This repository has been archived on 2026-03-14. You can view files and clone it, but cannot push or open issues or pull requests.
Files
ubports_kernel_google_msm/include/linux
Paul E. McKenney 25502a6c13 rcu: refactor RCU's context-switch handling
The addition of preemptible RCU to treercu resulted in a bit of
confusion and inefficiency surrounding the handling of context switches
for RCU-sched and for RCU-preempt.  For RCU-sched, a context switch
is a quiescent state, pure and simple, just like it always has been.
For RCU-preempt, a context switch is in no way a quiescent state, but
special handling is required when a task blocks in an RCU read-side
critical section.

However, the callout from the scheduler and the outer loop in ksoftirqd
still calls something named rcu_sched_qs(), whose name is no longer
accurate.  Furthermore, when rcu_check_callbacks() notes an RCU-sched
quiescent state, it ends up unnecessarily (though harmlessly, aside
from the performance hit) enqueuing the current task if it happens to
be running in an RCU-preempt read-side critical section.  This not only
increases the maximum latency of scheduler_tick(), it also needlessly
increases the overhead of the next outermost rcu_read_unlock() invocation.

This patch addresses this situation by separating the notion of RCU's
context-switch handling from that of RCU-sched's quiescent states.
The context-switch handling is covered by rcu_note_context_switch() in
general and by rcu_preempt_note_context_switch() for preemptible RCU.
This permits rcu_sched_qs() to handle quiescent states and only quiescent
states.  It also reduces the maximum latency of scheduler_tick(), though
probably by much less than a microsecond.  Finally, it means that tasks
within preemptible-RCU read-side critical sections avoid incurring the
overhead of queuing unless there really is a context switch.

Suggested-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Acked-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <peterz@infradead.org>
2010-05-10 11:08:33 -07:00
..
2010-03-12 15:53:10 -08:00
2009-12-16 07:20:13 -08:00
2010-04-25 08:54:42 +02:00
2010-03-12 15:52:32 -08:00
2010-02-05 07:35:05 -08:00
2009-12-15 08:53:33 -08:00
2009-12-23 13:33:54 +01:00
2010-04-23 10:39:48 -07:00
2010-03-16 08:55:32 +01:00
2010-02-09 11:13:56 +01:00
2010-03-07 22:17:09 +01:00
2010-03-02 12:23:42 +01:00
2010-04-01 01:31:13 -07:00
2010-02-18 14:08:38 -08:00
2010-02-17 16:35:17 -08:00
2010-02-02 07:32:29 -08:00
2010-01-11 16:28:01 -08:00
2010-03-12 15:52:40 -08:00
2010-03-12 15:52:36 -08:00
2010-03-12 15:52:40 -08:00
2010-03-12 15:53:10 -08:00
2009-12-16 06:56:12 -08:00
2010-04-06 10:55:33 -04:00
2009-12-15 08:53:36 -08:00
2009-12-26 20:40:34 -08:00
2010-02-03 17:39:50 +11:00
2010-03-12 15:53:10 -08:00
2010-02-09 11:13:56 +01:00
2010-02-19 03:35:12 -05:00
2010-03-12 15:52:38 -08:00
2010-03-12 15:52:28 -08:00
2010-04-07 08:38:04 -07:00
2010-02-10 23:49:08 +09:00
2010-02-03 20:32:28 -08:00
2010-04-12 07:55:15 -04:00
2010-02-19 15:52:40 -05:00
2010-02-14 07:13:47 -07:00
2010-02-23 23:14:47 -08:00
2010-02-04 10:23:02 -08:00
2010-03-12 15:53:11 -08:00
2010-03-05 00:20:31 +01:00
2010-02-10 17:47:17 -08:00
2010-03-12 15:52:44 -08:00
2010-03-02 14:28:49 -05:00
2010-03-12 15:53:10 -08:00
2010-03-12 15:52:43 -08:00
2010-01-21 23:55:25 -08:00
2010-02-16 13:38:56 +09:00
2010-03-12 15:52:44 -08:00
2009-12-16 22:32:29 -05:00
2010-01-14 22:38:09 -05:00
2010-03-12 15:52:36 -08:00
2010-02-18 15:43:09 -08:00
2010-01-05 09:17:33 +09:00
2010-03-12 10:03:42 +01:00
2009-12-16 12:16:49 -05:00