* >60% faster than the Euclidean algorithm in benchmarks.
     */
    int aTwos = Long.numberOfTrailingZeros(a);
    a >>= aTwos; // divide out all 2s
    int bTwos = Long.numberOfTrailingZeros(b);
    b >>= bTwos; // divide out all 2s
    while (a != b) { // both a, b are odd
      // The key to the binary GCD algorithm is as follows:
      // Both a and b are odd. Assume a > b; then gcd(a - b, b) = gcd(a, b).

            .build()

        val exchange = call.interceptorScopedExchange
        val followUp = followUpRequest(response, exchange)

        if (followUp == null) {
          if (exchange != null && exchange.isDuplex) {
            call.timeoutEarlyExit()
          }
          closeActiveExchange = false
          call.eventListener.followUpDecision(call, response, null)
          return response
        }

   * inserted into the multiset must be <i>mutually comparable</i> by the specified comparator:
   * {@code comparator.compare(e1, e2)} must not throw a {@code ClassCastException} for any elements
   * {@code e1} and {@code e2} in the multiset. If the user attempts to add an element to the
   * multiset that violates this constraint, the {@code add(Object)} call will throw a {@code
   * ClassCastException}.

 * {@link UnsupportedOperationException} when they are not implemented. Most implementations should
 * support either all add operations or none of them, all removal operations or none of them, and if
 * and only if all of these are supported, the {@code setCount} methods as well.
 *
 * <p>A multiset uses {@link Object#equals} to determine whether two instances should be considered

work.  (Executables containing this object code plus portions of the
Library will still fall under Section 6.)

  Otherwise, if the work is a derivative of the Library, you may
distribute the object code for the work under the terms of Section 6.
Any executables containing that work also fall under Section 6,
whether or not they are linked directly with the Library itself.

And it will save the returned value in a <abbr title="A utility/system to store computed/generated values, to reuse them instead of computing them again.">"cache"</abbr> and pass it to all the "dependants" that need it in that specific request, instead of calling the dependency multiple times for the same request.

            "No mapping between account names and security IDs was done.", "The security ID structure is invalid.",
            "The file system is full.", "All pipe instances are busy.", "All pipe instances are busy.", "The pipe state is invalid.",
            "All pipe instances are busy.", "No process is on the other end of the pipe.", "The pipe is being closed.",

 * **ExchangeFinder** chooses which connection carries each exchange. Where possible it will use the same connection for all exchanges in a single call. It prefers reusing pooled connections over establishing new connections.      

#### Per-Connection Locks

import java.util.concurrent.TimeUnit;
import org.jspecify.annotations.Nullable;

/**
 * A TimeLimiter implementation which actually does not attempt to limit time at all. This may be
 * desirable to use in some unit tests. More importantly, attempting to debug a call which is
 * time-limited would be extremely annoying, so this gives you a time-limiter you can easily swap in
 * for your real time-limiter while you're debugging.
 *

     *
     * @param flag the flag to set
     */
    public void setFlag(final int flag) {
        this.flags |= flag;
    }

    /**
     * Get the result of the DCERPC call.
     * @return result exception, if the call failed
     */
    public DcerpcException getResult() {
        if (this.result != 0) {
            return new DcerpcException(this.result);
        }
        return null;
    }