Thread: how to understand '|=' and '<<'

Hi all, I am a very beginner of C. There is a code which I cannot purely understand how does the MirrorBytes() function run.

there is the code:

Code:

function1(unsigned char *inBlock, unsigned char *outBlock,)
{
unsigned int left = MirrorBytes (((unsigned int *)inBlock)[0]);
unsigned int right = MirrorBytes (((unsigned int *)inBlock)[1]);
...
...
}

unsigned int MirrorBytes (unsigned int x)
{
unsigned int n = (unsigned char) x;
n<<=8; n|=(unsigned char) (x >>8);
n<<=8; n|=(unsigned char) (x>>16);
return (n<<8) | (unsigned char) (x>>24);
}

For example: if the value of the inBlock is '11010111', how does the MirrorBytes() function run?
if the value of the inBlock is '1101011101000110', how does the MirrorBytes() function run?

Thank you very much!!

Originally Posted by xnigel

Hi all, I am a very beginner of C. There is a code which I cannot purely understand how does the MirrorBytes() function run.

there is the code:

Code:

function1(unsigned char *inBlock, unsigned char *outBlock,)
{
unsigned int left = MirrorBytes (((unsigned int *)inBlock)[0]);
unsigned int right = MirrorBytes (((unsigned int *)inBlock)[1]);
...
...
}

unsigned int MirrorBytes (unsigned int x)
{
unsigned int n = (unsigned char) x;
n<<=8; n|=(unsigned char) (x >>8);
n<<=8; n|=(unsigned char) (x>>16);
return (n<<8) | (unsigned char) (x>>24);
}

For example: if the value of the inBlock is '11010111', how does the MirrorBytes() function run?
if the value of the inBlock is '1101011101000110', how does the MirrorBytes() function run?

Thank you very much!!

This code assumes a 32 bit, (4 byte) unsigned int. The function will reverse the order of the bytes.

If the original byte order is 1234 then the result byte order would be 4321.

It basically converts Bid-Endian, into Little-Endin, and vice-versa

Hi Rstanley and Matticus,

Thank both of you I think I have got something. Take a example for good practice: x = 0xACDB4213

Code:

firstly, if do n<<=8; n will be 0xDB421300, then n=n|(x>>8). That is n = 0xDB421300 | 0x00ACDB42. n = 0xDBEEDB42
secondly, do another n<<=8; n will be 0xEEDB4200, then n = 0xEEDB4200 | 0x0000ACDB. n = 0xEEDBEEDB
finally, n is 0xDBEEDB00, n = 0xDBEEDB00 | 0x 000000AC. n = 0xDBEEDBAC

According to your answer, the 'real' result should be 0x1342CDAB. So there must be somewhere I did wrong. Could you please correct me?

Originally Posted by xnigel

Hi Rstanley and Matticus,

Thank both of you I think I have got something. Take a example for good practice: x = 0xACDB4213

Code:

firstly, if do n<<=8; n will be 0xDB421300, then n=n|(x>>8). That is n = 0xDB421300 | 0x00ACDB42. n = 0xDBEEDB42
secondly, do another n<<=8; n will be 0xEEDB4200, then n = 0xEEDB4200 | 0x0000ACDB. n = 0xEEDBEEDB
finally, n is 0xDBEEDB00, n = 0xDBEEDB00 | 0x 000000AC. n = 0xDBEEDBAC

According to your answer, the 'real' result should be 0x1342CDAB. So there must be somewhere I did wrong. Could you please correct me?

You will need to post the full CURRENT code so we can examine.

Thank you.

Hi all, this afternoon i wrote a code and run in the Geany program though i do not have the full code. I just want work out why the result i calculated was not correct. The code written in the program is:

Code:

#include <stdio.h>
unsigned int n;
unsigned int m, m1, m2, m3, m4, m5, m6;
unsigned int z1, z2, z3, z4, z5, z6;

unsigned int MB32(unsigned int x)
{
	n = (unsigned char) x; // get last char of x, n == 'cd'.
	m = (unsigned char) x;
	
	n <<= 8;	// n == 'cd00'
	
	m1 = n;		// 
	z1 = x;		// z1 == x == '12 34 ab cd'
	
	n |= (unsigned char) (x >> 8);
	
	m2 = n;		//display the 1st result of n;
	z2 = x;		//display the 1st result of x;
	
	//=================================================
	
	n <<= 8;
	
	m3 = n;
	z3 = x;
		
	n |= (unsigned char) (x >> 16);
	
	m4 = n;
	z4 = x;
	
	//=================================================

	//return (n << 8) | (unsigned char) (x >> 24);
	n <<= 8;
	
	m5 = n;
	z5 = x;
	
	n |= (unsigned char) (x >> 24);
	
	m6 = n;
	z6 = x;
	
	return n;
}

int main()
{
	
	//int BlockIn = 0xABCD1234;
	int BlockIn = 0x1234ABCD;
	int BlockOut = MB32(BlockIn);
	printf ("BlockIn is: %x \n", BlockIn);
	printf ("BlockOut is: %x \n", BlockOut);
	printf ("m  = %x \n", m);
	printf ("m1 = %x \n", m1);
	printf ("z1 = %x \n \n", z1);
	printf ("m2 = %x \n", m2);
	printf ("z2 = %x \n \n", z2);
	printf ("m3 = %x \n", m3);
	printf ("z3 = %x \n \n", z3);
	printf ("m4 = %x \n", m4);
	printf ("z4 = %x \n \n", z4);
	printf ("m5 = %x \n", m5);
	printf ("z5 = %x \n \n", z5);
	printf ("m6 = %x \n", m6);
	printf ("z6 = %x \n \n", z6);

	return 0;
}

I think only the reason i got the wrong answer is because of the definition of 'unsigned char x', 'unsigned int x', and 'unsigned int n'. Some comment were added. I wish this can be used for other learner as well. This could could be run under 64bit Ubuntu.

Thank you all guys!

I think only the reason i got the wrong answer is because of the definition of 'unsigned char x', 'unsigned int x', and 'unsigned int n'.

What "wrong answer" did you get?

When I run your code with "1234abcd" as the input, the output is, "cdab3412".

Is this not the correct output? If not, what output did you expect?

BTW:

Code:

int main()

should be:

int main(void)