Using shared rings of buffers is almost but not quite what io_uring provides? It looks like it takes an io_uring style approach and natively so unless I’m misre

Using shared rings of buffers is almost but not quite what io_uring provides? It looks like it takes an io_uring style approach and natively so unless I’m misreading? Have the performance numbers somewhere been validated?