cpu: No more __stop_machine() in _cpu_down()
RFC PATCH 1/1 cpu: No more __stop_machine() in _cpu_down()
The _cpu_down() function invoked as part of the CPU-hotplug offlining process currently invokes __stop_machine(), which is slow and inflicts substantial real-time latencies on the entire system. This patch substitutes stop_cpus() for __stop_machine() in order to improve both performance and real-time latency.
This is currently unsafe, because there are a number of uses of preempt_disable() that are intended to block CPU-hotplug offlining. These will be fixed, but in the meantime, this commit is one way to help locate them. It nevertheless passes light rcutorture/hotplug stress testing. Meaning that we should not be relying on pure testing to find places where people are relying on preemption disabling to block CPUs from going offline. ;-)
kernel/cpu.c +7 -2
@@ -243,13 +243,18 @@ static int __ref take_cpu_down(void *_param) {struct take_cpu_down_param *param = _param;int err; +unsigned long flags;/* Ensure this CPU doesn't handle any more interrupts. */ +local_irq_save (flags); err =__cpu_disable (); -if (err <0 ) +if (err <0 ) { +local_irq_restore (flags);return err; + }cpu_notify (CPU_DYING | param->mod, param->hcpu); +local_irq_restore (flags);return 0 ; }
@@ -281,7 +286,7 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) goto out_release; } - err =__stop_machine (take_cpu_down, &tcd_param,cpumask_of (cpu)); + err =stop_cpus (cpumask_of (cpu), take_cpu_down, &tcd_param);if (err) {/* CPU didn't die: tell everyone. Can't complain. */ cpu_notify_nofail (CPU_DOWN_FAILED | mod, hcpu);