Post
Topic
Board Announcements (Altcoins)
Re: [ANN] [CPU mining] Yenten v1.2.1 [YTN] [YescryptR16] [Exchange avilable]
by
eule
on 06/12/2017, 19:42:24 UTC
Finally got my Odroid XU3 lite mining by compiling yentend, wasn't able to compile cpuminer-opt. Is the source of minerd available somewhere?
Mining on 7 of 8 cores at 93°C, Hashrate reports vary wildly: most often it's around 63 but rarely it's reported as 0 or up to 140, probably some kind of lag. Not too bad for an ARM CPU (Exynos 5422), OS is Debian Wheezy. My properly cooled and slightly overclocked FX 8320E does 650 hashes per second btw.

Had to configure with --ignore-incompatible-bdb and remove a "-msse41" from yescrypt.h in src/hash/yescrypt. More could be possible with some ARM optimization and compile option tuning, my only CXXFLAG was O3.

Too bad the hashrate almost doubled today.  Shocked

Wow nice, I tried to get it working on a pine64 which didn't work and it seemed to be missing x86 type operations so I assumed that yenten required x86 processors.  But that is interesting that you got it to work on ARM and not only that but on a non-ubuntu distro.  Would you mind writing up a little procedure on how to do it for a linux noob?  If this works on ARM it changes everything...

Also add up each core's hash/sec to get your total.
Found the minerd source at https://github.com/conan-equal-newone/yenten/releases, all I had to do to get running:
Code:
apt-get install build-essential autotools-dev autoconf libcurl3 libcurl4-gnutls-dev jansson
wget https://github.com/conan-equal-newone/yenten/releases/download/1.1.0/yenten_minerd_source.zip
unzip yenten_minerd_source.zip
Edit line 6 of configure.ac: AM_INIT_AUTOMAKE([gnu foreign])
./autogen.sh
./configure CFLAGS="-O3 -funroll-loops -fomit-frame-pointer -march=armv7-a -mfpu=vfpv3-d16 -mfloat-abi=hard"
make
strip minerd
-march=armv7-a -mfpu=vfpv3-d16 -mfloat-abi=hard will be different for most ARM devices, look up the recommended CFLAGS for your device.
To build yentend I followed the readme except the two steps mentioned above.
The 4 big cores report 0.02kH/s each, the 4 little cores 0.01kH/s. So with proper cooling I would get at least 120H/s, only running the 4 big cores atm.