the calculation of inverse operation and power operation are not correct on the DSP TMDSDK6416-t

Hi,

Your query is not device specific and moving your post to code composer studio (CCS) forum to be better answered.

Thanks & regards,

Sivaraj K

Hi,

Unfortunately with that level of detail is almost impossible to know where the issue may be coming from. It could be one or a range of factors that can be originated from the library you are using (is it this one?), an invalid call to the functions you are using (a usage error), or even a bug on its code.

Being a third party library, we here have very limited power to figure out its internals. Also, it is also possible the failing functions may require some effort to change its code to be compatible with a C6000 DSP (the library seems to be ported only for specific architectures) - in other words, the code may not be platform independent.

After verifying all these details you will probably either find and fix the issue or have some specific topics related to how the device behaves given a specific instruction is done (which the experts at the device forum would know better) or how the compiler should have addressed a specific operation (which the experts at the compiler forum would know better).

In all cases, we here in the CCS forum will probably only be able to redirect this question to a more appropriate forum as soon as you are able to debug the internals of these functions you mentioned.

Hope this helps,

Rafael

 I would like to ask what is the best library that i could use on code composer V3 on TMDSDSK6416-T dsp kit 

my code are working on large numbers i did use the Big digit library 

http://www.di-mgt.com.au/bigdigits.html

but the results are not correct 

i did compare the same code on visual studio and it generate the correct results 

but on the code composer the results are wrong . 

i  did create a simple code that multiply a number then get the imverse of it 

but the results are also not correct 

i will past this simple code 

but meanwhile , i need your feedback about what is the best library to be used with the above device to work on large numbers calculations ?

here it is a simple code that  i am trying it 

#include <stdio.h>
#include "hello.h"

#include <string.h>

#include "bigdigits.h"
#include "bigdtypes.h"
#include "bigd.h"

#define BUFSIZE 30
#define FILEIO 1
struct PARMS str =
{
2934,
9432,
213,
9432,
&str
};

/*
* ======== main ========
*/
void main()
{

char *p_IN="3167";
char *q_IN="1583";
char *g_IN="44";
BIGD p, q, g;
BIGD Xa, Ya, Xb,Yb;

BIGD temp1 , temp2,temp3,temp4;
char charNo[1024];
unsigned short int No;

puts("hello world111111!\n");
p = bdNew();
q = bdNew();
g = bdNew();
Xa = bdNew();
Ya = bdNew();
bdConvFromDecimal(p,p_IN);
bdConvFromDecimal(q,q_IN);
bdConvFromDecimal(g,g_IN);
bdSetShort(Xa,79);
Ya = bdNew();
temp1=bdNew();
temp2=bdNew();
bdConvToDecimal(Xa,charNo,sizeof(charNo)); /// convert the value of X into Character
No=atoi(charNo);
bdPower(temp1,g,No);
bdPrintDecimal(" \n g ^xa= ",temp1, "\n");
bdModulo(temp2,temp1,p);
bdPrintDecimal(" \n temp2 = ",temp2, "\n");
bdModInv(Ya,temp1,p);
bdPrintDecimal("Alice's privarte Key :Xa= ", Xa, "\n");
bdPrintDecimal("Alice's public Key :Ya= (g^-Xa) mod p =", Ya, "\n\n");
printf("press any key to exist\n");
/* write a string to stdout */
puts("hello world!\n");
getchar();
getchar();

}

on micorosoft visual studio Ya has a value 

but when run it on dsp 

the Ya has value 0 or 1 

which is not correct 

i am sure that the results on the visual studio are correct 

why on DSP not working 

the library are in the link in the previous post 

Need help