us.

 Rotate thru carry are for extending arithmetic operation over multi byte/word sizes and all processor work same way from era of electromechanics to mainframe and today computer...

http://www.computerhistory.org/fellowawards/hall/bios/Konrad,Zuse/

 please tell your professor and your colleagues university has to teach and student have to learn about here we don't solve troubles of and we can do limited support. Sucking this resource shorten future of technology and industry leaving unprepared people to real world.