In addition to NTLM, use Kerberos and SPNEGO as proxies to the crypto
feature.

...and converted tab characters, from commit 4b4e8a58, to spaces.

Such as !SSPI as we do for the NTLM and Digest tests.

Verifies the upgrade request, but gets a plain 1.1 response

This commit disables pipelining for HTTP/2 or upgraded connections.  For
HTTP/2, we do not support multiplexing.  In general, requests cannot be
pipelined in an upgraded connection, since it is now different protocol.

Updated following commit eda919f4 and a4b7f716

.

Acked-by: Brad King <brad.king@kitware.com>
Signed-off-by: Peter Wu <peter@lekensteyn.nl>

Updated following changes in commit f0d860d3

.

Acked-by: Brad King <brad.king@kitware.com>
Signed-off-by: Peter Wu <peter@lekensteyn.nl>

Prior to this change when no file descriptors were ready on platforms
other than Windows the multi examples would sleep whatever was in
timeout, which may or may not have been less than the minimum
recommended value [1] of 100ms.

[1]: http://curl.haxx.se/libcurl/c/curl_multi_fdset.html

When the connection code decides to close a socket it informs the multi
system via the Curl_multi_closed function. The multi system may, in
turn, invoke the CURLMOPT_SOCKETFUNCTION function with
CURL_POLL_REMOVE. This happens after the socket has already been
closed. Reorder the code so that CURL_POLL_REMOVE is called before the
socket is closed.

Moved target autodetection block after defining CC macro.

Remove use of an old hack that takes advantage of the auto-dereference
behavior of the if() command to detect if a variable is defined.  The
hack has the form:

 if("${VAR} MATCHES "^${VAR}$")

where "${VAR}" is a macro argument reference.  Use if(DEFINED) instead.
This also avoids warnings for CMake Policy CMP0054 in CMake 3.1.

Debug output 'typo' fix.

Don't print an extra "0x" in
  * Pipe broke: handle 0x0x2546d88, url = /

Add debug output.
Print the number of connections in the connection cache when
  adding one, and not only when one is removed.

Fix typos in comments.

... as it was before I changed the loop in commit e04ccbd5. It caused
test 2030 and 2032 to fail.

... and assign data->result only at the end. Makes the code more compact
(easier to read) and more similar to other code.

save 'result' for CURLcode types

simplifies the use of this function at little cost.

Fixes the problem when a transfer in a pipeline times out.

Even though commit 23e70e1c mentioned the v110_xp toolset, I had
forgotten to include the relevant pre-processor definitions.