* Fix pf_server_new: pass own copy of proxyConfig
The lifecycle of proxyConfig passed to pf_server_new is unknown,
so pass proxyServer->config copy to modules.
* Early free parsed proxyConfig
* Split out proxy headers and moved to public API to allow external
modules to be built.
* Split proxy into proxy library and proxy binary. The library
can be used by other applications and provides a simple API
* Improved channel passthrough, now all channels including dynamic
channels work.
* Extended module API to hook more events, improved module samples
* Cleaned up proxy code, removed global static variables used,
added WINPR_ASSERT
Proxy server might receive input updates while client side being
redirected (and reset), then a heap-after-free might occur. To solve
this, we register server input/update callbacks only after client is
fully activated.
* Fix some missing argument checks for function pointer implementations
* Fix broken return value check for client->SendChannelData
* Updated const correctness for function pointer implementations
* Add support for C++ plugins.
* Seperate between a "module" and a "plugin": a module is a shared
library that exposes a function called "proxy_module_entry_point". This
function gets a parameter to `proxyPluginsManager*`, which can be used to
register plugins.
* Refine the API of the modules infrastructure:
* use proxyData* instead of rdpContext* when firing a hook or calling a filter.
* use clearer names.
* Instead of having to specify a path for each module to be loaded in
the configuration, the proxy now loads modules from specificed
directory, using the CMAKE variable PROXY_PLUGINDIR.
* Add an option to specify required plugins: plugins that the proxy
wouldn't start without having them fully loaded.
To avoid a race in which the server received an input event before
proxy's client is initialized, we now only register input and update
callbacks after client is initialized and safe to use.
when disconnections/redirections occured, sometimes server/client
channels were not synced, meaning that for example the gfx server
received a message, then tried to use gfx client that was already freed.
