Tue, 13 Jul 2021 15:04:34 +0200 mod_s2s: Remove connection timeout once it's no longer needed
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 15:04:34 +0200] rev 11682
mod_s2s: Remove connection timeout once it's no longer needed Reduces the number of left-over timers to handle after many s2s connections were started, leaving only the ones related to incomplete connections.
Tue, 13 Jul 2021 14:58:50 +0200 net.server_epoll: Call onconnect immediately after TLS handshake completion
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:58:50 +0200] rev 11681
net.server_epoll: Call onconnect immediately after TLS handshake completion Skips a roundtrip through the main loop in case client-first data is available already, if not then :onreadable() will set the appropriate timeout.
Tue, 13 Jul 2021 14:55:21 +0200 net.server_epoll: Refactor immediate TLS handshake start
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:55:21 +0200] rev 11680
net.server_epoll: Refactor immediate TLS handshake start
Tue, 13 Jul 2021 14:51:05 +0200 net.server_epoll: Keep socket registered in epoll trough TLS wrapping
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:51:05 +0200] rev 11679
net.server_epoll: Keep socket registered in epoll trough TLS wrapping There's the theory that the socket isn't the same before/after wrap(), but since epoll operates on FD numbers this shouldn't matter.
Tue, 13 Jul 2021 14:27:46 +0200 net.server_epoll: Use TLS handshake timeout after initiating handshake
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:27:46 +0200] rev 11678
net.server_epoll: Use TLS handshake timeout after initiating handshake The :init() method sets a different timeout than the TLS related methods.
Tue, 13 Jul 2021 14:20:26 +0200 net.server_epoll: Start TLS handshake immediately on newly accepted connections
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:20:26 +0200] rev 11677
net.server_epoll: Start TLS handshake immediately on newly accepted connections Since TLS is a client-first protocol there is a chance that the ClientHello message is available already. TLS Fast Open and/or the TCP_DEFER_ACCEPT socket option would increase that chance.
Tue, 13 Jul 2021 14:20:24 +0200 net.server_epoll: Factor out TLS initialization into a method
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 14:20:24 +0200] rev 11676
net.server_epoll: Factor out TLS initialization into a method So there's :startls(), :inittls() and :tlshandshake() :starttls() prepares for plain -> TLS upgrade and ensures that the (unencrypted) write buffer is drained before proceeding. :inittls() wraps the connection and does things like SNI, DANE etc. :tlshandshake() steps the TLS negotiation forward until it completes
Tue, 13 Jul 2021 02:05:35 +0200 net.server_epoll: Fix typo
Kim Alvefur <zash@zash.se> [Tue, 13 Jul 2021 02:05:35 +0200] rev 11675
net.server_epoll: Fix typo
Sun, 11 Jul 2021 12:41:32 +0200 mod_s2s: Log debug message on attempted close of an connectionless session
Kim Alvefur <zash@zash.se> [Sun, 11 Jul 2021 12:41:32 +0200] rev 11674
mod_s2s: Log debug message on attempted close of an connectionless session This should probably never happen, but probably does anyways. A debug message would show the truth of the matter.
Sun, 11 Jul 2021 12:37:51 +0200 mod_s2s: Drop level of indentation by inverting a condition and early return
Kim Alvefur <zash@zash.se> [Sun, 11 Jul 2021 12:37:51 +0200] rev 11673
mod_s2s: Drop level of indentation by inverting a condition and early return Nicer to get rid of a conditional that covers such a large portion of a pretty big function.
(0) -10000 -3000 -1000 -300 -100 -10 +10 +100 +300 +1000 tip