The Program Status Register

We've seen how bits on a register can be used to give the answers of yes/no and true/false statements. On your computer, there is a register called the program status register. This register holds a lot of information about what happens in a computation. For example, have you ever wondered what would happen if you added two numbers and the result was larger than would fit in a register? The program status register has a flag called the carry flag. You can test it to see if the last computation overflowed the register. There are flags for a number of different statuses. In fact, when you do a compare (cmpl) instruction, the result is stored in this register. The conditional jump instructions (jge, jne, etc) use these results to tell whether or not they should jump. jmp, the unconditional jump, doesn't care what is in the status register, since it is unconditional.

Let's say you needed to store a number larger than 32 bits. So, let's say the number is 2 registers wide, or 64 bits. How could you handle this? If you wanted to add two 64 bit numbers, you would add the least significant registers first. Then, if you detected an carry, you could add I to the most significant register. In fact, this is probably the way you learned to do decimal addition. If the result in one column is more than 9, you simply carried the number to the next most significant column. If you added 65 and 37, first you add 7 and 4 to get 12. You keep the 2 in the right column, and carry the one to the next column. There you add 6, 3, and the 1 you carried. This results in 10. So, you keep the zero in that column and carry the one to the next most significant column, which is empty, so you just put the one there. Luckily, 32 bits is usually big enough to hold the numbers we use regularly.

Additional program status register flags are examined in Appendix B.