Add, Sub, Mul, Div

[Tisler]

Among those four operations, could you tell me which one run fastest ? I ask this here because I am taking a basic course on assembly and having a home work to solve bit counting problem.... I am wondering if i can make my program run a bit faster if I could make a correct choice, although admittedly, everything is just what I am....intending to do. I am not sure about this little problem though,

Thank you for any help,

Andreas Tisler

[Hobson]

Unfortunately, instruction clocks were included only in reference to 486 proccessor, and I cannot find them in new Intel Instruction Set Reference. So these numbers can be a bit outdated.

ADD
                                 Clocks                 Size
        Operands         808x  286   386   486          Bytes

        reg,reg           3     2     2     1             2
        mem,reg         16+EA   7     7     3            2-4  (W88=24+EA)
        reg,mem          9+EA   7     6     2            2-4  (W88=13+EA)
        reg,immed         4     3     2     1            3-4
        mem,immed       17+EA   7     7     3            3-6  (W88=23+EA)
        accum,immed       4     3     2     1            2-3


SUB
                                 Clocks                 Size
        Operands         808x  286   386   486          Bytes

        reg,reg           3     2     2     1             2
        mem,reg         16+EA   7     6     3            2-4  (W88=24+EA)
        reg,mem          9+EA   7     7     2            2-4  (W88=13+EA)
        reg,immed         4     3     2     1            3-4
        mem,immed       17+EA   7     7     3            3-6  (W88=25+EA)
        accum,immed       4     3     2     1            2-3

MUL
                                 Clocks                 Size
        Operands         808x  286   386   486          Bytes

        reg8            70-77   13   9-14  13-18          2
        reg16          118-113  21   9-22  13-26          2
        reg32             -     -    9-38  13-42         2-4
        mem8        (76-83)+EA  16  12-17  13-18         2-4
        mem16     (124-139)+EA  24  12-25  13-26         2-4
        mem32             -     -   12-21  13-42         2-4

DIV
                                 Clocks                 Size
        Operands         808x  286   386   486          Bytes

        reg8             80-90  14    14    16            2
        reg16           144-162 22    22    24            2
        reg32              -    -     38    40            2
        mem8        (86-96)+EA  17    17    16           2-4
        mem16     (150-168)+EA  25    25    24           2-4  (W88=158-176+EA)
        mem32              -    -     41    40           2-4
SHL/SAL & SHR
                                 Clocks                 Size
        Operands         808x  286   386   486          Bytes

        reg,1             2     2     3     3            2
        mem,1           15+EA   7     7     4           2-4  (W88=23+EA)
        reg,CL           8+4n  5+n    3     3            2
        mem,CL        20+EA+4n 8+n    7     4           2-4  (W88=28+EA+4n)
        reg,immed8        -    5+n    3     2            3
        mem,immed8        -    8+n    7     4           3-5

So, you can notice that you can put like 5-10 ADD\SUB\SHx instructions in place of single MUL or DIV.

And also a little piece of advice:
If you are not going to count all bits set in area of like 50MB of RAM, just use algorithm which is SIMPLER, not FASTER. Its easy to make problems hard with assembler and unnecessary optimizations.

Hob