News for hackers. Stuff that matters.
Re:Lord of the io_uring: io_uring tutorial, examples and ref... by tyingq on Sunday May 10, 2020 @11:41AM (#23133469)
If I understand the premise right, it should be fewer syscalls per IO. So even if it doesn't improve disk I/O, it might reduce CPU utilization.
Re:Lord of the io_uring: io_uring tutorial, examples and ref... by matheusmoreira on Sunday May 10, 2020 @12:03PM (#23133635)
It could also be nearly zero system calls per I/O operation. The kernel can poll the submission queue for new entries. This eliminates system call overhead at the cost of higher CPU utilization.
Re:Lord of the io_uring: io_uring tutorial, examples and ref... by jandrewrogers on Sunday May 10, 2020 @01:22PM (#23134297)
This is true, but for most intentionally optimized storage engines that syscall overhead is below the noise floor in practice, even on NVMe storage. A single core can easily drive gigabytes per second using the old Linux AIO interface.

It appears to primarily be an optimization for storage that was not well-optimized to begin with. It is not obvious that it would make high-performance storage engines faster.