CRC ALGORITHM FROM C TO LABVIEW

[jack1]

I have been given a CRC algorithim written in C for checking the packets of data coming down a serial link.

Have had a look on the forum and found examples of code suitable for CRC checking. Problem is that when I try the various Vi examples none give me the same result as that given by the one suggested by using C.

Pehaps I am missing something but from my dim and distant past I always thought that a fixed polynomial equation was used on each of the packets of data to arrive at a CRC value? If that assumption is correct then could that be why I do not get the same CRC value as the one given if implemented in the C example?

I have tried to execute the C code example but it failed on the compiler (it is just extracts of core code apparently) and I have hardly any knowledge of C so struggling to say the least.

[rsw]

Hi Jack,

This isn't something I've seen before but I would be more than happy to look into the issue with you, do you have the C code example you mention or the CRC type you are attempting under LabVIEW? Also which LabVIEW VI have you been using that gives you a different CRC result so I can take a look?

Thanks,

[jack1]

The link for CRC Vi is:- http://zone.ni.com/devzone/cda/epd/p/id/1568

and the C code I have been given for guidance is:-

First block of code:-

unsigned crc_table [256]

unsigned crc(unsigned char*buffer, unsigned count)

{

   unsigned i;

   unsigned c = 0;

   for (i = 0; i<count; i++)

{

  c = (c >> 😎 ^ crc_table[buffer[i] ^ (unsigned char)c];

}

return c;

Second block of code:-

void init_crc_table()

{

  unsigned z,i;

  for (i = 0, z = 0; i < 256; i++)

{

        z = i ^ (i << 4);

        crc_table [i]  (z << 😎 ^ (z << 3) ^ (z >> 4);

}

}

It seems that a table is being created to run the packets of data through but really no idea.

The data given to confirm correct operation is a 10 Byte sequence:-

0xb0, 0x3b, 0x4f, 0xfa, 0x00, 0x55, 0x89, 0x4c, 0x01, 0x35   result should be CRC value of 0xf268 (MSB first)

Thanks Jack

[rsw]

Hi Jack,

Just so I can understand what is your core aim from this, to create a CRC in C that is the same as in LabVIEW, vice versa or doesn't it matter just so long as they match for C and LabVIEW with regards to the outcome? Looking at the C code you provided it looks like just a few of the functions and doesn't appear to match the LabVIEW one at all, for example the LabVIEW application has initial CRC values set, and then runs 8 times every interation of the packet length where as the C code you have provided runs for the value of "count" and then runs for 256 times per iteration of count. I know you mentined you were unable to get the code to execute but do you have any CRC-16 C code that runs (with source) to verify the behaviour?

Once we have one that works as you require it's just a case of transferring the logic to the other language.

Kind Regards,

[jack1]

Hi Rob

At the moment there is a main control program that I have written in Labview 8.6 to run/monitor a combustion engine.

I now need to access the engines ECU to obtain temps/pressures from it and this information is available in serial format.

The serial data has a CRC check tacked onto the end of it for conformation that the data is valid.

The information given on the structure of that CRC routine is the one I detailed earlier.

So basically I need to be able to implement in Labview a CRC routine that can perform the same operations on the serial data packets so I can integrate the CRC Labview code into my main program.

Thanks Jack

[rsw]

Hi Jack,

Thanks for the additional information, as the serial CRC is already created (from what I understand) do you know exactly which type it uses as CRC-16 has many variants?

Thanks,

[jack1]

Hi Rob

That's the main problem. I know of the wide variation in CRC Polynomials and have tried quite a few of them without success. When the question was asked all I was told was that the C Code was written to carry out this function and no polynomial was available.

Could the C code given generate a CRC result that has no relationship to the standard CRC-16?

Thanks

jack

[rsw]

Hi Jack,

Is it that they don't wish to give you the polynomial/know what it is or does the CRC actually not use a polynomial? As you can see from this link: http://en.wikipedia.org/wiki/Cyclic_redundancy_check#Commonly_used_and_standardized_CRCs the majority of CRC's have a polynomial associated with them, one that doesn't is Fletcher's Checksum however looking at the C Code you included this doesn't appear to be it.

Without knowing the type of CRC it is (as it may be one of there own) then this might be quite a challenge!

[jack1]

Hi Rob

Unsure if they do not know or as you say are unwilling to let me know. That's the same link where I found all of (or most) of the CRC codes from originally when I thought this was going to be a 5 minute job!

I will contact them today and see if I can get more detail. Its just that I thought there was something obvious in the C code given that would have revealed all, glad there was not in one way because I could not see anything and makes me feel better!!

Will let you know how I get on,

Thanks Jack

[rsw]

Hi Jack, No problems, I know what you mean it's always good when someone else doesn't solve something you've been working on right away (bar it taking longer to solve the issue of course!) Let me know how it goes and if I can help further. Cheers,