smpboot_64.c

linux 内核源代码

	/* allocate memory for gdts of secondary cpus. Hotplug is considered */
	if (!cpu_gdt_descr[cpu].address &&
		!(cpu_gdt_descr[cpu].address = get_zeroed_page(GFP_KERNEL))) {
		printk(KERN_ERR "Failed to allocate GDT for CPU %d\n", cpu);
		return -1;
	}

	/* Allocate node local memory for AP pdas */
	if (cpu_pda(cpu) == &boot_cpu_pda[cpu]) {
		struct x8664_pda *newpda, *pda;
		int node = cpu_to_node(cpu);
		pda = cpu_pda(cpu);
		newpda = kmalloc_node(sizeof (struct x8664_pda), GFP_ATOMIC,
				      node);
		if (newpda) {
			memcpy(newpda, pda, sizeof (struct x8664_pda));
			cpu_pda(cpu) = newpda;
		} else
			printk(KERN_ERR
		"Could not allocate node local PDA for CPU %d on node %d\n",
				cpu, node);
	}

	alternatives_smp_switch(1);

	c_idle.idle = get_idle_for_cpu(cpu);

	if (c_idle.idle) {
		c_idle.idle->thread.rsp = (unsigned long) (((struct pt_regs *)
			(THREAD_SIZE +  task_stack_page(c_idle.idle))) - 1);
		init_idle(c_idle.idle, cpu);
		goto do_rest;
	}

	/*
	 * During cold boot process, keventd thread is not spun up yet.
	 * When we do cpu hot-add, we create idle threads on the fly, we should
	 * not acquire any attributes from the calling context. Hence the clean
	 * way to create kernel_threads() is to do that from keventd().
	 * We do the current_is_keventd() due to the fact that ACPI notifier
	 * was also queuing to keventd() and when the caller is already running
	 * in context of keventd(), we would end up with locking up the keventd
	 * thread.
	 */
	if (!keventd_up() || current_is_keventd())
		c_idle.work.func(&c_idle.work);
	else {
		schedule_work(&c_idle.work);
		wait_for_completion(&c_idle.done);
	}

	if (IS_ERR(c_idle.idle)) {
		printk("failed fork for CPU %d\n", cpu);
		return PTR_ERR(c_idle.idle);
	}

	set_idle_for_cpu(cpu, c_idle.idle);

do_rest:

	cpu_pda(cpu)->pcurrent = c_idle.idle;

	start_rip = setup_trampoline();

	init_rsp = c_idle.idle->thread.rsp;
	per_cpu(init_tss,cpu).rsp0 = init_rsp;
	initial_code = start_secondary;
	clear_tsk_thread_flag(c_idle.idle, TIF_FORK);

	printk(KERN_INFO "Booting processor %d/%d APIC 0x%x\n", cpu,
		cpus_weight(cpu_present_map),
		apicid);

	/*
	 * This grunge runs the startup process for
	 * the targeted processor.
	 */

	atomic_set(&init_deasserted, 0);

	Dprintk("Setting warm reset code and vector.\n");

	CMOS_WRITE(0xa, 0xf);
	local_flush_tlb();
	Dprintk("1.\n");
	*((volatile unsigned short *) phys_to_virt(0x469)) = start_rip >> 4;
	Dprintk("2.\n");
	*((volatile unsigned short *) phys_to_virt(0x467)) = start_rip & 0xf;
	Dprintk("3.\n");

	/*
	 * Be paranoid about clearing APIC errors.
	 */
	apic_write(APIC_ESR, 0);
	apic_read(APIC_ESR);

	/*
	 * Status is now clean
	 */
	boot_error = 0;

	/*
	 * Starting actual IPI sequence...
	 */
	boot_error = wakeup_secondary_via_INIT(apicid, start_rip);

	if (!boot_error) {
		/*
		 * allow APs to start initializing.
		 */
		Dprintk("Before Callout %d.\n", cpu);
		cpu_set(cpu, cpu_callout_map);
		Dprintk("After Callout %d.\n", cpu);

		/*
		 * Wait 5s total for a response
		 */
		for (timeout = 0; timeout < 50000; timeout++) {
			if (cpu_isset(cpu, cpu_callin_map))
				break;	/* It has booted */
			udelay(100);
		}

		if (cpu_isset(cpu, cpu_callin_map)) {
			/* number CPUs logically, starting from 1 (BSP is 0) */
			Dprintk("CPU has booted.\n");
		} else {
			boot_error = 1;
			if (*((volatile unsigned char *)phys_to_virt(SMP_TRAMPOLINE_BASE))
					== 0xA5)
				/* trampoline started but...? */
				printk("Stuck ??\n");
			else
				/* trampoline code not run */
				printk("Not responding.\n");
#ifdef APIC_DEBUG
			inquire_remote_apic(apicid);
#endif
		}
	}
	if (boot_error) {
		cpu_clear(cpu, cpu_callout_map); /* was set here (do_boot_cpu()) */
		clear_bit(cpu, &cpu_initialized); /* was set by cpu_init() */
		clear_node_cpumask(cpu); /* was set by numa_add_cpu */
		cpu_clear(cpu, cpu_present_map);
		cpu_clear(cpu, cpu_possible_map);
		per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID;
		return -EIO;
	}

	return 0;
}

cycles_t cacheflush_time;
unsigned long cache_decay_ticks;

/*
 * Cleanup possible dangling ends...
 */
static __cpuinit void smp_cleanup_boot(void)
{
	/*
	 * Paranoid:  Set warm reset code and vector here back
	 * to default values.
	 */
	CMOS_WRITE(0, 0xf);

	/*
	 * Reset trampoline flag
	 */
	*((volatile int *) phys_to_virt(0x467)) = 0;
}

/*
 * Fall back to non SMP mode after errors.
 *
 * RED-PEN audit/test this more. I bet there is more state messed up here.
 */
static __init void disable_smp(void)
{
	cpu_present_map = cpumask_of_cpu(0);
	cpu_possible_map = cpumask_of_cpu(0);
	if (smp_found_config)
		phys_cpu_present_map = physid_mask_of_physid(boot_cpu_id);
	else
		phys_cpu_present_map = physid_mask_of_physid(0);
	cpu_set(0, per_cpu(cpu_sibling_map, 0));
	cpu_set(0, per_cpu(cpu_core_map, 0));
}

#ifdef CONFIG_HOTPLUG_CPU

int additional_cpus __initdata = -1;

/*
 * cpu_possible_map should be static, it cannot change as cpu's
 * are onlined, or offlined. The reason is per-cpu data-structures
 * are allocated by some modules at init time, and dont expect to
 * do this dynamically on cpu arrival/departure.
 * cpu_present_map on the other hand can change dynamically.
 * In case when cpu_hotplug is not compiled, then we resort to current
 * behaviour, which is cpu_possible == cpu_present.
 * - Ashok Raj
 *
 * Three ways to find out the number of additional hotplug CPUs:
 * - If the BIOS specified disabled CPUs in ACPI/mptables use that.
 * - The user can overwrite it with additional_cpus=NUM
 * - Otherwise don't reserve additional CPUs.
 * We do this because additional CPUs waste a lot of memory.
 * -AK
 */
__init void prefill_possible_map(void)
{
	int i;
	int possible;

 	if (additional_cpus == -1) {
 		if (disabled_cpus > 0)
 			additional_cpus = disabled_cpus;
 		else
			additional_cpus = 0;
 	}
	possible = num_processors + additional_cpus;
	if (possible > NR_CPUS) 
		possible = NR_CPUS;

	printk(KERN_INFO "SMP: Allowing %d CPUs, %d hotplug CPUs\n",
		possible,
	        max_t(int, possible - num_processors, 0));

	for (i = 0; i < possible; i++)
		cpu_set(i, cpu_possible_map);
}
#endif

/*
 * Various sanity checks.
 */
static int __init smp_sanity_check(unsigned max_cpus)
{
	if (!physid_isset(hard_smp_processor_id(), phys_cpu_present_map)) {
		printk("weird, boot CPU (#%d) not listed by the BIOS.\n",
		       hard_smp_processor_id());
		physid_set(hard_smp_processor_id(), phys_cpu_present_map);
	}

	/*
	 * If we couldn't find an SMP configuration at boot time,
	 * get out of here now!
	 */
	if (!smp_found_config) {
		printk(KERN_NOTICE "SMP motherboard not detected.\n");
		disable_smp();
		if (APIC_init_uniprocessor())
			printk(KERN_NOTICE "Local APIC not detected."
					   " Using dummy APIC emulation.\n");
		return -1;
	}

	/*
	 * Should not be necessary because the MP table should list the boot
	 * CPU too, but we do it for the sake of robustness anyway.
	 */
	if (!physid_isset(boot_cpu_id, phys_cpu_present_map)) {
		printk(KERN_NOTICE "weird, boot CPU (#%d) not listed by the BIOS.\n",
								 boot_cpu_id);
		physid_set(hard_smp_processor_id(), phys_cpu_present_map);
	}

	/*
	 * If we couldn't find a local APIC, then get out of here now!
	 */
	if (!cpu_has_apic) {
		printk(KERN_ERR "BIOS bug, local APIC #%d not detected!...\n",
			boot_cpu_id);
		printk(KERN_ERR "... forcing use of dummy APIC emulation. (tell your hw vendor)\n");
		nr_ioapics = 0;
		return -1;
	}

	/*
	 * If SMP should be disabled, then really disable it!
	 */
	if (!max_cpus) {
		printk(KERN_INFO "SMP mode deactivated, forcing use of dummy APIC emulation.\n");
		nr_ioapics = 0;
		return -1;
	}

	return 0;
}

/*
 * Copy apicid's found by MP_processor_info from initial array to the per cpu
 * data area.  The x86_cpu_to_apicid_init array is then expendable and the
 * x86_cpu_to_apicid_ptr is zeroed indicating that the static array is no
 * longer available.
 */
void __init smp_set_apicids(void)
{
	int cpu;

	for_each_cpu_mask(cpu, cpu_possible_map) {
		if (per_cpu_offset(cpu))
			per_cpu(x86_cpu_to_apicid, cpu) =
						x86_cpu_to_apicid_init[cpu];
	}

	/* indicate the static array will be going away soon */
	x86_cpu_to_apicid_ptr = NULL;
}

/*
 * Prepare for SMP bootup.  The MP table or ACPI has been read
 * earlier.  Just do some sanity checking here and enable APIC mode.
 */
void __init smp_prepare_cpus(unsigned int max_cpus)
{
	nmi_watchdog_default();
	current_cpu_data = boot_cpu_data;
	current_thread_info()->cpu = 0;  /* needed? */
	smp_set_apicids();
	set_cpu_sibling_map(0);

	if (smp_sanity_check(max_cpus) < 0) {
		printk(KERN_INFO "SMP disabled\n");
		disable_smp();
		return;
	}


	/*
	 * Switch from PIC to APIC mode.
	 */
	setup_local_APIC();

	if (GET_APIC_ID(apic_read(APIC_ID)) != boot_cpu_id) {
		panic("Boot APIC ID in local APIC unexpected (%d vs %d)",
		      GET_APIC_ID(apic_read(APIC_ID)), boot_cpu_id);
		/* Or can we switch back to PIC here? */
	}

	/*
	 * Now start the IO-APICs
	 */
	if (!skip_ioapic_setup && nr_ioapics)
		setup_IO_APIC();
	else
		nr_ioapics = 0;

	/*
	 * Set up local APIC timer on boot CPU.
	 */

	setup_boot_APIC_clock();
}

/*
 * Early setup to make printk work.
 */
void __init smp_prepare_boot_cpu(void)
{
	int me = smp_processor_id();
	cpu_set(me, cpu_online_map);
	cpu_set(me, cpu_callout_map);
	per_cpu(cpu_state, me) = CPU_ONLINE;
}

/*
 * Entry point to boot a CPU.
 */
int __cpuinit __cpu_up(unsigned int cpu)
{
	int apicid = cpu_present_to_apicid(cpu);
	unsigned long flags;
	int err;

	WARN_ON(irqs_disabled());

	Dprintk("++++++++++++++++++++=_---CPU UP  %u\n", cpu);

	if (apicid == BAD_APICID || apicid == boot_cpu_id ||
	    !physid_isset(apicid, phys_cpu_present_map)) {
		printk("__cpu_up: bad cpu %d\n", cpu);
		return -EINVAL;
	}

	/*
	 * Already booted CPU?
	 */
 	if (cpu_isset(cpu, cpu_callin_map)) {
		Dprintk("do_boot_cpu %d Already started\n", cpu);
 		return -ENOSYS;
	}

	/*
	 * Save current MTRR state in case it was changed since early boot
	 * (e.g. by the ACPI SMI) to initialize new CPUs with MTRRs in sync:
	 */
	mtrr_save_state();

	per_cpu(cpu_state, cpu) = CPU_UP_PREPARE;
	/* Boot it! */
	err = do_boot_cpu(cpu, apicid);
	if (err < 0) {
		Dprintk("do_boot_cpu failed %d\n", err);
		return err;
	}

	/* Unleash the CPU! */
	Dprintk("waiting for cpu %d\n", cpu);

	/*
  	 * Make sure and check TSC sync:
 	 */
	local_irq_save(flags);
	check_tsc_sync_source(cpu);
	local_irq_restore(flags);

	while (!cpu_isset(cpu, cpu_online_map))
		cpu_relax();
	err = 0;

	return err;
}

/*
 * Finish the SMP boot.
 */
void __init smp_cpus_done(unsigned int max_cpus)
{
	smp_cleanup_boot();
	setup_ioapic_dest();
	check_nmi_watchdog();
}

#ifdef CONFIG_HOTPLUG_CPU

static void remove_siblinginfo(int cpu)
{
	int sibling;
	struct cpuinfo_x86 *c = &cpu_data(cpu);

	for_each_cpu_mask(sibling, per_cpu(cpu_core_map, cpu)) {
		cpu_clear(cpu, per_cpu(cpu_core_map, sibling));
		/*
		 * last thread sibling in this cpu core going down
		 */
		if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1)
			cpu_data(sibling).booted_cores--;
	}
			
	for_each_cpu_mask(sibling, per_cpu(cpu_sibling_map, cpu))
		cpu_clear(cpu, per_cpu(cpu_sibling_map, sibling));
	cpus_clear(per_cpu(cpu_sibling_map, cpu));
	cpus_clear(per_cpu(cpu_core_map, cpu));
	c->phys_proc_id = 0;
	c->cpu_core_id = 0;
	cpu_clear(cpu, cpu_sibling_setup_map);
}

void remove_cpu_from_maps(void)
{
	int cpu = smp_processor_id();

	cpu_clear(cpu, cpu_callout_map);
	cpu_clear(cpu, cpu_callin_map);
	clear_bit(cpu, &cpu_initialized); /* was set by cpu_init() */
	clear_node_cpumask(cpu);
}

int __cpu_disable(void)
{
	int cpu = smp_processor_id();

	/*
	 * Perhaps use cpufreq to drop frequency, but that could go
	 * into generic code.
 	 *
	 * We won't take down the boot processor on i386 due to some
	 * interrupts only being able to be serviced by the BSP.
	 * Especially so if we're not using an IOAPIC	-zwane
	 */
	if (cpu == 0)
		return -EBUSY;

	if (nmi_watchdog == NMI_LOCAL_APIC)
		stop_apic_nmi_watchdog(NULL);
	clear_local_APIC();

	/*
	 * HACK:
	 * Allow any queued timer interrupts to get serviced
	 * This is only a temporary solution until we cleanup
	 * fixup_irqs as we do for IA64.
	 */
	local_irq_enable();
	mdelay(1);

	local_irq_disable();
	remove_siblinginfo(cpu);

	spin_lock(&vector_lock);
	/* It's now safe to remove this processor from the online map */
	cpu_clear(cpu, cpu_online_map);
	spin_unlock(&vector_lock);
	remove_cpu_from_maps();
	fixup_irqs(cpu_online_map);
	return 0;
}

void __cpu_die(unsigned int cpu)
{
	/* We don't do anything here: idle task is faking death itself. */
	unsigned int i;

	for (i = 0; i < 10; i++) {
		/* They ack this in play_dead by setting CPU_DEAD */
		if (per_cpu(cpu_state, cpu) == CPU_DEAD) {
			printk ("CPU %d is now offline\n", cpu);
			if (1 == num_online_cpus())
				alternatives_smp_switch(0);
			return;
		}
		msleep(100);
	}
 	printk(KERN_ERR "CPU %u didn't die...\n", cpu);
}

static __init int setup_additional_cpus(char *s)
{
	return s && get_option(&s, &additional_cpus) ? 0 : -EINVAL;
}
early_param("additional_cpus", setup_additional_cpus);

#else /* ... !CONFIG_HOTPLUG_CPU */

int __cpu_disable(void)
{
	return -ENOSYS;
}

void __cpu_die(unsigned int cpu)
{
	/* We said "no" in __cpu_disable */
	BUG();
}
#endif /* CONFIG_HOTPLUG_CPU */