Bitwise And

Code:

---

#include <stdio.h>

int display_binary(unsigned short number);


int main()
{
    unsigned short number;


    printf("Enter a whole number between 0 and 65,535: ");
    scanf("%hu", &number);


    printf("\n\n    Decimal: %hu\n", number);
    printf("Hexadecimal: %x\n", number);


    display_binary(number);


    getch();
    return 0;
}


int display_binary(unsigned short number)
{
    unsigned short mask = 0x8000;
    int c, k;


    for(c = 0; c < 16; c++)
    {
        k = (number & mask);
        if(k == 0)
            printf("0");
        else
            printf("1");


        mask >> 1;


        if(c == 3 || c == 7 || c == 11)
            printf(" ");
    }
}

---

OK, I know that I am not supposed to just blatantly ask questions for homework, but I can't seem to figure this out... Here is what I was told to do:

Quote:

---

Hint: Declare an unsigned short mask = 0x8000 (which is 1000 0000 0000 0000). Write a for loop that iterates 16 times. In each cycle, do a bitwise & between the mask and the input parameter. Display 0 if the outcome is 0 and 1 otherwise. Shift the mask to the right one bit at the end of each cycle to retrieve the next bit in the next cycle. Display a space character after every 4 bits of data to improve readability.

---

My problem is, that I am getting the binary to be all "1's" for everything I put in. I think I am doing what is instructed, but obviously not.

Thanks in advance for any help,
~Alan

Code:

---

bitwise.c: In function ‘main’:
bitwise.c:22: warning: implicit declaration of function ‘getch’
bitwise.c: In function ‘display_binary’:
bitwise.c:48: warning: control reaches end of non-void function

---

getch is not portable, use getchar instead. Also, make display_binary return void, since you don't actually need to return anything.

Your problem is from the following line:

Code:

---

mask >> 1;

---

That doesn't actually do anything. That shifts mask to the right 1, but doesn't store the result anywhere, it's like saying x + 3; x doesn't change, nor mask in your case. It's causing your program to only look at the leftmost bit every time through the loop. Try:

Code:

---

mask >>= 1;

---

Line 42... mask = mask >> 1; or simply mask /= 2;

line 13... you should add some range checking here to make sure you don't get negative numbers or numbers > 65535.



edit: I see anduril beat me to it :D

Quote:

---

Originally Posted by anduril462

Code:

---

bitwise.c: In function ‘main’:
bitwise.c:22: warning: implicit declaration of function ‘getch’
bitwise.c: In function ‘display_binary’:
bitwise.c:48: warning: control reaches end of non-void function

---

getch is not portable, use getchar instead. Also, make display_binary return void, since you don't actually need to return anything.

Your problem is from the following line:

Code:

---

mask >> 1;

---

That doesn't actually do anything. That shifts mask to the right 1, but doesn't store the result anywhere, it's like saying x + 3; x doesn't change, nor mask in your case. It's causing your program to only look at the leftmost bit every time through the loop. Try:

Code:

---

mask >>= 1;

---

---

Thanks, that did it. :D

On a side note, what's the difference between getch and getchar? I'm not exactly sure how being 'portable' affects the outcome.

Quote:

---

Originally Posted by Alan Gott

On a side note, what's the difference between getch and getchar? I'm not exactly sure how being 'portable' affects the outcome.

---

getchar() is a standard function sanctioned by the C-99 standard, getch() is not. What this means is that one cannot reasonably expect to find getch() in every compiler's library, making your source code uncompilable on some systems. For example: I use Pelles C, which has it but in a different header, so your code does not compile for me *as is*, I had to add #include <conio.h> to make it work.

Thanks! That makes sense.

Quote:

---

Originally Posted by CommonTater

line 13... you should add some range checking here to make sure you don't get negative numbers or numbers > 65535.

---

Yeh, I figured that was needed. But my lab instructor frowns upon extra work. lol

Quote:

---

Originally Posted by Alan Gott

Yeh, I figured that was needed. But my lab instructor frowns upon extra work. lol

---

Any teacher who would discourage initiative in his students, should not be teaching.

Quote:

---

Originally Posted by CommonTater

Any teacher who would discourage initiative in his students, should not be teaching.

---

She claims that it's a waste of time since it isn't needed for the grade, it's probably because I turned my lab on switch statements from 3 to 15 cases to make it more detailed, and she has to read through everything before grading it. lol

Anyways, I have a second question. I am supposed to do this:

Quote:

---

The 3 leftmost bits represent the command, the lowest 4bits represent the frequency, and the 8th bit from the right, represents a switch value (on /
off). Your program should also display the input number in binary to verify the extracted
information.
Example: If the input number is 41851, which is 1010 0011 0111 1011 in binary.
The command value is 101 or 5 in decimal, the frequency is 1011 or 11 in decimal, and
the 8th bit from the right is 0, which means switch is off.

---

The only problem is, that my function "display_binary" can't return anything (as she say's not to in the instructions), and the frequency, switch and command are supposed to be extracted and displayed from the main function.

So, I can't return the values from the display_binary function, and I can't simply display them from that function either... Any suggestions?

Extracted and displayed in the same function? If not, create new functions to get each of those values, otherwise, pass in pointers to arguments and send back the values that way:

Code:

---

void display_binary(unsigned short number, unsigned short *command, unsigned short *frequency, unsigned short *switch)
{
    ...
    *command = // some bitwise operations to extract the command portion
    // likewise for frequency and switch
    ...
}

---

Quote:

---

Originally Posted by anduril462

Extracted and displayed in the same function? If not, create new functions to get each of those values, otherwise, pass in pointers to arguments and send back the values that way:

Code:

---

void display_binary(unsigned short number, unsigned short *command, unsigned short *frequency, unsigned short *switch)
{
    ...
    *command = // some bitwise operations to extract the command portion
    // likewise for frequency and switch
    ...
}

---

---

Like, after that function displays the binary I can't use it anymore. I have to somehow get the information I need while only being able to write in the main function.

Code:

---

int main()
{
...
display_binary(number);
//code for extracting/displaying the desired information goes here.
...
}

---

To extract just a set of bits, you need to first mask off the bits you don't want, using bitwise &, then shift them over so they're all the way to the right. Read this: http://www.cprogramming.com/tutorial...operators.html.

Quote:

---

Originally Posted by Alan Gott

She claims that it's a waste of time since it isn't needed for the grade, it's probably because I turned my lab on switch statements from 3 to 15 cases to make it more detailed, and she has to read through everything before grading it. lol

---

One has to wonder what some of these screwball teachers are thinking...
Is she teaching you how to program... or how to get good grades?

Quote:

---

Anyways, I have a second question. I am supposed to do this:


The only problem is, that my function "display_binary" can't return anything (as she say's not to in the instructions), and the frequency, switch and command are supposed to be extracted and displayed from the main function.

So, I can't return the values from the display_binary function, and I can't simply display them from that function either... Any suggestions?

---

That should be no problem...
For example:

Code:

---

frequency = number & 0xF;  // keep only 4 lsb
// etc.

---

Quote:

---

Originally Posted by anduril462

To extract just a set of bits, you need to first mask off the bits you don't want, using bitwise &, then shift them over so they're all the way to the right. Read this: Cprogramming.com - Tutorials - Bitwise Operators and Bit Manipulations in C and C++.

---

Wouldn't that involve doing the calculations in display_binary then return them to main? (If not, I'm totally clueless to what you are trying to teach me. :S

Quote:

---

Originally Posted by CommonTater

That should be no problem...
For example:

Code:

---

frequency = number & 0xF;  // keep only 4 lsb
// etc.

---

---

Thanks! That worked wonders! Does & 0xF do exactly what I did in my display_binary function without having to loop and whatnot like I had to do?

Quote:

---

Originally Posted by Alan Gott

Wouldn't that involve doing the calculations in display_binary then return them to main? (If not, I'm totally clueless to what you are trying to teach me. :S



Thanks! That worked wonders! Does & 0xF do exactly what I did in my display_binary function without having to loop and whatnot like I had to do?

---

No...

You had to do all that looping to display the binary value underlying your variable.

Bitwise and ( & ) operates directly on that underlying value.

Do a little reading on the logic functions AND OR NEG and XOR for a better insight.

Quote:

---

Originally Posted by Alan Gott

Wouldn't that involve doing the calculations in display_binary then return them to main? (If not, I'm totally clueless to what you are trying to teach me. :S

---

No. What I was getting at is this (note, these numbers are not correct for your example, you have to modify them):

Code:

---

command = (number & 0xF000) >> 12;

---

& with 0xF000 leaves the left 4 bits (0xF000 = 1111 0000 0000 0000) the way they are and clears the rest of them. The >> 12 moves everything to the right 12 places, so that you have numbers, from, say 0-8 instead of 8192-57344.

Quote:

---

Originally Posted by CommonTater

No...

You had to do all that looping to display the binary value underlying your variable.

Bitwise and ( & ) operates directly on that underlying value.

Do a little reading on the logic functions AND OR NEG and XOR for a better insight.

---

Ahh, that clears it up. Thanks!

Quote:

---

Originally Posted by anduril462

No. What I was getting at is this (note, these numbers are not correct for your example, you have to modify them):

Code:

---

command = (number & 0xF000) >> 12;

---

& with 0xF000 leaves the left 4 bits (0xF000 = 1111 0000 0000 0000) the way they are and clears the rest of them. The >> 12 moves everything to the right 12 places, so that you have numbers, from, say 0-8 instead of 8192-57344.

---

Thanks, that makes sense. :)

Sorry to bump this thread again, but I have ONE more question. When I am trying to identify the switch, something is happening incorrectly. I have:

Code:

---

switch = (number & 0xF000) >> 7;   
if(switch == 0)
        printf("    Switch: OFF");
else
        printf("    Switch: ON");

---

Where the switch is determined by the 8th digit from the right, 1 being on and 0 being off. But, when I enter 220, or 1101 1100, it is telling me that the switch is off. Any ideas as why?

You are applying the wrong value...

What hex value puts a 1 in only bit 7? (which is the 8th bit from the right)

Opens Win7 calculator, Programmer mode... click bit 7... = 0x80

Thus...

Code:

---

if (number & 0x80)
  puts("Eyyup, it's on!");
else
  puts("Where's that light switch, Martha?");

---

Quote:

---

Originally Posted by CommonTater

You are applying the wrong value...

What hex value puts a 1 in only bit 7? (which is the 8th bit from the right)

Opens Win7 calculator, Programmer mode... click bit 7... = 0x80

Thus...

Code:

---

if (number & 0x80)
  puts("Eyyup, it's on!");
else
  puts("Where's that light switch, Martha?");

---

---

So does that mean I am also finding the command wrong?

Code:

---

command = (number & 0xF000) >> 13;   
printf("    Command: %d\n", command);

---

The above works, but maybe what I need is a link to a thorough explanation of things such as 0xF and 0x80.

Quote:

---

Originally Posted by Alan Gott

So does that mean I am also finding the command wrong?

Code:

---

command = (number & 0xF000) >> 13;   
printf("    Command: %d\n", command);

---

---

Given that the command is supposed to be only the 3 msb (most significant bits) yes...

You probably want to AND with 0xE000 instead of ... 0xF000

However! ... since you are looking only for the three most significant bits, there's no need to strip lower bits with AND at all... You can just send them to the bit-bucket.

Code:

---

command = number >> 13:

---

will get the job done.


Something you will find very handy when bit twiddling is to take a second and write down the binary values for 0 to F ... then you have a visual reference that can show you what values you need.

Quote:

---

Originally Posted by CommonTater

Given that the command is supposed to be only the 3 msb (most significant bits) yes...

You probably want to AND with 0xE000 instead of ... 0xF000

However! ... since you are looking only for the three most significant bits, there's no need to strip lower bits with AND at all... You can just send them to the bit-bucket.

Code:

---

command = number >> 13:

---

will get the job done.


Something you will find very handy when bit twiddling is to take a second and write down the binary values for 0 to F ... then you have a visual reference that can show you what values you need.

---

Thanks a ton! I will definitely look into this more.