sech1: please review tevador/RandomX #179

I've added exactly this function for case when miner calculates the last nonce

but i didn't change "uint32_t chan" to "int chain", maybe it's needed

it's the same unless RANDOMX_PROGRAM_COUNT exceeds INT_MAX, which won't happen

btw, with batching, the browser miner will have ~95% of native performance

tevador looks good, but naming of these functions can be improved I think. It's not obvious that output in *_next function is for the previous hash. But this is for another PR.

maybe people actually read comments in the code, I don't know

I just did few recompilation until got why hashes from _first/_next doesn't match with _hash and reordered some lines of code

sech1: yes, I got an email the other day, some guy was asking why his hashes don't match - guess why, his nonce was one off because of how randomx_calculate_hash_next works