... to avoid a build failure when configured with --enable-debug

Previously, we just ignored error code passed to
on_stream_close_callback and just return 0 (success) after stream
closure even if stream was reset with error.  This patch records error
code in on_stream_close_callback, and return -1 and use CURLE_HTTP2
error code on abnormal stream closure.

As the 'error' and 'mute' options are now part of the GlobalConfig,
rather than per Operation, updated the warnf() function to use this
structure rather than the OperationConfig.

Removed the DataExecutionPrevention directive from the project files for
Visual Studio 2008 and above. The XML value in the VC9 project files was
set to "0" (Default) whilst the VC10+ project files contained an empty
XML element.

Visual Studio 2008 introduced support for the address space layout
randomization (ASLR) feature of Windows Vista. However, upgrading the
VC8 project files to VC9 and above disabled this feature.

Removed the RandomizedBaseAddress directive to enabled the default
setting (/DYNAMICBASE). Note: This doesn't appear to have any negative
impact when compiled and ran on Windows XP.

Pointed-out-by: Tatsuhiro Tsujikawa
Bug: http://curl.haxx.se/bug/view.cgi?id=1487

... in order to make test1135 succeed

Otherwise it expands to:

    echo ""/etc/pki/tls/certs/ca-bundle.crt""

Detected by ShellCheck:

    curl-config:74:16: warning: The double quotes around this do
    nothing.  Remove or escape them. [SC2140]

In that case, we only skip writing the error message for failed NSS
initialization (while still returning the correct error code).

The vtls layer now checks the return value, so it is no longer necessary
to abort if a random number cannot be provided by NSS.  This also fixes
the following Coverity report:

Error: FORWARD_NULL (CWE-476):
lib/vtls/nss.c:1918: var_compare_op: Comparing "data" to null implies that "data" might be null.
lib/vtls/nss.c:1923: var_deref_model: Passing null pointer "data" to "Curl_failf", which dereferences it.
lib/sendf.c:154:3: deref_parm: Directly dereferencing parameter "data".

... and remove some old ones

This reverts commit 03fa5768.

obj_count can be 1 if the custom read function is set or the stdin
handle is a reference to a pipe. Since the pipe should be handled
using the PeekNamedPipe-check below, the custom read function should
only be used if it is actually enabled.

According to [1]: "Returning 0 will signal end-of-file to the library
and cause it to stop the current transfer."
This change makes the Windows telnet code handle this case accordingly.

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

... by making sure the first request is completed before doing the
remainder.

... even if the last IPv6 connection attempt has failed.

Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1187531#c4

... in case the protocol versions are mixed in a DNS response
(IPv6 -> IPv4 -> IPv6).

Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1187531#c3

Remove cmake as marked for removal. It is in much better state now.

SSL_CTX_load_verify_locations by default (and if given non-Null
parameters) searches the CAfile first and falls back to CApath.  This
allows for CAfile to be a basis (e.g. installed by the package manager)
and CApath to be a user configured directory.

This wasn't reflected by the previous configure constraint which this
patch fixes.

Bug: https://github.com/bagder/curl/pull/139

Correctly check for memcmp() return value (it returns 0 if the strings match).

This is not really important, since curl is going to use http/1.1 anyway, but
it's still a bug I guess.

Correctly check for strncmp() return value (it returns 0 if the strings
match).

Use "cmake -E echo" instead of "echo".

Reviewed-by: Brad King <brad.king@kitware.com>

Set CMAKE_REQUIRED_DEFINITIONS to include definitions needed to get
the winsock2 API from windows.h.  Simplify the order of checks to
avoid extra conditions.

Use check_include_file instead of check_include_file_concat to look
for OpenSSL headers.  They do not need to participate in a sequence
of dependent system headers.  Also they may cause winsock.h to be
included before ws2tcpip.h, causing the latter to not be detected
in the sequence.

Reviewed-by: Brad King <brad.king@kitware.com>