Open david-j-smith opened 5 years ago
hello, sorry i focused my current release on the new Graft fork but then i'll fix that bug. each multi-hash is a dedicated assembly so it's very possible there's a bug in multi 2 and not 3. thanks for the report !
I just found the bug, and it's a lot worse than i expected, with other algos impacted like Electroneum (even if no body mine it with a CPU any longer...). I'll release an emergency fix sooner than planned.
It should be fixed now, with the 0.33q i just re-released. Note that only the multi_hash=2 with AES was impacted, all other from 1 to 6 were correct.
Greetings, in Version 0.33p CPU there seems to be a Bug regarding its extremely valuable Multihashing Feature. When using Variation 21 and a Masari-Pool (observed with pool.masari.hashvault.pro), an even Value of "multi_hash" in "cpu_threads_conf" in a config file leads to rejected Hashes (tested with Value 2). Using an odd Value (tested with 3) or disabling this handy feature gets the Hashes accepted as expected.
I can reproduce this on several Machines, the above was tested on a Ryzen 1700 (Windows 10) and 1 Thread, but it also happens on Ubuntu 18.04 on a Xeon 8XXX (also 1 Thread).
I really like this Miner and hope that this Bug gets addressed soon.
Thank you very much for your Time!