This is a respin of the "Address rcutorture issues" patchset, minus the actual rcutorture changes.
I still plan to implement detection of bad nesting scenarios, but it's complicated by the need to distinguish (on a non-RT kernel) between irq/preempt disabling that would and would not happen on an RT kernel (which would also have the benefit of being able to detect nesting regular spinlocks inside raw spinlocks on a non-RT debug kernel). In the meantime I could send the rcutorture changes as a PREEMPT_RT only patch, though the extent of the changes depends on whether my migrate disable patchset is applied since it removes a restriction. Scott Wood (3): rcu: Acquire RCU lock when disabling BHs sched: migrate_enable: Use sleeping_lock to indicate involuntary sleep rcu: Disable use_softirq on PREEMPT_RT include/linux/rcupdate.h | 4 ++++ include/linux/sched.h | 4 ++-- kernel/rcu/tree.c | 9 ++++++++- kernel/rcu/tree_plugin.h | 2 +- kernel/rcu/update.c | 4 ++++ kernel/sched/core.c | 8 ++++++++ kernel/softirq.c | 12 +++++++++--- 7 files changed, 36 insertions(+), 7 deletions(-) -- 1.8.3.1
