Previously, we use evbuffer_pullup(buf, -1) to linearize the memory
region and it may cause buffer copy. To avoid this, we use the return
value of evbuffer_get_contiguous_space() as 2nd parameter. According
to the libevent manual, by doing so evbuffer_pullup() will not copy or
modify any data in evbuffer.
It appears that some Android client does not like Connection: close in
200 CONNECT response. Browsers fine with this header field. It is
better to remove it. Squid does not emit it too.
If SPDY or HTTP/2 ustream is used and HTTP/2 downstream is used, only
call {spdylay,nghttp2}_resume_data when complete DATA frame was read
in backend to avoid to transmit too small DATA frame to the upstream.
4ed4efc does not disable TLS renegotiation at all, if client keeps
rengotiations without sending application data. In this change,
we intercept the raw incoming data from the client and if it is a
renegotiation, drop the connection immediately.
We thought that this kind of rewrite can be achieved by the configuration
of the backend severs, but in some configuration, however, it may get
complicated. So we decided to implement at least location rewrite in
nghttpx.
This commit also contains a fix to the bug which prevents the http2
backend request from concatenating header fields with the same value.
Now we set Downstream::set_response_connection_close(true) for
tunneled connections. Also call
Upstream::on_downstream_body_complete() callback when setting
MSG_COMPLETE in SpdySession when RST_STREAM is caught. Clean up EOF
handling in https_downstream_readcb.
Currently, resume_read() fails if on_read() returns -1 in case that
evbuffer_add failed, which means, most likely, memory allocation
failure. ClientHandler is marked "should be closed", but if
evbuffer_add is failed, write callback will not be invoked and its
marking is not evaluated. It will eventually be deleted when the
client is disconnected or backend failure though.
With --spdy-bridge option, it listens SPDY/HTTPS connections from
front end and forwards them to the backend in SPDY. The usage will be
written later. This change fixes the crash when more than 2
outstanding SpdyDownstreamConnection objects are added to SpdySession
and establishing connection to SPDY backend is failed.
INFO log and its surrounding code are now guarded by
LOG_ENABLED(SEVERITY) macro so that they don't run if log level
threshold is higher. This increases performance because log formatting
is somewhat expensive.
Added macros which log messages from the following components are
prefixed with their component name + object pointer address:
ListenHandler: LISTEN
ThreadEventReceiver: THREAD_RECV
Upstream: UPSTREAM
Downstream: DOWNSTREAM
DownstreamConnection: DCONN
SpdySession: DSPDY
In client mode, now SPDY connection to the backend server is
established per thread. The frontend connections which belong to the
same thread share the SPDY connection.
With --client-mode option, shrpx now accepts unencrypted HTTP
connections and communicates with backend server in SPDY. In short,
this is the "reversed" operation mode against normal mode. This may
be useful for testing purpose because it can sit between HTTP client
and shrpx "normal" mode.
When deciding whether to close the client connection, check
request_connection_close_ of Downstream in addition of
response_connection_close_. Also we only add "Connection: Keep-Alive"
header to the HTTP/1.0 or HTTP/0.9 clients.
To distinguish the to-be-installed programs and non-installable
example source code, the former programs, spdycat, spdydyd and shrpx,
were moved to src directory. spdynative was removed from Makefile
because it does not appeal to any users much.
Tatsuhiro Tsujikawa