remove fixed rx buffer allow definition per protocol
A new protocol member is defined that controls the size of rx
buffer allocation per connection. For compatibility 0 size
allocates 4096, but you should adapt your protocol definition
array in the user code to declare an appropriate value.
See the changelog for more detail.
The advantage is the rx frame buffer size is now tailored to
what is expected from the protocol, rather than being fixed
to a default of 4096. If your protocol only sends frames of
a dozen bytes this allows you to only allocate an rx frame
buffer of the same size.
For example the per-connection allocation (excluding headers)
for the test server fell from ~4500 to < 750 bytes with this.
Signed-off-by: Andy Green <andy.green@linaro.org>
diff --git a/libwebsockets-api-doc.html b/libwebsockets-api-doc.html
index e00dc59..c7129c7 100644
--- a/libwebsockets-api-doc.html
+++ b/libwebsockets-api-doc.html
@@ -934,6 +934,7 @@
<i>const char *</i> <b>name</b>;<br>
<i>callback_function *</i> <b>callback</b>;<br>
<i>size_t</i> <b>per_session_data_size</b>;<br>
+ <i>size_t</i> <b>rx_buffer_size</b>;<br>
<i>struct libwebsocket_context *</i> <b>owning_server</b>;<br>
<i>int</i> <b>protocol_index</b>;<br>
};<br>
@@ -951,6 +952,15 @@
this much memory allocated on connection establishment and
freed on connection takedown. A pointer to this per-connection
allocation is passed into the callback in the 'user' parameter
+<dt><b>rx_buffer_size</b>
+<dd>if you want atomic frames delivered to the callback, you
+should set this to the size of the biggest legal frame that
+you support. If the frame size is exceeded, there is no
+error, but the buffer will spill to the user callback when
+full, which you can detect by using
+<b>libwebsockets_remaining_packet_payload</b>. Notice that you
+just talk about frame size here, the LWS_SEND_BUFFER_PRE_PADDING
+and post-padding are automatically also allocated on top.
<dt><b>owning_server</b>
<dd>the server init call fills in this opaque pointer when
registering this protocol with the server.