s390/smp: lost IPIs on cpu hotplug (9acf73b7) · Commits · 戴 / test

IPIs might be lost when a cpu gets brought offline: When stop_machine executes its state machine there is a race window for the state STOPMACHINE_DISABLE_IRQ where the to be brought offline cpu might already have irqs disabled but a different cpu still may have irqs enabled. If the enabled cpu receives an interrupt and as a result sends an IPI to the to be offlined cpu in its bottom halve context, the IPI won't be noticed before the cpu is offline. In fact the race window is much larger since there is no guarantee when an IPI will be received. To fix this check for enqueued but not yet received IPIs in the cpu_disable() path and call the respective handlers before the cpu is marked offline. Reported-by:

Juergen Doelle <juergen.doelle@de.ibm.com> Signed-off-by:

Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by:

Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/kernel/smp.c

+11 −16

Original line number	Diff line number	Diff line
		@@ -428,34 +428,27 @@ void smp_stop_cpu(void)
		* This is the main routine where commands issued by other
		* cpus are handled.
		*/
		static void do_ext_call_interrupt(struct ext_code ext_code,
		unsigned int param32, unsigned long param64)
		static void smp_handle_ext_call(void)
		{
		unsigned long bits;
		int cpu;

		cpu = smp_processor_id();
		if (ext_code.code == 0x1202)
		inc_irq_stat(IRQEXT_EXC);
		else
		inc_irq_stat(IRQEXT_EMS);
		/*
		* handle bit signal external calls
		*/
		bits = xchg(&pcpu_devices[cpu].ec_mask, 0);

		/* handle bit signal external calls */
		bits = xchg(&pcpu_devices[smp_processor_id()].ec_mask, 0);
		if (test_bit(ec_stop_cpu, &bits))
		smp_stop_cpu();

		if (test_bit(ec_schedule, &bits))
		scheduler_ipi();

		if (test_bit(ec_call_function, &bits))
		generic_smp_call_function_interrupt();

		if (test_bit(ec_call_function_single, &bits))
		generic_smp_call_function_single_interrupt();
		}

		static void do_ext_call_interrupt(struct ext_code ext_code,
		unsigned int param32, unsigned long param64)
		{
		inc_irq_stat(ext_code.code == 0x1202 ? IRQEXT_EXC : IRQEXT_EMS);
		smp_handle_ext_call();
		}

		void arch_send_call_function_ipi_mask(const struct cpumask *mask)
		@@ -760,6 +753,8 @@ int __cpu_disable(void)
		{
		unsigned long cregs[16];

		/* Handle possible pending IPIs */
		smp_handle_ext_call();
		set_cpu_online(smp_processor_id(), false);
		/* Disable pseudo page faults on this cpu. */
		pfault_fini();

Admin message