Hi, iomap always punts async dio write completions to a workqueue, which has a cost in terms of efficiency (now you need an unrelated worker to process it) and latency (now you're bouncing a completion through an async worker, which is a classic slowdown scenario). This patchset intends to improve that situation. For polled IO, we always have a task reaping completions. Those do, by definition, not need to be punted through a workqueue. This is patch 1, and it adds an IOMAP_DIO_INLINE_COMP flag that tells the completion side that we can handle this dio completion without punting to a workqueue, if we're called from the appropriate (task) context. This is good for up to an 11% improvement in my testing. Details in that patch commit message. For IRQ driven IO, it's a bit more tricky. The iomap dio completion will happen in hard/soft irq context, and we need a saner context to process these completions. IOCB_DIO_DEFER is added, which can be set in a struct kiocb->ki_flags by the issuer. If the completion side of the iocb handling understands this flag, it can choose to set a kiocb->dio_complete() handler and just call ki_complete from IRQ context. The issuer must then ensure that this callback is processed from a task. io_uring punts IRQ completions to task_work already, so it's trivial wire it up to run more of the completion before posting a CQE. Patches 2 and 3 add the necessary flag and io_uring support, and patches 4 and 5 add iomap support for it. This is good for up to a 37% improvement in throughput/latency for low queue depth IO, patch 5 has the details. This work came about when Andres tested low queue depth dio writes for postgres and compared it to doing sync dio writes, showing that the async processing slows us down a lot. fs/iomap/direct-io.c | 44 +++++++++++++++++++++++++++++++++++++------- include/linux/fs.h | 30 ++++++++++++++++++++++++++++-- io_uring/rw.c | 24 ++++++++++++++++++++---- 3 files changed, 85 insertions(+), 13 deletions(-) Can also be found in a git branch here: https://git.kernel.dk/cgit/linux/log/?h=xfs-async-dio.2 Changelog: - Rewrite patch 1 to add an explicit flag to manage when dio completions can be done inline. This drops any write related checks. We set this flag by default for both reads and writes, and clear it for the latter if we need zero out or O_DSYNC handling. -- Jens Axboe
