[-next,0/6] epoll: some miscellaneous optimizations

Message

Davidlohr Bueso Nov. 8, 2018, 5:10 a.m. UTC

Hi,

The following are some incremental optimizations on some of the epoll
core. Each patch has the details, but together, the series is seen
to shave off measurable cycles on a number of systems and workloads.

For example, on a 40-core IB, a pipetest as well as parallel epoll_wait()
benchmark show around a 20-30% increase in raw operations per second when
the box is fully occupied (incremental thread counts), and up to 15%
performance improvement with lower counts.

Passes ltp epoll related testcases. Please consider for v4.21/5.0.

Thanks!

Davidlohr Bueso (6):
  fs/epoll: remove max_nests argument from ep_call_nested()
  fs/epoll: simplify ep_send_events_proc() ready-list loop
  fs/epoll: drop ovflist branch prediction
  fs/epoll: robustify ep->mtx held checks
  fs/epoll: reduce the scope of wq lock in epoll_wait()
  fs/epoll: avoid barrier after an epoll_wait(2) timeout

 fs/eventpoll.c | 206 ++++++++++++++++++++++++++++++---------------------------
 1 file changed, 108 insertions(+), 98 deletions(-)