artempyanykh

import java.util.*;
import java.util.stream.Collectors;
import java.util.stream.Stream;

sealed interface Expr permits Var, Lam, App {
    default Set<String> freeVars() {
        return switch (this) {
            case Var(var name) -> Set.of(name);
            case Lam(var bound, var body) -> {
                var vars = new HashSet<>(body.freeVars());

artempyanykh

type Vec =
    { x: float
      y: float
      z: float }

    static member (+)(v1, v2) = { x = v1.x + v2.x; y = v1.y + v2.y; z = v1.z + v2.z }
    static member (*)(factor, v2) = { x = factor + v2.x; y = factor + v2.y; z = factor + v2.z }

let test (v1: Vec) (v2: Vec) =
    let factorSum = 1.0 + 2.5

artempyanykh

Memory mapping on aarch64

System information:

lscpu
echo "================================================="
cat /proc/cpuinfo

Architecture:        aarch64

artempyanykh

import java.lang.management.ManagementFactory
import java.util.concurrent.ThreadPoolExecutor

import scalaz.zio.RTS
//import scalaz.zio.IO

object TheadPoolTestApp {
  def main(args: Array[String]): Unit = {
    val rts = new RTS {}
    val pool = rts.threadPool.asInstanceOf[ThreadPoolExecutor]

artempyanykh

cmake_minimum_required(VERSION 3.12)
project(jdk)

set(CMAKE_CXX_STANDARD 11)

set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=gnu++98 -DASSERT -D_GNU_SOURCE -D_REENTRANT -DLINUX -DVM_LITTLE_ENDIAN -D_LP64 -DTARGET_ARCH_x86 -DINCLUDE_SUFFIX_OS=_linux -DINCLUDE_SUFFIX_CPU=_x86 -DINCLUDE_SUFFIX_COMPILER=_gcc -DTARGET_COMPILER_gcc -DAMD64 -DHOTSPOT_LIB_ARCH='amd64' -DCOMPILER1 -DCOMPILER2")

include_directories(
        src/hotspot/cpu/x86
        src/hotspot/os/linux

artempyanykh

Scalaz RingBuffer

Goal

Scalaz RingBuffer adds a high-performance low-latency multi-producer/multi-consumer task queue to RTS for faster and smarter concurrency in ZIO.

Introduction & Highlights

ZIO RTS runs on a thread pool executor backed by LinkedBlockingQueue. This has the following consequences:

1. Performance and latency are inhibited because: a) waiting on locks cause threads to be de-scheduled, b) traversing a linked list is not cache-friendly.
2. Optimizations to Fiber cooperative multi-tasking strategy is limited because the thread pool doesn't expose any metrics on its queue remaining capacity\processing times.