RCU fix for v6.12
These commits fix rcuog kthread wakeup invocation on a CPU
which has been marked offline. This can happen if there are
new callbacks enqueued (for example from an interrupt) on a
CPU before it calls rcutree_report_cpu_dead(). When this
happens on NOCB configuration, the rcuog wake-up is deferred
through an IPI to an online CPU. This is done to avoid call
into the scheduler which can risk arming the RT-bandwidth after
hrtimers have been migrated out and disabled. However, this
IPI call is done from IRQ context, which is buggy. Fix this
by adding a deferred wakeup mode, which waits for
rcu_report_cpu_dead() to be called in order to flush any
pending rcuog wake up.
rcu/nocb: Fix rcuog wake-up from offline softirq
After a CPU has set itself offline and before it eventually calls
rcutree_report_cpu_dead(), there are still opportunities for callbacks
to be enqueued, for example from an IRQ. When that happens on NOCB, the
rcuog wake-up is deferred through an IPI to an online CPU in order not
to call into the scheduler and risk arming the RT-bandwidth after
hrtimers have been migrated out and disabled.
But performing a synchronized IPI from an IRQ is buggy as reported in
the following scenario:
WARNING: CPU: 1 PID: 26 at kernel/smp.c:633 smp_call_function_single
Modules linked in: rcutorture torture
CPU: 1 UID: 0 PID: 26 Comm: migration/1 Not tainted 6.11.0-rc1-00012-g9139f93209d1 #1
Stopper: multi_cpu_stop+0x0/0x320 <- __stop_cpus+0xd0/0x120
RIP: 0010:smp_call_function_single
<IRQ>
swake_up_one_online
__call_rcu_nocb_wake
__call_rcu_common
? rcu_torture_one_read
call_timer_fn
__run_timers
run_timer_softirq
handle_softirqs
irq_exit_rcu
? tick_handle_periodic
sysvec_apic_timer_interrupt
</IRQ>
The periodic tick must be shutdown when the CPU is offline, just like is
done for oneshot tick. This must be fixed but this is not enough:
softirqs can happen on any hardirq tail and reproduce the above scenario.
Fix this with introducing a special deferred rcuog wake up mode when the
CPU is offline. This deferred wake up doesn't arm any timer and simply
wait for rcu_report_cpu_dead() to be called in order to flush any
pending rcuog wake up.
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202409231644.4c55582d-lkp@intel.com
Fixes: 9139f93209d1 ("rcu/nocb: Fix RT throttling hrtimer armed from offline CPU")
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
2 files changed
