loop for finding possible odd divisors

[JosAH]

is there a way I could define a data type that will fit all data under 171 undecillion ?

As I wrote before: you need a 'big integer' package; several are available, Google is your friend. About the definition of S(n): if it is what I think it is it's the sum of all proper divisors of n except n itself, e.g. S(6) == 1+2+3 = 6; S(28) == 1+2+4+7+14 == 28. When you see large primes you see perfect numbers around the corner (a perfect number equals the sum of its divisors, i.e. n == S(n)).

kind regards,

Jos

[scienceman88]

could a code be added to check those if so we would fit all needs for a mersenne prime tester and I know what else to change to find a new record faster.

[scienceman88]

found a full code for mersennes on rosettacode.org if it's true I can whip Gimps want to try ?

[scienceman88]

#include <math.h>
#include <stdio.h>
#include <limits.h>
#pragma precision=log10l(ULLONG_MAX)/2

typedef enum { FALSE=0, TRUE=1 } BOOL;

BOOL is_prime( int p ){
  if( p == 2 ) return TRUE;
  else if( p <= 1 || p % 2 == 0 ) return FALSE;
  else {
	BOOL prime = TRUE;
	const int to = sqrt(p);
	int i;
    for(i = 3; i <= to; i+=2)
	  if (!(prime = p % i))break;
	return prime;
  }
}

BOOL is_mersenne_prime( int p ){
  if( p == 2 ) return TRUE;
  else {
	const long long unsigned m_p = ( 1LLU << p ) - 1;
	long long unsigned s = 4;
	int i;
	for (i = 3; i <= p; i++){
	  s = (s * s - 2) % m_p;
    }
	return s == 0;
  }
}

int main(int argc, char **argv){

  const int upb = log2l(ULLONG_MAX)/2;
  int p;
  printf(" Mersenne primes:\n");
  for( p = 43112609; p <= upb; p += 1 ){
	if( is_prime(p) && is_mersenne_prime(p) ){
	  printf (" M%u",p);
	}
  }
  printf("\n");
}

is an edited version of the code on rosettacode.org it gives me these warning can you help me solve them ?


[BCC32 Warning] File2.c(36): W8065 Call to function 'log2l' with no prototype
[BCC32 Warning] File2.c(45): W8070 Function should return a value
[ILINK32 Error] Error: Unresolved external '_log2l' referenced from C:\USERS\RODDY\ DOCUMENTS\RAD STUDIO\PROJECTS \DEBUG\FILE2.OB J

[JosAH]

[BCC32 Warning] File2.c(36): W8065 Call to function 'log2l' with no prototype
[BCC32 Warning] File2.c(45): W8070 Function should return a value
[ILINK32 Error] Error: Unresolved external '_log2l' referenced from C:\USERS\RODDY\ DOCUMENTS\RAD STUDIO\PROJECTS \DEBUG\FILE2.OB J

There is no prototype of function log2l( ... ) in scope anywhere; check your manual for which header file includes that prototype; #include that header file.

Your function main( ... ) is supposed to return an int. You are not returning anything at the end of that function. Insert 'return 0;' there.

Those old systems want you to explicitly link in the math library; add the '-lm' flag on the linker's command line.

kind regards,

Jos

[scienceman88]

There is no prototype of function log2l( ... ) in scope anywhere; check your manual for which header file includes that prototype; #include that header file.

Your function main( ... ) is supposed to return an int. You are not returning anything at the end of that function. Insert 'return 0;' there.

Those old systems want you to explicitly link in the math library; add the '-lm' flag on the linker's command line.

kind regards,

Jos

lm flag ? god I suck at C
even I figured the return part
I go by the standard headers and i have no idea what one would have it maybe it's a C99 complex.h function.

[scienceman88]

Think if my search is right it's in math.h but I could be wrong http://www.codecogs.com/reference/c/...hp?alias=log2l

[scienceman88]

I didn't even know there was a search.h see I learn things too much so I forget most things I think I love learning though so I'll have to adapt

[JosAH]

lm flag ? god I suck at C
even I figured the return part
I go by the standard headers and i have no idea what one would have it maybe it's a C99 complex.h function.

I don't know what you mean by 'Im flag ?'. The log2l( ... ) function most likely is a Borland function so they must have a header file available for it and, more important, a library that contains the definition of it. It is not a C99 thing.

kind regards,

Jos

[scienceman88]

by what I've found it should be in math.h so I have no idea why it doesn't work I know what its supposed to do but I don't know how I can do it

[JosAH]

by what I've found it should be in math.h so I have no idea why it doesn't work I know what its supposed to do but I don't know how I can do it

Check your math.h file and see if you can find a prototype of that function in there.

kind regards,

Jos

[scienceman88]

Works with: gcc version 4.1.2 20070925 (Red Hat 4.1.2-27) Works with: C99 Compiler options: gcc -std=c99 -lm Lucas-Lehmer_test.c -o Lucas-Lehmer_test
#include <math.h>
#include <stdio.h>
#include <limits.h>
#pragma precision=log10 l(ULLONG_MAX)/2

typedef enum { FALSE=0, TRUE=1 } BOOL;

BOOL is_prime( int p ){
if( p == 2 ) return TRUE;
else if( p <= 1 || p % 2 == 0 ) return FALSE;
else {
BOOL prime = TRUE;
const int to = sqrt(p);
int i;
for(i = 3; i <= to; i+=2)
if (!(prime = p % i))break;
return prime;
}
}

BOOL is_mersenne_pri me( int p ){
if( p == 2 ) return TRUE;
else {
const long long unsigned m_p = ( 1LLU << p ) - 1;
long long unsigned s = 4;
int i;
for (i = 3; i <= p; i++){
s = (s * s - 2) % m_p;
}
return s == 0;
}
}

int main(int argc, char **argv){

const int upb = log2l(ULLONG_MA X)/2;
int p;
printf(" Mersenne primes:\n");
for( p = 2; p <= upb; p += 1 ){
if( is_prime(p) && is_mersenne_pri me(p) ){
printf (" M%u",p);
}
}
printf("\n");
}
Output:
Mersenne primes:
M2 M3 M5 M7 M13 M17 M19 M31

[JosAH]

Works with: gcc version 4.1.2 20070925 (Red Hat 4.1.2-27) Works with: C99 Compiler options: gcc -std=c99 -lm Lucas-Lehmer_test.c -o Lucas-Lehmer_test

[ ... ]

Output:
Mersenne primes:
M2 M3 M5 M7 M13 M17 M19 M31

So it works for small values of p; now go for the really big ones. btw, please use those code tags around your code for readability reasons.

kind regards,

Jos

[scienceman88]

I can't get it to work for any

[JosAH]

I can't get it to work for any

You have to be more exact: first you write that the output happened to be correct and showed it to us and now you're writing that it doesn't work at all. I you want help you have to help us understand you. btw, the probability of a thread ending successfully is inversely proportional to the length of the thread.

kind regards,

Jos