LibProtocol: Exit fatally if the connection to RequestServer disappears

The default implementation of die() causes the client process to simply
exit cleanly. This prevents any tests from recognizing that something
went wrong, as the process exits with a code of 0. With this patch, we
still just exit when the connection dies, but with a fatal signal. In
the future, we will want to launch a new RequestServer process and
re-establish client connections.

Userland/Libraries/LibProtocol/RequestClient.cpp

@@ -14,6 +14,13 @@ RequestClient::RequestClient(NonnullOwnPtr<Core::LocalSocket> socket)
 {
 }
 
+void RequestClient::die()
+{
+    // FIXME: Gracefully handle this, or relaunch and reconnect to RequestServer.
+    warnln("\033[31;1mLost connection to RequestServer\033[0m");
+    VERIFY_NOT_REACHED();
+}
+
 void RequestClient::ensure_connection(URL::URL const& url, ::RequestServer::CacheLevel cache_level)
 {
     async_ensure_connection(url, cache_level);

Userland/Libraries/LibProtocol/RequestClient.h

@@ -36,6 +36,8 @@ public:
     bool set_certificate(Badge<Request>, Request&, ByteString, ByteString);
 
 private:
+    virtual void die() override;
+
     virtual void request_started(i32, IPC::File const&) override;
     virtual void request_progress(i32, Optional<u64> const&, u64) override;
     virtual void request_finished(i32, bool, u64) override;