Simple array and arthmetics

[Evoprog]

Hi everyone !
I've started learning python a few months ago, but I'm stuck since a few weeks on a C script.

I've got an array :
char local_1 [4] : FF 5F 09 CF

and I have this code

if ((byte)(local_1[0] - 0x30U) < 10 {
   mac1 = (short) local_1[0] + -0x30;
}
else {
   if ((byte)(local_1[0] +0xbfU) < 6 {
   mac1 = (short) local_1[0] + -0x37;
}
      else {
         if ((byte)(local_1[0] +0x9fU) < 6 {
         mac1 = (short) local_1[0] + -0x57;
}
            else {
               mac1 = 0x10
   }
 }
}

I would tend to say first of all that local_1[0] = FF

The line that i don't understand is (byte)(FF-0x30) =?

Does the result refer to the number of bits in the result of the substraction?

FF-0x30 = CF in hex = 207 in decimal = 11001111 in bin

so the answer would equal 8 ?


Thanks very much for your precious help

[hamster_nz]

I haven't looked at your code deeply, what platform are you on?

On Intel, 'char' is a signed type, so 0xFF is -1.

On ARM 'char' is unsigned, so 0xFF is 255.

If you have all the compiler warnings turned on ("-Wall -pedantic -Wextra" options for GCC) it will point out these sorts of issues

[hamster_nz]

You can't have that code. THere are unbalanced brackets in the 'it' statements.

And " mac1 = 0x10" is missing the semicolon.

I've made a compellable program:

#include <stdio.h>
#include <stdint.h>
typedef uint8_t byte;

char local_1[4] = {0xFF, 0x5F, 0x09, 0xCF};


int main(int argc, char *argv[]) {
   char mac1;
   if ((byte)(local_1[0] - 0x30U) < 10) {
      mac1 = (short) local_1[0] + -0x30;
   } else if ((byte)(local_1[0] +0xbfU) < 6) {
      mac1 = (short) local_1[0] + -0x37;
   } else if ((byte)(local_1[0] +0x9fU) < 6) {
      mac1 = (short) local_1[0] + -0x57;
   } else {
      mac1 = 0x10;
   }
   printf("mac1 is %u\n",mac1);
   return 0;
}

Here's the compilation output with warnings on:

$ gcc -o help help.c -Wall -pedantic
help.c:4:20: warning: overflow in implicit constant conversion [-Woverflow]
 char local_1[4] = {0xFF, 0x5F, 0x09, 0xCF};
                    ^~~~
help.c:4:38: warning: overflow in implicit constant conversion [-Woverflow]
 char local_1[4] = {0xFF, 0x5F, 0x09, 0xCF};
                                      ^~~~

As a general rule of thumb, all the cool kids working on portable system-level code are using the types in stdint.h:

uint8_t - 8-bit unsigned
uint16_t - 16-bit unsigned
uint32_t - 32-bit unsigned
uint64_t - 64-bit unsigned

int8_t - 8-bit signed
int16_t - 16-bit signed
int32_t - 32-bit signed
int64_t - 64-bit signed

[Evoprog]

Hi, thanks very much for your help

I used ghidra to decompile an exe. this is only a small part of the code used as an example to understand.

Processor is motorola = ARM technology. So FF = 255.

I just wanted to be sure that arrays work like in python or other basic languages : [4] = [0,1,2,3]

and especially this simple maths result :

(byte)(0xFF - 0x30U) = ?

don't really know if i have to convert the answer etc?

thanks very much for your time

[hamster_nz]

(byte)(0xFF - 0x30U) = ?

(byte)(0xFF - 0x30U) = 0xCF.

If that is a + or - number depends on the definition of 'byte'.

[Evoprog]

I guess these are all positive, but not sure.

I was also guessing that this would equal 0xCF.
But when having a look at the code, first condition is that the answer <10 . Second and third conditions are <6.
I was trying to figure out if this "10" or "6" is a decimal or decimal from signed 2nd complement answer?

So in this cas 0xCF = 207 (decimal) or -49 (decimal from signed 2nd complement)


First case (decimal answer) seems strange bcause the second "if" condition is (byte)(local_1[0] + 0xbfu) <6

but 0xbfu by itself in decimal is 191, so an addition with a positive number can't ever be under 6 ....

thanks

[Evoprog]

Hi, any ideas about this equation ?

I'll try out all different cases mabye . This is a tricky one

thanks!