• uses an event loop created with io_uring
• uses liburing HEAD https://github.com/axboe/liburing
• Linux 5.7 or higher with IORING_FEAT_FAST_POLL and IORING_OP_PROVIDE_BUFFERS required

make liburing

make io_uring_echo_server

./io_uring_echo_server [port_number]

https://github.com/frevib/epoll-echo-server

https://github.com/frevib/io_uring-echo-server/blob/io-uring-feat-fast-poll/benchmarks/benchmarks.md

• Use IORING_OP_PROVIDE_BUFFERS

• Use IORING_FEAT_FAST_POLL, which increases performance a lot
• Remove all polling, this is now handled by IORING_FEAT_FAST_POLL

Fix bug that massively overstated the performance.

Use pre-allocated sqe->user_data instead of dynamically allocating memory.

Fix memory leak.

Working release.