We Keep Coding

Crc64 LSB based reverse implementation

I have tried to implement LSB based reversal function. The code is in C#. There's a bug in reversal of the checksum. The function can only reverse the Crc64 checksum when it was computed with only 8 bytes original data. When I try to reverse the checksum through FixChecksum method, the middle 6 bytes are reversed, but the first & the last byte are reversed corrupted. Please inform what is wrong, what's needed to be implemented or fixed. I will appreciate any solution.
[UPDATED]
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using MYC;
using primaryMethodsCS_2;
using System.Globalization;
using UnitsFramework.Numerics;
using UnitsFramework;
namespace MYC
{
public class Crc64_LSB_Reverse
{
public const UInt64 POLY64REV = 0xD800000000000000; //0xffffffffffffff00;
public static ulong TOPBIT_MASK = 0x8000000000000000;
public static ulong LOWBIT_MASK = 0x0000000000000001;
public const ulong startxor = 0; //0xffffffffffffffff;
public const ulong FinalXor = 0; // 0xffffffffffffffff;
public UInt64[] CRCTable;
public UInt64[] revCRCTable;
public UInt64 crc = 0;
public Crc64_LSB_Reverse(UInt64 POLY = POLY64REV)
{
List<ulong> listforward = new List<ulong>();
List<ulong> listreverse = new List<ulong>();
for (int i = 0; i <= 255; i++)
{
List<ulong> forward = generateCrcTableConstants(new List<ulong>() { (UInt64)i }, POLY);
List<ulong> reverse = generateRevCrcTableConstants(new List<ulong>() { (UInt64)i }, POLY);
listforward.AddRange(forward);
listreverse.AddRange(reverse);
}
this.CRCTable = listforward.ToArray();
this.revCRCTable = listreverse.ToArray();
return;
}
public static List<UInt64> generateCrcTableConstants(List<UInt64> initialValues, UInt64 POLY)
{
List<UInt64> list = new List<ulong>();
for (int thisValue = 0; thisValue < initialValues.Count; thisValue++)
{
UInt64 currentValue = initialValues[thisValue];
UInt64 initialValue = currentValue;
currentValue <<= 56; // is valid for MSB forward table creation
// MSB based forward table implementation.
for (byte bit = 0; bit < 8; bit++)
{
if ((currentValue & TOPBIT_MASK) != 0)
{
//currentValue <<= 1;
//currentValue ^= CrcFramework.Reflect64(POLY);
currentValue = (currentValue << 1) ^ ((0 - (currentValue >> 63)) & POLY); // fwd
}
else
{
currentValue <<= 1;
}
}
list.Add(currentValue);
}
return list;
}
public static List<UInt64> generateRevCrcTableConstants(List<UInt64> initialValues, UInt64 POLY)
{
List<UInt64> list = new List<ulong>();
for (int thisValue = 0; thisValue < initialValues.Count; thisValue++)
{
UInt64 initialValue = initialValues[thisValue];
UInt64 currentValue = initialValues[thisValue];
// LSB based reverse table implementation for MSB based forward table function.
for (byte bit = 0; bit < 8; bit++)
{
if ((currentValue & LOWBIT_MASK) != 0)
{
//currentValue ^= POLY; // CrcFramework.Reflect64(POLY); //POLY;
currentValue = (currentValue >> 1) ^ ((0 - (currentValue & 1)) & POLY); // rvs
//currentValue >>= 1;
//currentValue |= 1; // TOPBIT_MASK;
}
else
{
currentValue >>= 1;
}
}
list.Add(currentValue);
}
return list;
}
public ulong Compute_LSB(byte[] bytes, bool reset = true)
{
if (reset) this.crc = startxor;
foreach (byte b in bytes)
{
byte curByte = b;
/* update the LSB of crc value with next input byte */
crc = (ulong)(crc ^ (ulong)(curByte));
/* this byte value is the index into the lookup table */
byte pos = (byte)(crc & 0xFF); // tushar: original 12-September-2019-1: & 0xFF);
/* shift out this index */
crc = (ulong)(crc >> 8);
/* XOR-in remainder from lookup table using the calculated index */
crc = (ulong)(crc ^ (ulong)CRCTable[pos]);
/* shorter:
byte pos = (byte)((crc ^ curByte) & 0xFF);
crc = (ulong)((crc >> 8) ^ (ulong)(crcTable[pos]));
*/
}
return (ulong)(crc ^ FinalXor);
}
public ulong Compute_MSB(byte[] bytes, bool reset = true)
{
if (reset) this.crc = startxor;
foreach (byte b in bytes)
{
byte curByte = b;
/* update the MSB of crc value with next input byte */
crc = (ulong)(crc ^ (ulong)((ulong)curByte << 56));
/* this MSB byte value is the index into the lookup table */
byte pos = (byte)(crc >> 56);
/* shift out this index */
crc = (ulong)(crc << 8);
/* XOR-in remainder from lookup table using the calculated index */
crc = (ulong)(crc ^ (ulong)CRCTable[pos]);
/* shorter:
byte pos = (byte)((crc ^ (curByte << 56)) >> 56);
crc = (uint)((crc << 8) ^ (ulong)(crcTable[pos]));
*/
}
return (ulong)(crc ^ FinalXor);
}
public UInt64 FixChecksum(byte[] bytes, Int64 length, Int64 fixpos, UInt64 wantcrc)
{
if (fixpos + 8 > length) return 0;
UInt64 crc = startxor;
for (Int64 i = 0; i < fixpos; i++)
{
crc = (crc >> 8) ^ CRCTable[(crc ^ bytes[i]) & 0xff];
}
Array.Copy(BitConverter.GetBytes(crc), 0, bytes, fixpos, 8);
List<UInt64> list = new List<UInt64>();
crc = wantcrc ^ startxor;
for (Int64 i = length - 1; i >= fixpos; i--)
{
UInt64 param0 = (UInt64)(crc >> 56);
list.Add(param0);
crc = (crc << 8) ^ revCRCTable[param0] ^ bytes[i]; //
}
Array.Copy(BitConverter.GetBytes(crc), 0, bytes, fixpos, 8);
return crc;
}
}
}

The solution to this problem involves reverse cycling a CRC.
If using tables, for CRC generation (forward cycling), a left shifting CRC uses the left byte of the CRC to index a table, and a right shifting CRC uses the right byte of the CRC to index a table. For CRC reverse cycling, a left shifting CRC uses the right byte of the CRC to index a table, and a right shifting CRC uses the left byte of the CRC to index a table.
Example code below for both left shifting and right shifting CRCs. Two additional tables are used for reverse cycling CRCs. The CRC functions take in a data length parameter, in case the buffer has additional space to store the CRC after generating it.
To simplify the reverse cycling functions, a CRC is generated and xor'ed with the wanted CRC, that can be used with an imaginary buffer of zeroes, to generate the 8 bytes of data needed to produce the xor'ed CRC. The 8 bytes of data are then xor'ed to the original buffer. The xor's cancel out the CRC initial and xor out values, so the reverse cycling functions don't have to take those values into account.
Note that the CRC bytes can go anywhere in a message (although they are usually at the end of a message). For example a 24 byte message consisting of 8 bytes of data, 8 bytes of CRC, 8 bytes of data, which can be created using crc64fw(0ul, bfr, 24, 8); or crc64rw(0ul, bfr, 24, 8);.
namespace crc64
{
public class crc64
{
public const ulong poly64f = 0x000000000000001bul;
public const ulong poly64r = 0xd800000000000000ul;
public const ulong crcin = 0x0000000000000000ul; // initial value
public const ulong xorot = 0x0000000000000000ul; // xorout
public static ulong[] crctblf; // fwd tbl
public static ulong[] crctblg; // fwd tbl reverse cycle
public static ulong[] crctblr; // ref tbl
public static ulong[] crctbls; // ref tbl reverse cycle
// generate tables
public static void gentbls()
{
ulong crc;
byte b;
b = 0;
do
{
crc = ((ulong)b)<<56;
for(int i = 0; i < 8; i++)
crc = (crc<<1)^((0-(crc>>63))&poly64f);
crctblf[b] = crc;
b++;
}while (b != 0);
do
{
crc = ((ulong)b) << 0;
for (int i = 0; i < 8; i++)
crc = (crc<<63)^((crc^((0-(crc&1))&poly64f))>>1);
crctblg[b] = crc;
b++;
} while (b != 0);
do
{
crc = ((ulong)b)<<0;
for(int i = 0; i < 8; i++)
crc = (crc>>1)^((0-(crc&1))&poly64r);
crctblr[b] = crc;
b++;
}while (b != 0);
do
{
crc = ((ulong)b) << 56;
for (int i = 0; i < 8; i++)
crc = (crc>>63)^((crc^((0-(crc>>63))&poly64r))<<1);
crctbls[b] = crc;
b++;
} while (b != 0);
}
// generate forward crc
public static ulong crc64f(byte[] bfr, int len)
{
ulong crc = crcin;
for(int i = 0; i < len; i++)
crc = (crc<<8)^(crctblf[(crc>>56)^bfr[i]]);
return (crc^xorot);
}
// append forward crc
public static void crc64fa(ulong crc, byte[] bfr, int pos)
{
bfr[pos+0] = (byte)(crc>>56);
bfr[pos+1] = (byte)(crc>>48);
bfr[pos+2] = (byte)(crc>>40);
bfr[pos+3] = (byte)(crc>>32);
bfr[pos+4] = (byte)(crc>>24);
bfr[pos+5] = (byte)(crc>>16);
bfr[pos+6] = (byte)(crc>> 8);
bfr[pos+7] = (byte)(crc>> 0);
}
// "fix" bfr to generate wanted forward crc
public static void crc64fw(ulong crc, byte[] bfr, int len, int pos)
{
crc ^= crc64f(bfr, len);
for(int i = pos; i < len; i++)
crc = (crc>>8)^(crctblg[crc&0xff]);
bfr[pos+0] ^= (byte)(crc>>56);
bfr[pos+1] ^= (byte)(crc>>48);
bfr[pos+2] ^= (byte)(crc>>40);
bfr[pos+3] ^= (byte)(crc>>32);
bfr[pos+4] ^= (byte)(crc>>24);
bfr[pos+5] ^= (byte)(crc>>16);
bfr[pos+6] ^= (byte)(crc>> 8);
bfr[pos+7] ^= (byte)(crc>> 0);
}
// generate reflected crc
public static ulong crc64r(byte[] bfr, int len)
{
ulong crc = crcin;
for(int i = 0; i < len; i++)
crc = (crc>>8)^(crctblr[(crc&0xff)^bfr[i]]);
return (crc^xorot);
}
// append reflected crc
public static void crc64ra(ulong crc, byte[] bfr, int pos)
{
bfr[pos+0] = (byte)(crc>> 0);
bfr[pos+1] = (byte)(crc>> 8);
bfr[pos+2] = (byte)(crc>>16);
bfr[pos+3] = (byte)(crc>>24);
bfr[pos+4] = (byte)(crc>>32);
bfr[pos+5] = (byte)(crc>>40);
bfr[pos+6] = (byte)(crc>>48);
bfr[pos+7] = (byte)(crc>>56);
}
// "fix" bfr to generate wanted reflected crc
public static void crc64rw(ulong crc, byte[] bfr, int len, int pos)
{
crc ^= crc64r(bfr, len);
for (int i = pos; i < len; i++)
crc = (crc<<8)^(crctbls[crc>>56]);
bfr[pos+0] ^= (byte)(crc>> 0);
bfr[pos+1] ^= (byte)(crc>> 8);
bfr[pos+2] ^= (byte)(crc>>16);
bfr[pos+3] ^= (byte)(crc>>24);
bfr[pos+4] ^= (byte)(crc>>32);
bfr[pos+5] ^= (byte)(crc>>40);
bfr[pos+6] ^= (byte)(crc>>48);
bfr[pos+7] ^= (byte)(crc>>56);
}
static void Main(string[] args)
{
crctblf = new ulong[256];
crctblg = new ulong[256];
crctblr = new ulong[256];
crctbls = new ulong[256];
byte[] bfr = {0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37, // data (16 bytes)
0x38,0x39,0x61,0x62,0x63,0x64,0x65,0x66,
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00}; // space for crc
ulong crcf, crcr, crcw, crcg, crcs;
int dl = bfr.Length-8; // length of data
int bl = bfr.Length; // length of bfr
gentbls();
crcf = crc64f(bfr, dl); // forward crc
crc64fa(crcf, bfr, dl); // append crc
crcw = crc64f(bfr, bl); // crcw == 0
crcr = crc64r(bfr, dl); // reflected crc
crc64ra(crcr, bfr, dl); // append crc
crcw = crc64r(bfr, bl); // crcw == 0
Console.WriteLine(crcf.ToString("x16") + " " + crcr.ToString("x16"));
crcw = 0x0001020304050607ul; // wanted crc
crc64fw(crcw, bfr, dl, 8); // "fix" for forward
crcg = crc64f(bfr, dl); // crcg == crcw
crc64fw(crcf, bfr, dl, 8); // undo "fix" for forward (restore bfr)
crc64rw(crcw, bfr, dl, 8); // "fix" for reflected
crcs = crc64r(bfr, dl); // crcs == crcw
Console.WriteLine(crcg.ToString("x16") + " " + crcs.ToString("x16"));
}
}
}

How to generate Crc-64 table having all negative integer Constants and checksum?

I have some example code for Crc-64 Table generator. I tried to check the unsigned integer's sign and discovered that it generates mixed table Constants both negative & positive integers. Same for the Crc-64 Checksum, it may be negative or positive. Is it possible to implement a modified Crc-64 Table generator that should produce all negative signed Constants and also the Checksum? Or otherwise all positive signed Constants and Checksum. Kindly help me with some information and example implementation.
Here is the example code for the Crc-64 Table generator:
public class Crc64
{
public const UInt64 POLYNOMIAL = 0xD800000000000000;
public UInt64[] CRCTable;
public Crc64()
{
this.CRCTable = new ulong[256];
for (int i = 0; i <= 255; i++)
{
int j;
UInt64 part = (UInt64)i;
for (j = 0; j < 8; j++)
{
if ((part & 1) != 0)
part = (part >> 1) ^ POLYNOMIAL;
else
part >>= 1;
}
CRCTable[i] = part;
}
}
}
UPDATE: Kindly inform, is this implementation correct as per my question:
public static List<UInt64> generateCrcTable(UInt64 POLY)
{
const ulong TOPBIT_MASK = 0x8000000000000000;
const ulong LOWBIT_MASK = 0x0000000000000001;
List<UInt64> list = new List<UInt64>();
for (int i = 0; i <= 255; i++)
{
UInt64 part = (UInt64)(i); // << 56;
for (byte bit = 0; bit < 63; bit++)
{
if ((part & LOWBIT_MASK) != 0) // 0x8000000000000000) != 0) //1) != 0) // 0x8000000000000000) != 0)
{
part >>= 1;
part ^= POLY;
}
else
{
part >>= 1;
//currentValue |= 0x8000000000000000;
}
}
part |= TOPBIT_MASK; // 0x8000000000000000;
list.Add(part);
}
return list;
}

I got code to work. No table. crc2 is always zero. I run CRC twice. Once to create CRC then add CRC to original data. Then run again to validate.
if you only want 63 bits then change the shift from 56 to 55. Then use mask starting with 0x4 instead of 0x8. And then on the return AND with 0x7FFFFFFFFFFFFFFF
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
namespace ConsoleApplication1
{
class Program
{
static void Main(string[] args)
{
Random rand = new Random();
List<byte> data = Enumerable.Range(0, 1000).Select(x => (byte)rand.Next(0,256)).ToList();
ulong crc = Crc64.ComputeCrc64(data);
List<byte> results = new List<byte>();
for (int i = 7; i >= 0; i--)
{
results.Add((byte)((crc >> (8 * i)) & 0xFF));
}
data.AddRange(results);
ulong crc2 = Crc64.ComputeCrc64(data);
}
}
public class Crc64
{
const ulong POLYNOMIAL = 0xD800000000000000;
public static ulong ComputeCrc64(List<byte> data)
{
ulong crc = 0; /* CRC value is 64bit */
foreach (byte b in data)
{
crc ^= (ulong)b << 56; /* move byte into MSB of 63bit CRC */
for (int i = 0; i < 8; i++)
{
if ((crc & 0x8000000000000000) != 0) /* test for MSB = bit 63 */
{
crc = (ulong)((crc << 1) ^ POLYNOMIAL);
}
else
{
crc <<= 1;
}
}
}
return crc;
}
}
}

Convert CRC calculation from C# to C

I am currently working on a serial monitor, to ensure some data integrity I am trying to implement a CRC8 checksum, below is the calculation i do on any messages before i send them.
public byte Checksum(params byte[] val)
{
if (val == null)
throw new ArgumentNullException("val");
byte c = 0;
foreach (byte b in val)
{
c = table[c ^ b];
}
return c;
}
I generate a table using 0xD8:
public byte[] GenerateTable(CRC8_POLY polynomial)
{
byte[] csTable = new byte[256];
for (int i = 0; i < 256; ++i)
{
int curr = i;
for (int j = 0; j < 8; ++j)
{
if ((curr & 0x80) != 0)
{
curr = (curr << 1) ^ (int)polynomial;
}
else
{
curr <<= 1;
}
}
csTable[i] = (byte)curr;
}
return csTable;
}
This is a code i have used for testing the setup:
private void btnSend_Click(object sender, EventArgs e)
{
ProtoFrame rxFrame = new ProtoFrame();
if (cboParam.Text == "test")
{
rxFrame.Start = 0x73;
rxFrame.Size = 9;
rxFrame.Command = 01;
rxFrame.Unused = 0;
rxFrame.ParamId = 0x0100;
rxFrame.Param = 8000;
}
byte[] rxBuffer = getBytes(rxFrame); //call to byte array formatter
rxBuffer[rxBuffer.Length-1] = Checksum(rxBuffer); //append crc at end of array
ComPort.Write(rxBuffer, 0, rxBuffer.Length);
}
static byte[] getBytes(object str) //input struct
{
int size = Marshal.SizeOf(str) + 1;
byte[] arr = new byte[size];
IntPtr ptr = Marshal.AllocHGlobal(size);
Marshal.StructureToPtr(str, ptr, true);
Marshal.Copy(ptr, arr, 0, size);
Marshal.FreeHGlobal(ptr);
return arr;
}
As far as i know this code work as intendended, and im using the table generator to implement a hardcoded table in my microcontroller, to speed up the process.
What i dont quite get is how i implement a function to calculate the CRC in a similar way as i do here.
Any help or guides in the right direction is aprreciated.
So far i have come up with this function:
uint8_t crc8(uint8_t *crc)
{
uint8_t crcVal;
int m;
for (m = 0; m < PacketSize ;m++ )startbyte
{
*crc = crc8_table[(*crc) ^ m];
*crc &= 0xFF;
}
}
where table is:
uint8_t crc8_table[256] = {0,24,48,40,96,120,80,72,192,216,240,232,160,184,144,136,88,64,104,112,56,32,8,16,
152,128,168,176,248,224,200,208,176,168,128,152,208,200,224,248,112,104,64,88,16,8,
32,56,232,240,216,192,136,144,184,160,40,48,24,0,72,80,120,96,184,160,136,144,216,
192,232,240,120,96,72,80,24,0,40,48,224,248,208,200,128,152,176,168,32,56,16,8,64,
88,112,104,8,16,56,32,104,112,88,64,200,208,248,224,168,176,152,128,80,72,96,120,48,
40,0,24,144,136,160,184,240,232,192,216,168,176,152,128,200,208,248,224,104,112,88,
64,8,16,56,32,240,232,192,216,144,136,160,184,48,40,0,24,80,72,96,120,24,0,40,48,120,
96,72,80,216,192,232,240,184,160,136,144,64,88,112,104,32,56,16,8,128,152,176,168,224,
248,208,200,16,8,32,56,112,104,64,88,208,200,224,248,176,168,128,152,72,80,120,96,40,
48,24,0,136,144,184,160,232,240,216,192,160,184,144,136,192,216,240,232,96,120,80,72,
0,24,48,40,248,224,200,208,152,128,168,176,56,32,8,16,88,64,104,112
};
and PacketSize is found from rxFrame.Size

So you simply need to port your C# function to C
public byte Checksum(params byte[] val)
{
if (val == null)
throw new ArgumentNullException("val");
byte c = 0;
foreach (byte b in val)
{
c = table[c ^ b];
}
return c;
}
I. there are no exceptions in C. Use return value to indicate errors and add an argument that you'll use as a return value. It's up to you to decide whether to pass a message length as a parameter or leave it at the global scope:
int checksum(uint8_t const *msg, size_t msglen, uint8_t *result)
II. foreach loop is converted to for loop, where i is index and msg[i] is the b from the foreach:
int checksum(uint8_t const *msg, size_t msglen, uint8_t *result)
{
if (msg == NULL || msglen == 0)
return 0;
uint8_t crc = 0;
for (int i = 0; i < msglen; i++)
{
crc = table[crc ^ msg[i]];
}
III. Store the result and return a success code:
int checksum(uint8_t const *msg, size_t msglen, uint8_t *result)
{
if (msg == NULL || msglen == 0)
return 0;
uint8_t crc = 0;
for (int i = 0; i < msglen; i++)
{
crc = table[crc ^ msg[i]];
}
*result = crc;
return 1;
}
IV. Usage:
uint8_t crc;
if (!checksum(message, PacketSize, &crc))
report_error();

Looking at the C# code should be:
uint8_t crc8(uint8_t const *crc, size_t size)
{
unit8_t c = 0;
size_t m;
for (m = 0; m < size; m++ )
{
c = crc8_table[c ^ *crc];
crc++;
}
return c;
}

Don't use pointers, convert to arrays
public static byte crc8(byte[] crc)
{
byte crcVal;
int m;
for (m = 0; m < PacketSize; m++)
{
crc[m] = crc8_table[crc[m] ^ m];
crc[m] &= 0xFF;
}
return crcVal;
}

Cant get CRC function with Big Endian to work

I am having trouble getting a CRC function to work in C#. Below is the example C code given, below is the C# code I use. I suspect the culprit is the big endian, but when I switch the byte order of the checksum it still gives different values.
Any ideas where I am going wrong?
Many thanks
Matt
CRC is calculated using CRC-CCITT (Kermit) algorithm (http://www.lammertbies.nl/comm/info/crccalculation.html).
Parameter length – 2 bytes. Format is big endian. That is why bytes 0 and 1 are “switched”.
CRC calculation algorithm is shown below (C programming language).
/*----------------------------------------------------------------------------
* FUNCTION: CRC16
*---------------------------------------------------------------------------
*/
unsigned short crc_16_rec (unsigned char *pucData, unsigned short ucLen) {
//--------------------------------------------------------------------
unsigned int i;
unsigned char ucBit, ucCarry;
//--------------------------------------------------------------------
unsigned short usPoly = 0x8408;//reversed 0x1021
unsigned short usCRC = 0;
//--------------------------------------------------------------------
for (i = 0; i < ucLen; i++) {
usCRC ^= pucData[i];
for (ucBit = 0; ucBit < 8; ucBit++) {
ucCarry = usCRC & 1;
usCRC >>= 1;
if (ucCarry) {
usCRC ^= usPoly;
}
}
}
//--------------------------------------------------------------------
return usCRC;
//--------------------------------------------------------------------
}
The code I use
public enum Crc16Mode : ushort { Standard = 0xA001, CcittKermit = 0x8408 }
Crc16 CRC = new Crc16(Crc16Mode.CcittKermit);
public class Crc16
{
readonly ushort[] table = new ushort[256];
public ushort ComputeChecksum(params byte[] bytes)
{
ushort crc = 0;
for (int i = 0; i < bytes.Length - 2; ++i)
{
byte index = (byte)(crc ^ bytes[i]);
crc = (ushort)((crc >> 8) ^ table[index]);
}
return crc;
}
public bool ChecksumValid(ref byte[] bytes)
{
ushort checksum = (ushort)((bytes[bytes.Length - 2] << 8) | bytes[bytes.Length - 1]);
ushort checksum_received = CRC16(bytes);
if (checksum == checksum_received)
{
return true;
}
else
{
return false;
}
}
public ushort CRC16(byte[] bytes)
{
ushort crc = 0; //(ushort.maxvalue, 65535)
for (int j = 0; j < bytes.Length; j++)
{
crc = (ushort)(crc ^ bytes[j]);
for (int i = 0; i < 8; i++)
{
if ((crc & 0x0001) == 1)
crc = (ushort)((crc >> 1) ^ 0x8408);
else
crc >>= 1;
}
}
return (ushort)~(uint)crc;
}
public byte[] ComputeChecksumBytes(params byte[] bytes)
{
ushort crc = ComputeChecksum(bytes);
return BitConverter.GetBytes(crc);
}
public Crc16(Crc16Mode mode)
{
ushort polynomial = (ushort)mode;
ushort value;
ushort temp;
for (ushort i = 0; i < table.Length; ++i)
{
value = 0;
temp = i;
for (byte j = 0; j < 8; ++j)
{
if (((value ^ temp) & 0x0001) != 0)
{
value = (ushort)((value >> 1) ^ polynomial);
}
else
{
value >>= 1;
}
temp >>= 1;
}
table[i] = value;
}
}
}

Migrating CRC calculations between C and C#

I'm trying to implement crc calculation into my C# program. I've got code snippet from device C program but I'm unable to implement it into my program.
Here is code I've got:
uint16_t oblicz_crc(uint8_t* buffer, uint8_t length)
{
uint16_t crc = 0xffff;
uint8_t i, j;
for(i=0;i<length;i++)
{
crc ^= *(buffer+i);
for(j=0;j<8;j++)
{
3
if(crc & 0x0001)
{
crc >>= 1;
crc ^= 0xA001;
}
else
{
crc >>= 1;
}
}
}
return(crc);
}
And there's a code I already have in my program. I assume that some of the binary operators, an "if ((crc & 0x0001) == 1)" fragment, or some of data types are wrong. Can you help me with this?
public void Send_Frame(byte[] buffer)
{
byte[] checksum = Crc(buffer); //Array.Reverse(arg) ??
Array.Resize(ref buffer, buffer.Length + 2);
Array.Copy(checksum, 0, buffer, buffer.Length - 2, 2);
#if DEBUG
richTextBoxPreview.AppendText( BitConverter.ToString(buffer) + "\n");
#endif
//Port.Write(buffer);
}
private byte[] Crc(byte[] buffer)
{
UInt16 crc = 0xFFFF;
int i, j;
for (i = 0; i < buffer.Length; i++)
{
crc ^= (ushort)(BitConverter.ToUInt16(buffer, 0) + (byte)i);
for (j = 0; j < 8; j++)
{
if ((crc & 0x0001) == 1)
{
crc >>= 1;
crc ^= 0xA001;
}
else
{
crc >>= 1;
}
}
}
return BitConverter.GetBytes(crc);
}
PS: This question is about my specific code snippet. When I get answer will it be best to delete it? Or should I leave it ?

The problem is at this line:
crc ^= (ushort)(BitConverter.ToUInt16(buffer, 0) + (byte)i);
In the original code its:
crc ^= *(buffer+i);
The original code is adding i to the buffer pointer, and dereferencing that location. If you change the code to:
crc ^= (ushort)(buffer[i]);
then you're actually doing the same operation.

Does this work?
public static ushort oblicz_crc(byte[] buffer, int length)
{
unchecked
{
ushort crc = 0xffff;
for (int i = 0; i < length; i++)
{
crc ^= buffer[i];
for (int j = 0; j < 8; j++)
{
if ((crc & 0x0001) != 0)
{
crc >>= 1;
crc ^= 0xA001;
}
else
{
crc >>= 1;
}
}
}
return (crc);
}
}