Messages in this thread | | | Date | Tue, 30 Oct 2012 20:45:52 -0700 | From | "Paul E. McKenney" <> | Subject | [PATCH tip/core/rcu 0/2] v2 Add callback-free CPUs |
| |
Hello!
RCU callback execution can add significant OS jitter and also can degrade scheduling latency. This commit therefore adds the ability for selected CPUs ("rcu_nocbs=" boot parameter) to have their callbacks offloaded to kthreads, inspired by Joe Korty's and Jim Houston's JRCU. If the "rcu_nocb_poll" boot parameter is also specified, these kthreads will do polling, removing the need for the offloaded CPUs to do wakeups. At least one CPU must be doing normal callback processing: currently CPU 0 cannot be selected as a no-CBs CPU. In addition, attempts to offline the last normal-CBs CPU will fail.
Changes since v1 (https://lkml.org/lkml/2012/9/5/572):
1. Contains fixes for a few problems located by Fengguang Wu's kbuild test robot. 2. Counters are now atomic_long_t rather than atomic_t, as suggested by Peter Zijlstra. 3. The rcu_nocbs= kernel boot parameter is now documented. 4. Fixed a bug introduced by commit c96ea7cf (Avoid spurious RCU CPU stall warnings) that resulted in boot-time NULL-pointer dereferences (reported by Paul Gortmaker). 4. Account for normal and offloaded callbacks separately, so that offloading is represented in debugfs output.
The patches in this series are as follows:
1. Offload RCU callbacks based on boot-time kernel parameter. 2. Account for normal and offloaded callbacks separately, so that offloading is represented in debugfs output.
Thanx, Paul
b/Documentation/kernel-parameters.txt | 5 b/include/trace/events/rcu.h | 1 b/init/Kconfig | 19 + b/kernel/rcutree.c | 63 ++++- b/kernel/rcutree.h | 48 ++++ b/kernel/rcutree_plugin.h | 399 +++++++++++++++++++++++++++++++++- b/kernel/rcutree_trace.c | 12 - 7 files changed, 528 insertions(+), 19 deletions(-)
| |