- 05 Feb, 2019 2 commits
-
-
Simon Giesecke authored
Solution: use tcp_open_socket function
-
Simon Giesecke authored
Problem: tcp_connecter_t and tcp_listener_t duplicate code around opening and configuring a TCP socket Solution: extract common parts into tcp_open_socket function
-
- 04 Feb, 2019 2 commits
-
-
Simon Giesecke authored
Solution: add const
-
Simon Giesecke authored
Solution: let socks_connecter_t derive from stream_connecter_base_t and remove duplicate code
-
- 03 Feb, 2019 1 commit
-
-
Luca Boccassi authored
Solution: change back the indentifier endpoint to the one passed by the user rather than the resolved one, otherwise when the user passes the same string to the disconnect call they do not match anymore
-
- 02 Feb, 2019 9 commits
-
-
Simon Giesecke authored
Solution: fix address passed and refactor set_local_address to remove code duplication and unnecessary address parsing when ZMQ_USE_FD is used
-
Simon Giesecke authored
Problem: stream listeners do not provide correct local and remote address information (test_monitor is failing) Solution: query local and remote addresses on accepting a connection
-
Simon Giesecke authored
Solution: query the local address and set it
-
Simon Giesecke authored
Solution: pass the local address by subclasses to create_engine
-
Simon Giesecke authored
Problem: get_socket_address and get_socket_name not available throughout libzmq and restricted to local address Solution: move to address.hpp/.cpp and generalize
-
Simon Giesecke authored
Solution: add monitor event socket option
-
Simon Giesecke authored
Problem: socks_connecter_t, vmci_connecter_t and vmci_listener_t duplicate code with stream_*_base_t Solution: add TODO comments to resolve this debt
-
Simon Giesecke authored
Solution: change to refer to the context
-
Simon Giesecke authored
Solution: pull up to base class
-
- 01 Feb, 2019 25 commits
-
-
Simon Giesecke authored
Solution: also use get_socket_address from base class
-
Simon Giesecke authored
Solution: also use get_socket_address from base class
-
Simon Giesecke authored
Solution: perform explicit cast
-
Simon Giesecke authored
Solution: extract function create_engine into base class
-
Simon Giesecke authored
Solution: extract common base class stream_listener_base_t
-
Simon Giesecke authored
Solution: make them private and adapt initialization order
-
Simon Giesecke authored
Solution: removed definitions and uses in subclasses
-
Simon Giesecke authored
Solution: add underscore prefix
-
Simon Giesecke authored
Solution: add underscore prefix
-
Simon Giesecke authored
Solution: pull up common code, introduce new create_engine function in base class
-
Simon Giesecke authored
Solution: pull up
-
Simon Giesecke authored
Solution: pull up tcp_connecter_t::rm_handle and use in all subclasses
-
Simon Giesecke authored
Problem: process_plug, get_new_reconnect_ivl and add_reconnect_timer duplicated across subclasses of stream_connector_base_t Solution: pull up to stream_connector_base_t
-
Simon Giesecke authored
Solution: pull up to stream_connecter_base_t
-
Simon Giesecke authored
Solution: extract common base class stream_connecter_base_t
-
Simon Giesecke authored
Solution: replace by checks against valid value of _handle
-
Simon Giesecke authored
Solution: replace by checks against valid value of _handle
-
Simon Giesecke authored
Solution: rename to _reconnect_timer_started
-
Simon Giesecke authored
Solution: rename to _reconnect_timer_started
-
Simon Giesecke authored
Solution: add underscore prefix to data members
-
Simon Giesecke authored
Solution: add underscore prefix to data members
-
Simon Giesecke authored
Solution: declare _endpoint const
-
Simon Giesecke authored
Solution: extract initialization code into get_peer_address function and declare _peer_address const
-
Simon Giesecke authored
Solution: assign result to address Fixes #3376
-
Simon Giesecke authored
Solution: remove them
-
- 19 Jan, 2019 1 commit
-
-
Luca Boccassi authored
Solution: prefer CXX11 atomics if they are available to compiler intrinsics. test_hwm_pubsub fails 50% of the times on PPC64 and RISC-V with an apparent memory corruption of messages sent by the application thread to the I/O thread when using compiler intrinsics. Switching to CXX11 atomics makes the test reliable again. The standard API should be preferred anyway, if available.
-