Count Down Latch

A one-shot concurrency primitive that blocks any fibers that wait on it. It is initialized with a positive integer n latches and waiting fibers are semantically blocked until all n latches are released. After this, further awaits are no-ops (equivalent to IO.unit).

trait CountDownLatch[F[_]] {

  def release: F[Unit]

  def await: F[Unit]

}

Cancelation of await is supported.

import cats.implicits._
import cats.effect._
import cats.effect.std.CountDownLatch
import cats.effect.unsafe.implicits.global

val run = (
  for {
    c <- CountDownLatch[IO](2)
    f <- (c.await >> IO.println("Countdown latch unblocked")).start
    _ <- c.release
    _ <- IO.println("Before latch is unblocked")
    _ <- c.release
    _ <- f.join
  } yield ()
)
// run: IO[Unit] = FlatMap(
//   ioe = FlatMap(
//     ioe = Map(
//       ioe = Delay(
//         thunk = cats.effect.IO$$$Lambda$17622/0x0000000803cc6440@3b3435d6,
//         event = cats.effect.tracing.TracingEvent$StackTrace
//       ),
//       f = cats.effect.std.CountDownLatch$State$$$Lambda$17623/0x0000000803cc4840@63462f8a,
//       event = cats.effect.tracing.TracingEvent$StackTrace
//     ),
//     f = cats.effect.std.CountDownLatch$$$Lambda$17624/0x0000000803cc4040@59885d12,
//     event = cats.effect.tracing.TracingEvent$StackTrace
//   ),
//   f = <function1>,
//   event = cats.effect.tracing.TracingEvent$StackTrace
// )

run.unsafeRunSync()