search

kramble / FPGA-Litecoin-Miner

A litecoin scrypt miner implemented with FPGA on-chip memory.
GNU General Public License v3.0
277 stars 125 forks source link

Test data minor issue #6

Open razorfish-sl opened 10 years ago

razorfish-sl commented 10 years ago

Getting the following with the latest code, set for two cores There seems to be one more valid hash for 71 !!, everything else is no error

Tested 70 passed 70 fail 0 unmatched 0 Test 000007ff0043e4ce9a6c491b3119fb516030a368196617e5c5c901d562f24416dd1dda66db011278b8c2b2cb93383c7de354106474bee7afe991e590271b6e679e74eb1716340b5a0957589f04a1cf3601000000 Share found on Sat Nov 30 15:00:50 2013 nonce 0043e500 ... CORRECT Tested 71 passed 71 fail 0 unmatched 0 Test 000007ff00227f0b9a6c491bad19fb51e6597f5ca35e7f9f83b9aa228e457c5d7b4693c473028b34be5db64fe8178b93330429ae8107d2beddb7ff90a46f7fe0b1321e5ad4c3dba805341698445f6f7102000000 Share found on Sat Nov 30 15:00:50 2013 nonce 10227f2b ... ERROR expected nonce 0x227f3d Share found on Sat Nov 30 15:00:50 2013 nonce 00227f3d ... CORRECT

razorfish-sl commented 10 years ago

Sorry this highlights it better.. we have un-mached which dissapears

Tested 175 passed 175 fail 1 unmatched -1 Test 000007ff00a7a67bc7bb4f1bdc64fb5161bd3cdce2d157f21531044822bc20beb390ccf1d60527bed8b797e06755ed456c59ede5e903ba9d1a2ea2550872c200fd58dcf587c3759a0597dbf44e9bee0a02000000 Share found on Sun Dec 1 09:52:19 2013 nonce 00a7a6ad ... CORRECT Tested 176 passed 176 fail 1 unmatched -1 Test 000007ff010e1b58c7bb4f1be464fb5145564d126b26cbf45c6d39e8c2b7c5ecdaf1db2db1187ceadd54a6cc08db11993c86ea2ceddb902870e17c0142fbfd89a2d637d8cd8a669feac4df0d6b22fad802000000 Tested 177 passed 176 fail 1 unmatched 0 Test 000007ff003fee77c7bb4f1b7265fb514a6ee37d8e140b1c9c6e873810f90b31e5f7b14a06c61a20443402b59e40fc71bcf3ca733f83c119e111545ff65ab282d56e01cc2dc7e6b35e46b4889a3de1ad02000000 Share found on Sun Dec 1 09:52:21 2013 nonce 003feea9 ... CORRECT Tested 178 passed 177 fail 1 unmatched 0 Test 000007ff0007cda8c7bb4f1b5565fb517b57ade3f0113cbf0946da892dfc13b9d0c63578a6906c85e702ca920625629ab9dd24540a7d642af33b799bddd688c6d64b40740727ba133b0a851f3b7d9d2301000000 Share found on Sun Dec 1 09:52:21 2013 nonce 0007cdda ... CORRECT

kramble commented 10 years ago

As you correctly concluded, the ERROR in the first post is explained as its matching on a different core, and the solution would be to omit this from the test data set. The second is simply sloppy coding on my part (see line 97 https://github.com/kramble/FPGA-Litecoin-Miner/blob/master/ICARUS-LX150/MiningSoftware/ltcminer-testmode.py#L97) as unmached is calculated assuming at most one match per test whereas we get two in this case. I'll have a think about a better way to calculate this, thanks for the report.