opentracker/ot_udp.c, branch v2.0.0An open and free bittorrent tracker
https://erdgeist.org/gitweb/opentracker/atom?h=v2.0.02026-11-02T20:14:02ZMove the build sources into src/2026-11-02T20:14:02ZDirk Englingerdgeist@erdgeist.org2026-11-02T19:50:24Zurn:sha1:ad541c67761b8aa6da1566802c347cac4018d0c2Sweep the obsolete CVS Id keyword out of the tree2026-11-02T15:55:53ZDirk Englingerdgeist@erdgeist.org2026-11-02T15:55:53Zurn:sha1:28d6bf03e1bb8c38fb7e09ade3c5a9c58d68182eGeneralize the worker setup/runloop/teardown logic by adding a smol ot_worker helper2026-10-31T23:53:53ZDirk Englingerdgeist@erdgeist.org2026-10-31T23:53:53Zurn:sha1:af0f2cb1be5aa10bb6a4bdd9f7ef4aa3950ecb31Make tcp worker multithreaded, too2026-10-31T14:59:41ZDirk Englingerdgeist@erdgeist.org2026-10-31T14:59:41Zurn:sha1:4d030871c004c917458b5e1c296edcfc689a2e66
This shamelessly steals the model established in the UDP workers, i.e.
if the kernel is able to bind multiple listening sockets to the same
endpoint, we will share them across threads. If it can't, all threads
will accept on the same socket.
The fast path handles most of the connections, only the rare case of
no data yet on the socket is offloaded to the slow path (which used to
be the mainloop) for completion. This scales nicely, as the higher the
load is, the higher chances are that an accept happens only after the
network roundtrip to the peer completes, leading to more connections
that already have data waiting for us.
We also consolidate the workstruct init and deinit code, preparation
for later unifying all thread state structs for all protocols (mca,
udc and tcp).
The config file gains a per tcp connection worker count and some docs
explaining what this knob buys and costs, nudging the user to either
keep it 0 or dial it up to at least 2.
Revert last wake-up experiment. Having each udp worker wake up once a second to check if the tracker stops is wasteful2026-10-28T23:53:50ZDirk Englingerdgeist@erdgeist.org2026-10-28T23:53:50Zurn:sha1:0f95c7d1056cab5bd8c6334cf601349ac4631457With more workers blockingly waiting on udp sockets, we need to keep track of all sockets to shut them down on exit2026-10-28T22:31:07ZDirk Englingerdgeist@erdgeist.org2026-10-28T22:31:07Zurn:sha1:772873104658182a5b76f5c0eed76047c62c6651Performance improvement, where it works: Let different udp worker threads have their own socket2026-10-28T16:08:17ZDirk Englingerdgeist@erdgeist.org2026-10-28T16:08:17Zurn:sha1:f3f640b6bcf10ba463fc81cec97cfa212c520cfePlug in the new per-thread stats counter, also make sure that the ot_clean sleep does not stall tracker exit2026-10-27T20:32:24ZDirk Englingerdgeist@erdgeist.org2026-10-27T20:32:24Zurn:sha1:6cbd21542cf7a751e6f0f38864e7e9eae8fa3b14All OSes now do have a cheap monotonic clock, so no caching thread is needed. Also add some relative time helpers2026-10-27T18:15:22ZDirk Englingerdgeist@erdgeist.org2026-10-27T18:15:22Zurn:sha1:8fb1c09df1540aa3fca1148bed6a08ffc3090669Send udp error packet in case of blacklisted torrent2026-10-27T02:20:19ZDirk Englingerdgeist@erdgeist.org2026-10-27T02:20:19Zurn:sha1:6251f74a8d37fa896c8ad78b478ce7ae5f0273ae