int j = i & ARRAY_MASK;
      tmp += LongMath.sqrt(positive[j], mode);
    }
    return tmp;
  }

  @Benchmark
  int divide(int reps) {
    int tmp = 0;
    for (int i = 0; i < reps; i++) {
      int j = i & ARRAY_MASK;
      tmp += LongMath.divide(longs[j], nonzero[j], mode);
    }
    return tmp;
  }

  }

  /**
   * Binary search [sections] for [codePoint], looking at its top 14 bits.
   *
   * This binary searches over 4-byte entries, and so it needs to adjust binary search indices
   * in (by dividing by 4) and out (by multiplying by 4).
   */
  private fun findSectionsIndex(codePoint: Int): Int {
    val target = (codePoint and 0x1fff80) shr 7
    val offset =
      binarySearch(
        position = 0,

  }

  /**
   * Verifies that the specified expected result is always produced by collecting the specified
   * inputs, regardless of how the elements are divided.
   */
  @SafeVarargs
  @CanIgnoreReturnValue
  @SuppressWarnings("nullness") // TODO(cpovirk): Remove after we fix whatever the bug is.

        result *= i;
      }
      return BigInteger.valueOf(result);
    }

    /*
     * We want each multiplication to have both sides with approximately the same number of digits.
     * Currently, we just divide the range in half.
     */
    int mid = (n1 + n2) >>> 1;
    return oldSlowFactorial(n1, mid).multiply(oldSlowFactorial(mid, n2));
  }

  @Benchmark
  int slowFactorial(int reps) {
    int tmp = 0;

        result *= i;
      }
      return BigInteger.valueOf(result);
    }

    /*
     * We want each multiplication to have both sides with approximately the same number of digits.
     * Currently, we just divide the range in half.
     */
    int mid = (n1 + n2) >>> 1;
    return oldSlowFactorial(n1, mid).multiply(oldSlowFactorial(mid, n2));
  }

  @Benchmark
  int slowFactorial(int reps) {
    int tmp = 0;

    int bytesToCopy = bufferSize - buffer.position();
    for (int i = 0; i < bytesToCopy; i++) {
      buffer.put(readBuffer.get());
    }
    munch(); // buffer becomes empty here, since chunkSize divides bufferSize

    // Now process directly from the rest of the input buffer
    while (readBuffer.remaining() >= chunkSize) {
      process(readBuffer);
    }

      // Avoid delegating to this(long[]), since AtomicLongArray(long[]) will clone its input and
      // thus double memory usage.
      this.data =
          new AtomicLongArray(Ints.checkedCast(LongMath.divide(bits, 64, RoundingMode.CEILING)));
      this.bitCount = LongAddables.create();
    }

    // Used by serialization
    LockFreeBitArray(long[] data) {
      checkArgument(data.length > 0, "data length is zero!");

        if (b != 0) {
          UnsignedInteger aUnsigned = UnsignedInteger.fromIntBits(a);
          UnsignedInteger bUnsigned = UnsignedInteger.fromIntBits(b);
          int expected = aUnsigned.bigIntegerValue().divide(bUnsigned.bigIntegerValue()).intValue();
          UnsignedInteger unsignedDiv = aUnsigned.dividedBy(bUnsigned);
          assertThat(unsignedDiv.intValue()).isEqualTo(expected);
        }
      }
    }
  }

      // non-negative int, we can do long-arithmetic on index * (dataset.length - 1) / scale to get
      // a rounded ratio and a remainder which can be expressed as ints, without risk of overflow:
      int quotient = (int) LongMath.divide(numerator, scale, RoundingMode.DOWN);
      int remainder = (int) (numerator - (long) quotient * scale);
      selectInPlace(quotient, dataset, 0, dataset.length - 1);
      if (remainder == 0) {

 * to that of {@code ConcurrentHashMap} in a reusable form, and extends it for semaphores and
 * read-write locks. Conceptually, lock striping is the technique of dividing a lock into many
 * <i>stripes</i>, increasing the granularity of a single lock and allowing independent operations
 * to lock different stripes and proceed concurrently, instead of creating contention for a single
 * lock.
 *