-Alongflagname
    	disable bounds checking
  -C	a boolean defaulting to true (default true)
  -D path
    	set relative path for local imports
  -E string
    	issue 23543 (default "0")
  -F number
    	a non-zero number (default 2.7)
  -G float
    	a float that defaults to zero
  -M string
    	a multiline
    	help
    	string
  -N int

* Add a timeout to the sshDialer to prevent indefinite hangs. ([#23843](https://github.com/kubernetes/kubernetes/pull/23843), [@cjcullen](https://github.com/cjcullen))
* AWS kube-up: tolerate a lack of ephemeral volumes ([#23776](https://github.com/kubernetes/kubernetes/pull/23776), [@justinsb](https://github.com/justinsb))

//              U(z) = U0[0] + U0[1]*z + ... + U0[4]*z**4
//              V(z) = 1  + v0[0]*z + ... + v0[4]*z**5
//         Note: For tiny x, 1/x dominate y1 and hence
//              y1(tiny) = -2/pi/tiny, (choose tiny<2**-54)
//      3. For x>=2.
//               y1(x) = sqrt(2/(pi*x))*(p1(x)*sin(x1)+q1(x)*cos(x1))
//         where x1 = x-3*pi/4. It is better to compute sin(x1),cos(x1)
//         by method mentioned above.

						t.Error(err)
					}
					w.SetWriteDeadline(time.Now().Add(time.Millisecond))
				}
			}
		}()
	}
}

// Closing a TTY while reading from it should not hang.  Issue 23943.
func TestTTYClose(t *testing.T) {
	// Ignore SIGTTIN in case we are running in the background.
	signal.Ignore(syscall.SIGTTIN)
	defer signal.Reset(syscall.SIGTTIN)

	f, err := os.Open("/dev/tty")

	err = p.copyResponse(rw, res.Body, p.flushInterval(res))
	if err != nil {
		defer res.Body.Close()
		// Since we're streaming the response, if we run into an error all we can do
		// is abort the request. Issue 23643: ReverseProxy should use ErrAbortHandler
		// on read error while copying body.
		if !shouldPanicOnCopyError(req) {
			p.logf("suppressing panic for copyResponse error in test; copy error: %v", err)
			return
		}

// The initial hash generation is omitted from the test, because it takes a long time.
// The test contains some already-generated states, and their expected sums
// Tests a problem that is outlined in GitHub issue #29543
// The problem is triggered when an amount of data has been hashed for which
// the data length has a 1 in the 32nd bit. When casted to int, this changes
// the sign of the value, and causes the modulus operation to return a

	// half (a quarter overall) will be SLLS; an eighth will be SLLLS, and so on.
	// Not counting the final S in each (which overlaps the first S in the next),
	// This works out to an average length 2×½ + 3×¼ + 4×⅛ + ... = 3.
	// The space we need is further reduced by the fact that many of the
	// short patterns like SLS will often be the same character sequences
	// repeated throughout the text, reducing maxID relative to numLMS.
	//