.. and enable SSLpinning tests for mbedTLS, BoringSSL and LibreSSL.

Previously, when HTTP/2 is enabled and used, and stream has content
length known, Curl_read was not called when there was no bytes left to
read. Because of this, we could not make sure that
http2_handle_stream_close was called for every stream. Since we use
http2_handle_stream_close to emit trailer fields, they were
effectively ignored. This commit changes the code so that Curl_read is
called even if no bytes left to read, to ensure that
http2_handle_stream_close is called for every stream.

Discussed in https://github.com/bagder/curl/pull/564

This regression landed in 5778e6f5 and made libcurl not act on received
settings and instead stayed with its internal defaults.

Bug: http://curl.haxx.se/mail/lib-2016-01/0031.html
Reported-by: Bankde

This reverts commit 46cb70e9.

Bug: http://curl.haxx.se/mail/lib-2016-01/0031.html

Discussed in https://github.com/bagder/curl/pull/564

Use the ACE form of IDN hostnames as key in the connection cache.  Add
new tests.

Closes #592

- Fix ALPN reply detection.

- Wrap nghttp2 code in ifdef USE_NGHTTP2.


Prior to this change ALPN and HTTP/2 did not work properly in mbedTLS.

Check that the trailer buffer exists before attempting a client write
for trailers on stream close.

Refer to comments in https://github.com/bagder/curl/pull/564

Mistake from commit a464f338

To make sure curl doesn't allow multiplexing before a connection is
upgraded to HTTP/2 (like when Upgrade: h2c fails), we must make sure the
connection uses HTTP/2 as well and not only check what's wanted.

Closes #584

Patch-by: c0ff

- Add to both curl_global_init.3 and libcurl.3 the caveat for Windows
that initializing libcurl via a DLL's DllMain or static initializer
could cause a deadlock.

Bug: https://github.com/bagder/curl/issues/586


Reported-by:  <marc-groundctl@users.noreply.github.com>

MSYS would otherwise turn a /-style path into a C:\-style path.

While still using datacheck mode binary for the inline reply data.

Move the text-mode conversion for reply/replycheck from the verify
section into the load section and add support for 4 more check parts.

Not converting to double caused small timeouts to be skipped.

Not converting to double caused small timeouts to be skipped.

Previously file.txt[CR][LF] would have been returned as file.tx
(without the last t) if filetype is symlink. Now the t is
included and the internal item_length includes the zero byte.

Spotted using test 576 on Windows.

Follow up on b064ff0c, thanks Daniel.

This allows the root Makefile.am to include the Makefile.inc without
causing automake to warn on it (variables named *_SOURCES are
magic). curl_SOURCES is then instead assigned properly in
src/Makefile.am only.

Closes #577