Daniel Verkamp 52bbb267d8 event: process events in batches ... Since we are usually going to be removing multiple events from the queue at once, use the DPDK burst dequeue interface to improve efficiency. Also rework the event queue runner to always process a fixed maximum number of events per timeslice for simplicity. This removes the rte_ring_count() call from the hot path and improves fairness between events and pollers. Now that events are dequeued in bulk, we can also put the event objects back into the mempool in bulk. Add an env wrapper around rte_mempool_put_bulk() and use it to free all of the events at once. Basic performance benchmark using test/lib/event/event/event -t 10 is improved: previously ~40 million events per second, now ~46 million events per second. Change-Id: I432e8a48774a087eec2be3a64c38c339608af42a Signed-off-by: Daniel Verkamp <daniel.verkamp@intel.com>		2016-11-07 09:32:25 -07:00
..
cpp_headers	test/cpp_headers: autogenerate header test files	2016-10-03 10:24:18 -07:00
iscsi_tgt	bdev: make construct RPCs return the bdev names	2016-10-31 11:17:46 -07:00
lib	event: process events in batches	2016-11-07 09:32:25 -07:00
nvmf	test: Use lspci to discover default kernel drivers	2016-11-02 16:37:40 -07:00
Makefile	test: validate that all headers compile as C++	2016-09-21 15:27:08 -07:00
spdk_cunit.h	spdk_cunit: only evaluate assert condition once	2016-10-19 08:09:45 -07:00