Print Page - Hashing code explanation?

Title: Hashing code explanation?
Post by: XaiZou on August 29, 2015, 02:22:45 AM

Can anyone please explane the code below with comments ?

what are `pbegin` & `pend` ?

what is ( i believe it is data, where it comes from ? ): `(pbegin == pend ? pblank : static_cast<const void*>(&pbegin[0]))`

What is (i believe it is the size of data 80 ?) : `(pend - pbegin) * sizeof(pbegin[0])`

what is : `pblank[1]`

template<typename T1> inline uint256 HashShabal(const T1 pbegin, const T1 pend) { sph_shabal256_context ctx_shabal; static unsigned char pblank[1]; uint256 hash[1]; sph_shabal256_init(&ctx_shabal); // ZSHABAL; sph_shabal256 (&ctx_shabal, (pbegin == pend ? pblank : static_cast<const void*>(&pbegin[0])), (pend - pbegin) * sizeof(pbegin[0])); sph_shabal256_close(&ctx_shabal, static_cast<void*>(&hash[0])); return hash[0]; }

Title: Re: Hashing code explanation?
Post by: hhanh00 on August 29, 2015, 02:41:51 AM

pbegin is a pointer to the beginning of the data, pend is a pointer to the element *just after* the end of the array.
Therefore, *pbegin is the first element but *pend is invalid since it is past your array.
However, defining pend this way simplifies many formulas. For instance, (pend-pbegin) is the number of elements of your array. (pend-pbegin)*sizeof(T) is the size of your array in bytes. sizeof(T) == sizeof(&pbegin[0]) but sizeof(T) is more readable.

if (pend==pbegin), the array is empty. The code handles this boundary case specially by passing pblank instead of pbegin, possibly because sph_shabal256(p) evaluates *p which in this case is *pend and then could crash.
However, it is a bug of sph_shabal256 to do so. The caller passes a dummy array of one byte to make it happy.

Title: Re: Hashing code explanation?
Post by: XaiZou on August 30, 2015, 01:10:13 PM

Thanks, also what does this mean and what are the difference , second one is scrypt, what thash is ?? why second one is BEGIN-BEGIN not BEGIN-END ?

Code:

 uint256 GetPoWHash() const
    {
	return Hash(BEGIN(nVersion), END(nNonce));
    }

Code:

 uint256 GetPoWHash() const
    {
       uint256 thash;
       scrypt_1024_1_1_256(BEGIN(nVersion), BEGIN(thash));
       return thash;

    }

Title: Re: Hashing code explanation?
Post by: hhanh00 on August 30, 2015, 02:39:39 PM

Hash(pbegin, pend) computes the hash between [pbegin, pend) and returns the value
but scrypt(pbegin, presult) hashes starting from pbegin and returns the result in presult. The length of the data is probably hard coded.

Title: Re: Hashing code explanation?
Post by: wanghaoqd on May 13, 2016, 07:18:10 PM

Quote from: hhanh00 on August 30, 2015, 02:39:39 PM

Thanks for the explanation!

Bitcoin Forum

Alternate cryptocurrencies => Altcoin Discussion => Topic started by: XaiZou on August 29, 2015, 02:22:45 AM