libera #monero-pow

25 Sep 2023

Monero's Proof of Work algorithm (RandomX) and related topics. Use #p2pool-log and #p2pool-mini for P2Pool discussions. Anti-spam active: PM hyc for voice.

• 16:35
  tevador
  Btw, this whole thing is one FDIV_M RandomX instruction for RV64GC: paste.debian.net/hidden/93bdf103
• 16:45
  sech1
  Where did you get this code? RV64GC + vector instructions will be more compact
• 16:50
  tevador
  I wrote it. We have to target RV64GC, which has no vector support.
• 16:50
  tevador
  RV64GC is the "base" ISA for Linux.
• 16:53
  tevador
  When I compile some code with GCC, it will use -march=rv64imafdc_zicsr_zifencei, so this is our base ISA, to be more precise. It has RV64GC + CSR + FENCE.I.
• 16:55
  tevador
  It's safe to assume that anything that can run Linux will be at least rv64imafdc_zicsr_zifencei since that's what the linux kernel uses.
• 16:56
  sech1
  X5 has vector instructions by the way
• 16:56
  hyc
  then it must be C920, not C910
• 16:56
  tevador
  Yes, I don't think you can get anywhere near x86 performance without the vector extension.
• 16:57
  sech1
  are you using qemu, or did you get a real risc-v board?
• 16:57
  tevador
  I have a board with SiFive U74
• 16:58
  sech1
  nice
• 16:58
  sech1
  qemu is too slow even on 7950X. But being able to run 32 threads kind of fixes it :D
• 16:59
  hyc
  I'm still waiting for delivery of my lichee pi 4
• 17:00
  tevador
  paste.debian.net/hidden/0faa7063
• 17:01
  hyc
  isa-ext seems rather unhelpful
• 17:04
  tevador
  based on my research, the only reliable way to detect extensions is to run it and catch SIGILL...
• 17:06
  tevador
  I'm really starting to appreciate the x86 cpuid instruction.
• 17:09
  sech1
  Just assume RV64GC
• 17:09
  sech1
  as minimum supported for RandomX
• 17:09
  sech1
  and yes, test everything else
• 17:10
  sech1
  As far as I can see, you only need to test vector instructions and aes instructions
• 17:10
  sech1
  Everything else is covered by RV64GC
• 17:17
  sech1
  I started working on ARM64 code for RandomX CFROUND abd AES tweaks
• 17:17
  sech1
  Then I realized my RPi doesn't have AES, so I spent the day setting up aarch64 ubuntu in qemu
• 17:23
  tevador
  RV64GC has no rotate and scaled addition, these need Zba and Zbb extensions.
• 17:23
  sech1
  hmm, interesting
• 17:23
  sech1
  no rotate, as in ROR/ROL?
• 17:23
  tevador
  As for AES, I don't know any chips that support it.
• 17:24
  sech1
  not a big deal, ROR/ROL can be replaced by a couple bitshifts + logical or
• 17:24
  tevador
  Yes, rotate has to be emulated with shifts.
• 17:24
  tevador
  4 instructions instead of 1
• 17:25
  sech1
  risc-v cryptography extensions are ratified already, so future chips will have aes
• 19:37
  tevador
  Btw, there are at least 2 incompatible vector extensions being used in the wild: version 0.7.1 (SG2042 has it) and version 1.0.0. arxiv.org/abs/2304.10324
• 21:31
  hyc
  sech1 you don't have an arm64 smartphone to run on?
• 21:36
  sech1
  I have
• 21:39
  sech1
  But I need to ssh into it somehow to use it as a "aarch64 dev machine", and I'm always too lazy to set it up. I only use it for final testing in termux
• 21:39
  sech1
  btw I get 22 h/s in qemu on 7950X (single thread JIT)
• 21:40
  sech1
  and overall it's fast enough to not be annoying :D
• 21:42
  hyc
  heh
• 21:42
  hyc
  I build it all on termux.
• 21:42
  hyc
  doesn't take much to get the build env set up
• 21:43
  hyc
  and you can install openssh in termux too
• 21:43
  hyc
  ssh in is easy enough
• 21:44
  hyc
  of course I usually clone the repos from my laptop. don't have the patience to download everything from the web again
• 21:45
  hyc
  the other way I go, if it's just a quick compile/test, is leave the binary on my laptop, running tinyhttpd
• 21:45
  hyc
  then just grab the binary on the phone using any browser
• 21:49
  sech1
  plus, I can run 32 threads in randomx-benchmakr (dataset init is only 7 seconds)
• 21:49
  sech1
  and 32 threads when compiling