logstage perf tests

PerfTest.scala

package com.github.pshirshov.izumi

import com.github.pshirshov.izumi.logstage.api.logger.LogRouter
import com.github.pshirshov.izumi.logstage.api.{IzLogger, Log}

object PerfTest {
  final val acceptingNullRouter = new LogRouter {
    override def acceptable(id: Log.LoggerId, messageLevel: Log.Level): Boolean = true

    override protected def doLog(entry: Log.Entry): Unit = {}
  }

  def main(args: Array[String]): Unit = {
    val logger = IzLogger(acceptingNullRouter)
//    val logger = IzLogger.NullLogger

    var i0 = 0
    while (i0 < 10000) {
      i0+=1
      logger.debug("jit")
    }

    val results = 1 to 1000 map { _ =>
      val start = System.nanoTime()
      var i = 0
      while (i < 100000) {
        i += 1

        logger.debug("heya-ya-hey")
      }
      val end = System.nanoTime()

      end-start: BigInt
    }

    println(s"Result: avg of 1000 runs ${results.sum[BigInt] / results.size} nanos")

  }
}

// accepting null tests:
// new (lambda, anyval):
// Result: avg of 1000 runs 14876492 nanos
// Result: avg of 1000 runs 15245759 nanos
// Result: avg of 1000 runs 13890930 nanos
// Result: avg of 1000 runs 15674452 nanos
// Result: avg of 1000 runs 18266810 nanos
// Result: avg of 1000 runs 15266548 nanos

// new (nolambda, anyval):
// Result: avg of 1000 runs 13813806 nanos
// Result: avg of 1000 runs 13838628 nanos
// Result: avg of 1000 runs 13879927 nanos

// new (abstract, nolambda, anyval, applicationpointid)
// Result: avg of 1000 runs 13506974 nanos
// Result: avg of 1000 runs 13489046 nanos

// old (no anyval, no lambda):
// Result: avg of 1000 runs 15135465 nanos
// Result: avg of 1000 runs 14133218 nanos
// Result: avg of 1000 runs 14609388 nanos
// Result: avg of 1000 runs 14005947 nanos
// Result: avg of 1000 runs 14039547 nanos
// Result: avg of 1000 runs 14214438 nanos
// Result: avg of 1000 runs 14177798 nanos

// null tests:

// new:
// Result: avg of 1000 runs 14102548 nanos
// Result: avg of 1000 runs 13658187 nanos

// inline all:
// Result: avg of 1000 runs 12602929 nanos
// Result: avg of 1000 runs 1054526 nanos

// inline constants:
// Result: avg of 1000 runs 6696346 nanos

// val all:
// Result: avg of 1000 runs 832194 nanos

// lambda + no-anyval-codeposition
// Result: avg of 1000 runs 1031065 nanos

// lambda + anyval codeposition
// Result: avg of 1000 runs 461838 nanos
// Result: avg of 1000 runs 483007 nanos

// lambda-abstract + anyval codeposition
// Result: avg of 1000 runs 447830 nanos

// nolambda + applicationPointId literal
// Result: avg of 1000 runs 7035 nanos

// old:
// Result: avg of 1000 runs 1114780 nanos


// old = inlined:
// Result: avg of 1000 runs 14489897 nanos

// old = inlined-with-proper-if:
// Result: avg of 1000 runs 1066996 nanos

// print tests:

// Result: 2733118063 nanos
// Result: 2803161680 nanos
// Result: 2991024260 nanos

// new:
// Result: avg of 10 runs 2285790767 nanos
// Result: avg of 10 runs 2216757380 nanos
// Result: avg of 10 runs 2292143744 nanos
// Result: avg of 10 runs 3205364461 nanos
// Result: avg of 10 runs 1931434831 nanos
// Result: avg of 10 runs 1842183675 nanos
// Result: avg of 10 runs 1871253507 nanos
// Result: avg of 10 runs 1976296972 nanos

// old:
// Result: avg of 10 runs 2188380910 nanos
// Result: avg of 10 runs 2275079815 nanos
// Result: avg of 10 runs 1860807122 nanos
// Result: avg of 10 runs 2013032035 nanos