Файл: Microcontroller Programming. Thi Micro Chip PIC (Julio Sanchez, 2007).pdf

Note in Figure 4-3 that the contents of the carry flag are first copied to the low-order bit of the destination operand, then the individual bits of the source (in gray in the illustration) are shifted left and moved to the destination. Finally the high-order bit of the source is copied to the carry flag.

There are several possible variations of the rotate operation. The Intel microprocessors distinguish between arithmetic and logic rotates. In the arithmetic rotation the high-order bit is preserved in order to maintain the sign of the operand. The rotate shown in Figure 4-3 is the one most common in microcontroller hardware. Clearing the carry flag before the rotate takes place makes the operation identical to a shift.

4.3.2 Comparison Operations

An interesting property of subtraction is its use in finding the relative size of two operands. This interesting action of subtraction is based on the following logic:

1.If the result of a subtraction is zero, then both operands were of the same size.

2.If the result of a subtraction is a positive number, then the subtrahend was smaller than the minuend.

3.If the result of a subtraction is a negative number, the subtrahend was larger than the minuend.

In a binary/digital device the result of a subtraction operation can be determined by observing the flags. If the zero flag is set, then the operands were the same (case 1, above). If the carry flag is set, then the subtrahend was larger than the minuend (case 3, above). If neither the carry nor the zero flag is set, then resulting subtrahend was smaller than the minuend (case 2, above). Since all microcontrollers offer some mechanism for re-directing execution according to the state of the flags, a program can use subtraction to make these decisions.

The one objection to the use of subtraction in comparing the size of two operands is that the process will change one of them. To use subtraction in comparison operations the programmer has to find some way of preserving the minuend. Alternatively, some devices contain a comparison operator that sets the flags as if a subtraction had taken place but without changing the operands. High-end microcontrollers are equipped with dedicated comparison operators, but the middle- and low-range devices usually are not.

4.3.3 Other Support Operations

Midand high-range microcontrollers contain other auxiliary bitwise, arithmetic, and logic operators that can be useful to the programmer. These include instructions to:

1.Increment and decrement operands

2.Clear registers or storage locations

3.Swap nibbles

4.Clear and set individual bits

5.Test individual bits

Usually instructions to increment and decrement and to test individual bits are also capable of redirecting execution according to the result. For example, a special decrement can be followed by a jump if decrementing sets the zero flag. Or a bit test instruction can include a jump that is taken if the tested bit is set or reset.

4.4 Unsigned Binary Arithmetic

Since microcontrollers are not used in data processing, microcontroller programming does not usually require the development of powerful or sophisticated numerical routines. At the same time, because microcontrollers often lack primitive support for even the most essential calculations, the programmer makes up for this deficiency. For example, mid-range PIC microcontrollers contain primitive instructions for signed and unsigned addition and subtraction of byte-size operands. Unsigned addition and subtraction operations that exceed one byte, as well as unsigned multiplication and division, must be provided in software.

In unsigned arithmetic all bits of the binary encoding are interpreted as magnitude bits and all numbers are positive. Addition of unsigned binary numbers is limited by the machine’s word size. For example, a mid-range PIC microcontroller performs unsigned addition on 8-bit operands. An overflow of the sum is reported by the carry flag set. In this case the carry flag clear indicates that the sum is within the storage capacity of the format. In unsigned arithmetic processing, routines for extending operations to multiple bytes are straightforward and relatively simple.

4.4.1 Multi-byte Unsigned Addition

Many microcontrollers are one-byte machines, so operands and results for arithmetic operations must be contained within eight bits. The largest unsigned value that can be represented in a single byte is the decimal number 255. But often applications require adding operands that are larger than a single byte and storing results that exceed this limit. In these cases multi-byte routines become necessary.

The simplest case is the addition of two unsigned byte-size operands whose sum exceeds 255 decimal. This case requires storing the result in a two-byte area and detecting those cases in which there is a carry into the high-order byte. In this case the largest possible operands for byte addition are the hexadecimal numbers FF. Addition is as follows:

Binary:

1 1 1 1 1 1 1 1

+1 1 1 1 1 1 1 1

---------------

11 1 1 1 1 1 0

C <=

In this example the symbol C <= represents a carry out of the high-order bit, the case when the sum exceeds the capacity of a single byte. In hexadecimal, the sum of HFF + HFF = H1FE. You can add two byte-size operands into a two-byte storage area by using byte addition to determine the low-order byte of the result and testing for a carry out of the high-order bit. If there is a carry, then the high-order byte of the result is 1; otherwise the high-order byte is 0.

ing operator. This section explains multiplication in devices that lack a dedicated multiplication operation code.

Arithmetically, multiplication is performed by repeated addition. The multiplier represents the number of times that the multiplicand must be added to itself. Therefore, 3 times 4 is the same as 3 + 3 + 3 + 3. This fact allows implementing multiplication routines in software as long as the device contains an addition operator. The logic is based on using the multiplier as a counter. This counter is decremented each time that the multiplicand is added to itself. The routine ends when the counter is exhausted, as shown in the flowchart in Figure 4-5.

START

A = MULTIPLICAND

B = MULTIPLIER

PERFORM P = A * B

P (PRODUCT) = 0

C (COUNTER) = B

Figure 4-5 Unsigned Multiplication Flowchart

The beauty of the repeated addition algorithm is its simplicity and its main shortcoming is its slowness. An alternative way of performing multiplication is by shifting the bits of the operand. This method is based on the properties of a binary positional system, in which the value of each digit is a successive power of 2. Therefore, by shifting all digits to the left, the value 0001B (1 decimal) successively becomes 0010B (2 decimal), 0100B (4 decimal), 1000B (8 decimal), and so on.

Binary multiplication by means of bit shifting has the downside that the multiplier must be a power of 2. Otherwise, the software must shift by a power of 2 that is smaller than the multiplier and then add the multiplier as many times as necessary to complete the product. In this manner, to multiply by 5 we can shift left twice and add once the value of the multiplicand. To multiply by 7 we would shift left twice and then add three times the value of the multiplicand. As the multiplier gets larger and more distant from the smaller power of 2, the number of addition operations required is also larger, and the effectiveness of the algorithm diminishes.