smp.c

linux 内核源代码

/*
 *  arch/s390/kernel/smp.c
 *
 *    Copyright IBM Corp. 1999,2007
 *    Author(s): Denis Joseph Barrow ([e-mail],[e-mail]),
 *		 Martin Schwidefsky ([e-mail])
 *		 Heiko Carstens ([e-mail])
 *
 *  based on other smp stuff by
 *    (c) 1995 Alan Cox, CymruNET Ltd  <[e-mail]>
 *    (c) 1998 Ingo Molnar
 *
 * We work with logical cpu numbering everywhere we can. The only
 * functions using the real cpu address (got from STAP) are the sigp
 * functions. For all other functions we use the identity mapping.
 * That means that cpu_number_map[i] == i for every cpu. cpu_number_map is
 * used e.g. to find the idle task belonging to a logical cpu. Every array
 * in the kernel is sorted by the logical cpu number and not by the physical
 * one which is causing all the confusion with __cpu_logical_map and
 * cpu_number_map in other architectures.
 */

#include <linux/module.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/err.h>
#include <linux/spinlock.h>
#include <linux/kernel_stat.h>
#include <linux/delay.h>
#include <linux/cache.h>
#include <linux/interrupt.h>
#include <linux/cpu.h>
#include <linux/timex.h>
#include <linux/bootmem.h>
#include <asm/ipl.h>
#include <asm/setup.h>
#include <asm/sigp.h>
#include <asm/pgalloc.h>
#include <asm/irq.h>
#include <asm/s390_ext.h>
#include <asm/cpcmd.h>
#include <asm/tlbflush.h>
#include <asm/timer.h>
#include <asm/lowcore.h>
#include <asm/cpu.h>

/*
 * An array with a pointer the lowcore of every CPU.
 */
struct _lowcore *lowcore_ptr[NR_CPUS];
EXPORT_SYMBOL(lowcore_ptr);

cpumask_t cpu_online_map = CPU_MASK_NONE;
EXPORT_SYMBOL(cpu_online_map);

cpumask_t cpu_possible_map = CPU_MASK_NONE;
EXPORT_SYMBOL(cpu_possible_map);

static struct task_struct *current_set[NR_CPUS];

static void smp_ext_bitcall(int, ec_bit_sig);

/*
 * Structure and data for __smp_call_function_map(). This is designed to
 * minimise static memory requirements. It also looks cleaner.
 */
static DEFINE_SPINLOCK(call_lock);

struct call_data_struct {
	void (*func) (void *info);
	void *info;
	cpumask_t started;
	cpumask_t finished;
	int wait;
};

static struct call_data_struct *call_data;

/*
 * 'Call function' interrupt callback
 */
static void do_call_function(void)
{
	void (*func) (void *info) = call_data->func;
	void *info = call_data->info;
	int wait = call_data->wait;

	cpu_set(smp_processor_id(), call_data->started);
	(*func)(info);
	if (wait)
		cpu_set(smp_processor_id(), call_data->finished);;
}

static void __smp_call_function_map(void (*func) (void *info), void *info,
				    int nonatomic, int wait, cpumask_t map)
{
	struct call_data_struct data;
	int cpu, local = 0;

	/*
	 * Can deadlock when interrupts are disabled or if in wrong context.
	 */
	WARN_ON(irqs_disabled() || in_irq());

	/*
	 * Check for local function call. We have to have the same call order
	 * as in on_each_cpu() because of machine_restart_smp().
	 */
	if (cpu_isset(smp_processor_id(), map)) {
		local = 1;
		cpu_clear(smp_processor_id(), map);
	}

	cpus_and(map, map, cpu_online_map);
	if (cpus_empty(map))
		goto out;

	data.func = func;
	data.info = info;
	data.started = CPU_MASK_NONE;
	data.wait = wait;
	if (wait)
		data.finished = CPU_MASK_NONE;

	spin_lock(&call_lock);
	call_data = &data;

	for_each_cpu_mask(cpu, map)
		smp_ext_bitcall(cpu, ec_call_function);

	/* Wait for response */
	while (!cpus_equal(map, data.started))
		cpu_relax();
	if (wait)
		while (!cpus_equal(map, data.finished))
			cpu_relax();
	spin_unlock(&call_lock);
out:
	if (local) {
		local_irq_disable();
		func(info);
		local_irq_enable();
	}
}

/*
 * smp_call_function:
 * @func: the function to run; this must be fast and non-blocking
 * @info: an arbitrary pointer to pass to the function
 * @nonatomic: unused
 * @wait: if true, wait (atomically) until function has completed on other CPUs
 *
 * Run a function on all other CPUs.
 *
 * You must not call this function with disabled interrupts, from a
 * hardware interrupt handler or from a bottom half.
 */
int smp_call_function(void (*func) (void *info), void *info, int nonatomic,
		      int wait)
{
	cpumask_t map;

	preempt_disable();
	map = cpu_online_map;
	cpu_clear(smp_processor_id(), map);
	__smp_call_function_map(func, info, nonatomic, wait, map);
	preempt_enable();
	return 0;
}
EXPORT_SYMBOL(smp_call_function);

/*
 * smp_call_function_single:
 * @cpu: the CPU where func should run
 * @func: the function to run; this must be fast and non-blocking
 * @info: an arbitrary pointer to pass to the function
 * @nonatomic: unused
 * @wait: if true, wait (atomically) until function has completed on other CPUs
 *
 * Run a function on one processor.
 *
 * You must not call this function with disabled interrupts, from a
 * hardware interrupt handler or from a bottom half.
 */
int smp_call_function_single(int cpu, void (*func) (void *info), void *info,
			     int nonatomic, int wait)
{
	preempt_disable();
	__smp_call_function_map(func, info, nonatomic, wait,
				cpumask_of_cpu(cpu));
	preempt_enable();
	return 0;
}
EXPORT_SYMBOL(smp_call_function_single);

void smp_send_stop(void)
{
	int cpu, rc;

	/* Disable all interrupts/machine checks */
	__load_psw_mask(psw_kernel_bits & ~PSW_MASK_MCHECK);

	/* write magic number to zero page (absolute 0) */
	lowcore_ptr[smp_processor_id()]->panic_magic = __PANIC_MAGIC;

	/* stop all processors */
	for_each_online_cpu(cpu) {
		if (cpu == smp_processor_id())
			continue;
		do {
			rc = signal_processor(cpu, sigp_stop);
		} while (rc == sigp_busy);

		while (!smp_cpu_not_running(cpu))
			cpu_relax();
	}
}

/*
 * Reboot, halt and power_off routines for SMP.
 */
void machine_restart_smp(char *__unused)
{
	smp_send_stop();
	do_reipl();
}

void machine_halt_smp(void)
{
	smp_send_stop();
	if (MACHINE_IS_VM && strlen(vmhalt_cmd) > 0)
		__cpcmd(vmhalt_cmd, NULL, 0, NULL);
	signal_processor(smp_processor_id(), sigp_stop_and_store_status);
	for (;;);
}

void machine_power_off_smp(void)
{
	smp_send_stop();
	if (MACHINE_IS_VM && strlen(vmpoff_cmd) > 0)
		__cpcmd(vmpoff_cmd, NULL, 0, NULL);
	signal_processor(smp_processor_id(), sigp_stop_and_store_status);
	for (;;);
}

/*
 * This is the main routine where commands issued by other
 * cpus are handled.
 */

static void do_ext_call_interrupt(__u16 code)
{
	unsigned long bits;

	/*
	 * handle bit signal external calls
	 *
	 * For the ec_schedule signal we have to do nothing. All the work
	 * is done automatically when we return from the interrupt.
	 */
	bits = xchg(&S390_lowcore.ext_call_fast, 0);

	if (test_bit(ec_call_function, &bits))
		do_call_function();
}

/*
 * Send an external call sigp to another cpu and return without waiting
 * for its completion.
 */
static void smp_ext_bitcall(int cpu, ec_bit_sig sig)
{
	/*
	 * Set signaling bit in lowcore of target cpu and kick it
	 */
	set_bit(sig, (unsigned long *) &lowcore_ptr[cpu]->ext_call_fast);
	while (signal_processor(cpu, sigp_emergency_signal) == sigp_busy)
		udelay(10);
}

#ifndef CONFIG_64BIT
/*
 * this function sends a 'purge tlb' signal to another CPU.
 */
void smp_ptlb_callback(void *info)
{
	__tlb_flush_local();
}

void smp_ptlb_all(void)
{
	on_each_cpu(smp_ptlb_callback, NULL, 0, 1);
}
EXPORT_SYMBOL(smp_ptlb_all);
#endif /* ! CONFIG_64BIT */

/*
 * this function sends a 'reschedule' IPI to another CPU.
 * it goes straight through and wastes no time serializing
 * anything. Worst case is that we lose a reschedule ...
 */
void smp_send_reschedule(int cpu)
{
	smp_ext_bitcall(cpu, ec_schedule);
}

/*
 * parameter area for the set/clear control bit callbacks
 */
struct ec_creg_mask_parms {
	unsigned long orvals[16];
	unsigned long andvals[16];
};

/*
 * callback for setting/clearing control bits
 */
static void smp_ctl_bit_callback(void *info)
{
	struct ec_creg_mask_parms *pp = info;
	unsigned long cregs[16];
	int i;

	__ctl_store(cregs, 0, 15);
	for (i = 0; i <= 15; i++)
		cregs[i] = (cregs[i] & pp->andvals[i]) | pp->orvals[i];
	__ctl_load(cregs, 0, 15);
}

/*
 * Set a bit in a control register of all cpus
 */
void smp_ctl_set_bit(int cr, int bit)
{
	struct ec_creg_mask_parms parms;

	memset(&parms.orvals, 0, sizeof(parms.orvals));
	memset(&parms.andvals, 0xff, sizeof(parms.andvals));
	parms.orvals[cr] = 1 << bit;
	on_each_cpu(smp_ctl_bit_callback, &parms, 0, 1);
}
EXPORT_SYMBOL(smp_ctl_set_bit);

/*
 * Clear a bit in a control register of all cpus
 */
void smp_ctl_clear_bit(int cr, int bit)
{
	struct ec_creg_mask_parms parms;

	memset(&parms.orvals, 0, sizeof(parms.orvals));
	memset(&parms.andvals, 0xff, sizeof(parms.andvals));
	parms.andvals[cr] = ~(1L << bit);
	on_each_cpu(smp_ctl_bit_callback, &parms, 0, 1);
}
EXPORT_SYMBOL(smp_ctl_clear_bit);

#if defined(CONFIG_ZFCPDUMP) || defined(CONFIG_ZFCPDUMP_MODULE)

/*
 * zfcpdump_prefix_array holds prefix registers for the following scenario:
 * 64 bit zfcpdump kernel and 31 bit kernel which is to be dumped. We have to
 * save its prefix registers, since they get lost, when switching from 31 bit
 * to 64 bit.
 */
unsigned int zfcpdump_prefix_array[NR_CPUS + 1] \
	__attribute__((__section__(".data")));

static void __init smp_get_save_area(unsigned int cpu, unsigned int phy_cpu)
{
	if (ipl_info.type != IPL_TYPE_FCP_DUMP)
		return;
	if (cpu >= NR_CPUS) {
		printk(KERN_WARNING "Registers for cpu %i not saved since dump "
		       "kernel was compiled with NR_CPUS=%i\n", cpu, NR_CPUS);
		return;
	}
	zfcpdump_save_areas[cpu] = alloc_bootmem(sizeof(union save_area));
	__cpu_logical_map[1] = (__u16) phy_cpu;
	while (signal_processor(1, sigp_stop_and_store_status) == sigp_busy)
		cpu_relax();
	memcpy(zfcpdump_save_areas[cpu],
	       (void *)(unsigned long) store_prefix() + SAVE_AREA_BASE,
	       SAVE_AREA_SIZE);
#ifdef CONFIG_64BIT
	/* copy original prefix register */
	zfcpdump_save_areas[cpu]->s390x.pref_reg = zfcpdump_prefix_array[cpu];
#endif
}

union save_area *zfcpdump_save_areas[NR_CPUS + 1];
EXPORT_SYMBOL_GPL(zfcpdump_save_areas);

#else

static inline void smp_get_save_area(unsigned int cpu, unsigned int phy_cpu) { }

#endif /* CONFIG_ZFCPDUMP || CONFIG_ZFCPDUMP_MODULE */

/*
 * Lets check how many CPUs we have.
 */
static unsigned int __init smp_count_cpus(void)
{
	unsigned int cpu, num_cpus;
	__u16 boot_cpu_addr;

	/*
	 * cpu 0 is the boot cpu. See smp_prepare_boot_cpu.
	 */