rcu-tasks: Maintain real-time response in rcu_tasks_postscan()
authorPaul E. McKenney <paulmck@kernel.org>
Thu, 1 Feb 2024 14:10:26 +0000 (06:10 -0800)
committerBoqun Feng <boqun.feng@gmail.com>
Sun, 25 Feb 2024 22:27:49 +0000 (14:27 -0800)
commit0bb11a372fc8d7006b4d0f42a2882939747bdbff
treeb87f28feefac58a52950abe40bdf1901bc7b32bb
parent1612160b91272f5b1596f499584d6064bf5be794
rcu-tasks: Maintain real-time response in rcu_tasks_postscan()

The current code will scan the entirety of each per-CPU list of exiting
tasks in ->rtp_exit_list with interrupts disabled.  This is normally just
fine, because each CPU typically won't have very many tasks in this state.
However, if a large number of tasks block late in do_exit(), these lists
could be arbitrarily long.  Low probability, perhaps, but it really
could happen.

This commit therefore occasionally re-enables interrupts while traversing
these lists, inserting a dummy element to hold the current place in the
list.  In kernels built with CONFIG_PREEMPT_RT=y, this re-enabling happens
after each list element is processed, otherwise every one-to-two jiffies.

[ paulmck: Apply Frederic Weisbecker feedback. ]

Link: https://lore.kernel.org/all/ZdeI_-RfdLR8jlsm@localhost.localdomain/
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Sebastian Siewior <bigeasy@linutronix.de>
Cc: Anna-Maria Behnsen <anna-maria@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
kernel/rcu/tasks.h