rcu-tasks: Set ->percpu_enqueue_shift to zero upon contention
authorPaul E. McKenney <paulmck@kernel.org>
Thu, 3 Feb 2022 00:34:40 +0000 (16:34 -0800)
committerPaul E. McKenney <paulmck@kernel.org>
Tue, 8 Feb 2022 18:13:22 +0000 (10:13 -0800)
commit00a8b4b54cd69d9f7ba1730d3b266469a778b1d7
tree9ffd61b4b7a88dc9d1cb00cdea5e28d0f2dce0b1
parent2bcd18e041fc3c2ae58f41eb5e18790c7c82c674
rcu-tasks: Set ->percpu_enqueue_shift to zero upon contention

Currently, call_rcu_tasks_generic() sets ->percpu_enqueue_shift to
order_base_2(nr_cpu_ids) upon encountering sufficient contention.
This does not shift to use of non-CPU-0 callback queues as intended, but
rather continues using only CPU 0's queue.  Although this does provide
some decrease in contention due to spreading work over multiple locks,
it is not the dramatic decrease that was intended.

This commit therefore makes call_rcu_tasks_generic() set
->percpu_enqueue_shift to 0.

Reported-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
kernel/rcu/tasks.h