Optimise CQ locking for event posting depending on a number of ring setup flags.
QD1 nop benchmark showed 12.067 -> 12.565 MIOPS increase, which more than 8.5%
of the io_uring kernel overhead (taking into account that the syscall overhead
is over 50%) or 4.12% of the total performance. Naturally, it's not only about
QD1, applications can submit a bunch of requests but their completions will may
arrive randomly hurting batching and so performance (or latency).
The downside is that we have to punt all io-wq completions to the
original task. The performance win should diminish with better
completion batching, but it should be worth it for as it also helps tw,
which in reality often don't complete too many requests.
The feature depends on DEFER_TASKRUN but can be relaxed to SINGLE_ISSUER
v2: some general msg_ring fixes (patche 1,2)
fix exiting ring potentially modifying CQ in parallel (8/12)
use task_work instead of overflowing msg_ring CQEs, which could've
messed with CQE ordering (9-11)
Pavel Begunkov (12):
io_uring: dont remove file from msg_ring reqs
io_uring: improve io_double_lock_ctx fail handling
io_uring: skip overflow CQE posting for dying ring
io_uring: don't check overflow flush failures
io_uring: complete all requests in task context
io_uring: force multishot CQEs into task context
io_uring: use tw for putting rsrc
io_uring: never run tw and fallback in parallel
io_uring: get rid of double locking
io_uring: extract a io_msg_install_complete helper
io_uring: do msg_ring in target task via tw
io_uring: skip spinlocking for ->task_complete
include/linux/io_uring.h | 2 +
include/linux/io_uring_types.h | 3 +
io_uring/io_uring.c | 165 ++++++++++++++++++++++-----------
io_uring/io_uring.h | 12 ++-
io_uring/msg_ring.c | 163 ++++++++++++++++++++++----------
io_uring/msg_ring.h | 1 +
io_uring/net.c | 21 +++++
io_uring/opdef.c | 8 ++
io_uring/opdef.h | 2 +
io_uring/rsrc.c | 19 +++-
io_uring/rsrc.h | 1 +
11 files changed, 291 insertions(+), 106 deletions(-)
--
2.38.1
