Rotate function not working for uint3_t?

Projects Programming
1

When I wrtie a functionn to rotate uint16_t it works as expected

#define INT_BITS 16
/* rotate n by d bits*/
int leftRotate(uint16_t n, uint16_t d)
{
    return (n << d) | (n >> (INT_BITS - d));
}

But when I do the same for uint32_t

#define INT_BITS_32  32

uint32_t leftRotate_uint32(uint32_t n, uint32_t d)
{
    return (n << d) | (n >> (INT_BITS_32 - d));
}

This just results in a left shift?

Any clues anyone?

Repeated rotation just sift the contents out to the left

image
image176×420 1.24 KB

2

Perhaps begin by printing out the values of the partial expressions to see what's happening.

Edit: and by the way, where is n defined in leftRotate_uint32? You use it but I don't see a definition for it.

3

what microcontroller are you using?

4

Have edited OP accoringly

5

Nano

6

INT_BITS_32 is not the same as IN_BITS_32

can you post the real code ?

1 Like
7

Nano V3 (ATMEGA328)
Nano Every
Nano 33 IoT
Nano 33 BLE
Nano 33 BLE Sense
Nano RP2040 Connect
Nano 33 BLE Sense Rev 2
Nano ESP32
Nano R4

It makes a difference

8

Typo in the OP that isnt in the actual code will fix now.

9

ATMega328P

10

can you post your full code printing the details

11

Worth a try:

#define INT_BITS_32  32UL

uint32_t leftRotate_uint32(uint32_t n, uint32_t d)
{
    return (n << d) | (n >> (INT_BITS_32 - d));
}

Do you have any other Arduino, with 32-bit chip? What happens with your code on that board?

12

My compiler threw a fit when I added that, nope just a couple of nanos

13

Nope. No emotional responses. Maybe error or warning messages. You could post them (use code tags!).

14

D:\My Pics\Pacman\pac_code\pac_code.ino: In function 'uint32_t leftRotate_uint32(uint32_t, uint32_t)':
D:\My Pics\Pacman\pac_code\pac_code.ino:7:26: error: expected ')' before numeric constant
#define INT_BITS_32 32 32UL
^
D:\My Pics\Pacman\pac_code\pac_code.ino:265:30: note: in expansion of macro 'INT_BITS_32'
return (n << d) | (n >> (INT_BITS_32 - d));
^~~~~~~~~~~
D:\My Pics\Pacman\pac_code\pac_code.ino:265:47: error: expected ')' before ';' token
return (n << d) | (n >> (INT_BITS_32 - d));
^
D:\My Pics\Pacman\pac_code\pac_code.ino: In function 'uint32_t rightRotate_int32(uint32_t, uint32_t)':
D:\My Pics\Pacman\pac_code\pac_code.ino:7:26: error: expected ')' before numeric constant
#define INT_BITS_32 32 32UL
^
D:\My Pics\Pacman\pac_code\pac_code.ino:272:30: note: in expansion of macro 'INT_BITS_32'
return (n >> d) | (n << (INT_BITS_32 - d));
^~~~~~~~~~~
D:\My Pics\Pacman\pac_code\pac_code.ino:272:47: error: expected ')' before ';' token
return (n >> d) | (n << (INT_BITS_32 - d));
^

exit status 1

Compilation error: expected ')' before numeric constant

15

try

#define INT_BITS_32 32
uint32_t leftRotate_uint32(uint32_t n, uint32_t d) {
  return (n << d) | (n >> (INT_BITS_32 - d));
}

void printBinary32(uint32_t value) {
  for (int8_t i = 31; i >= 0; i--) Serial.print(bitRead(value, i));
  Serial.println();
}

void setup() {
  Serial.begin(115200);
  uint32_t value = 0b1111111111111111;

  for (uint32_t i = 0; i < INT_BITS_32; i++)   {
    uint32_t rotated = leftRotate_uint32(value, i);
    Serial.print("Rotate by ");
    if (i < 10) Serial.write(' ');
    Serial.print(i);
    Serial.print(": ");
    printBinary32(rotated);
  }
}

void loop() {}

you should see

Rotate by  0: 00000000000000001111111111111111
Rotate by  1: 00000000000000011111111111111110
Rotate by  2: 00000000000000111111111111111100
Rotate by  3: 00000000000001111111111111111000
Rotate by  4: 00000000000011111111111111110000
Rotate by  5: 00000000000111111111111111100000
Rotate by  6: 00000000001111111111111111000000
Rotate by  7: 00000000011111111111111110000000
Rotate by  8: 00000000111111111111111100000000
Rotate by  9: 00000001111111111111111000000000
Rotate by 10: 00000011111111111111110000000000
Rotate by 11: 00000111111111111111100000000000
Rotate by 12: 00001111111111111111000000000000
Rotate by 13: 00011111111111111110000000000000
Rotate by 14: 00111111111111111100000000000000
Rotate by 15: 01111111111111111000000000000000
Rotate by 16: 11111111111111110000000000000000
Rotate by 17: 11111111111111100000000000000001
Rotate by 18: 11111111111111000000000000000011
Rotate by 19: 11111111111110000000000000000111
Rotate by 20: 11111111111100000000000000001111
Rotate by 21: 11111111111000000000000000011111
Rotate by 22: 11111111110000000000000000111111
Rotate by 23: 11111111100000000000000001111111
Rotate by 24: 11111111000000000000000011111111
Rotate by 25: 11111110000000000000000111111111
Rotate by 26: 11111100000000000000001111111111
Rotate by 27: 11111000000000000000011111111111
Rotate by 28: 11110000000000000000111111111111
Rotate by 29: 11100000000000000001111111111111
Rotate by 30: 11000000000000000011111111111111
Rotate by 31: 10000000000000000111111111111111
16

Looks like your code is using 16-bit integer for printing, are you actually sending a 32-bit value to the function?

17

Yes I get the exact same output

18

so all works :slight_smile:

19

Ok, be like that, no more help from me.

20

Of course that causes warning, that replaces "INT_BITS_32" with "32 32UL", the compiler does not like the resulting two integers where it expects only one.