Fix cronet_transport.c failure handling bug

Fix cronet header memory free

Fix grpc._server._Context.time_remaining.

Update boringssl

A weak test is included; proper test coverage will come later.

Update path to http2 Java test client

Failing test is known issue #9150

Metadata handling rewrite

Optimizer was spotting that src1 and src2 would be identical and merging
them.

Fixing msan issue in client_ssl.c

adding content-type to the example response

Tweaked cpp stress client flags

Fix GRPC_TRACE doc formatting

Stop generating method name array when there is no method.

Change expected error code in Python negative interop client

See this error (somehow triggered by a new version of boringssl):
https://grpc-testing.appspot.com/job/gRPC_pull_requests_msan_c/1154/testReport/junit/(root)/c_linux_msan/bins_msan_handshake_client_GRPC_POLL_STRATEGY_poll_cv/
In the alpn callback, in_len is the size of the in buffer and not the number of alpn elements.

Merge pull request #47 from soltanmm-google/if-you-keep-getting-disappointed-do-you-start-to-just-accept-good-faith-or-get-jaded-about-good-faith

Fix Python memory errors

Doc string fix and tweaks.

... but for real this time.

In those places where we return an object that implements two
interfaces (let's say P and Q), consistently describe it as a "P-Q"
rather than once mentioning that it is a P and describing it as a "Q"
for the rest of prose.

pyformat_code.sh → yapf_code.sh.

C++11: Support compat with 1.0.x generated PB files

The test now expects to receive StatusCode.INTERNAL when it receives a
RST_STREAM from the server in order to comply with
https://github.com/grpc/grpc/blob/master/doc/PROTOCOL-HTTP2.md

Also added wait for ready behavior on the client channel.

Add ChannelArguments methods for setting max send/recv message size.