tick/rcu: Stop allowing RCU_SOFTIRQ in idle
authorFrederic Weisbecker <frederic@kernel.org>
Tue, 8 Feb 2022 16:16:34 +0000 (17:16 +0100)
committerPaul E. McKenney <paulmck@kernel.org>
Thu, 24 Feb 2022 17:56:46 +0000 (09:56 -0800)
commit8001c4ebd4f70bf4732dd116654bdb8183d99e7b
treed46474ec1721f2e670a806935f66bf96d99b7b88
parent2253d02cf83cdc01b50a47a69e81ec525f362944
tick/rcu: Stop allowing RCU_SOFTIRQ in idle

RCU_SOFTIRQ used to be special in that it could be raised on purpose
within the idle path to prevent from stopping the tick. Some code still
prevents from unnecessary warnings related to this specific behaviour
while entering in dynticks-idle mode.

However the nohz layout has changed quite a bit in ten years, and the
removal of CONFIG_RCU_FAST_NO_HZ has been the final straw to this
safe-conduct. Now the RCU_SOFTIRQ vector is expected to be raised from
sane places.

A remaining corner case is admitted though when the vector is invoked
in fragile hotplug path.

Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
include/linux/interrupt.h
kernel/time/tick-sched.c