/*
* arch/s390/kernel/smp.c
*
- * Copyright IBM Corp. 1999,2007
+ * Copyright IBM Corp. 1999, 2009
* Author(s): Denis Joseph Barrow (djbarrow@de.ibm.com,barrow_dj@yahoo.com),
* Martin Schwidefsky (schwidefsky@de.ibm.com)
* Heiko Carstens (heiko.carstens@de.ibm.com)
* cpu_number_map in other architectures.
*/
+#define KMSG_COMPONENT "cpu"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
#include <linux/module.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/delay.h>
#include <linux/cache.h>
#include <linux/interrupt.h>
+#include <linux/irqflags.h>
#include <linux/cpu.h>
#include <linux/timex.h>
#include <linux/bootmem.h>
+#include <linux/slab.h>
+#include <asm/asm-offsets.h>
#include <asm/ipl.h>
#include <asm/setup.h>
#include <asm/sigp.h>
#include <asm/timer.h>
#include <asm/lowcore.h>
#include <asm/sclp.h>
+#include <asm/cputime.h>
+#include <asm/vdso.h>
#include <asm/cpu.h>
+#include "entry.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_ALL;
-EXPORT_SYMBOL(cpu_possible_map);
+/* logical cpu to cpu address */
+unsigned short __cpu_logical_map[NR_CPUS];
static struct task_struct *current_set[NR_CPUS];
CPU_STATE_CONFIGURED,
};
-#ifdef CONFIG_HOTPLUG_CPU
-static DEFINE_MUTEX(smp_cpu_state_mutex);
-#endif
+DEFINE_MUTEX(smp_cpu_state_mutex);
+int smp_cpu_polarization[NR_CPUS];
static int smp_cpu_state[NR_CPUS];
+static int cpu_management;
static DEFINE_PER_CPU(struct cpu, cpu_devices);
-DEFINE_PER_CPU(struct s390_idle_data, s390_idle);
-
-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;
+static void smp_ext_bitcall(int, int);
-/*
- * '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)
+static int raw_cpu_stopped(int cpu)
{
- 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);
+ u32 status;
- /* 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();
+ switch (raw_sigp_ps(&status, 0, cpu, sigp_sense)) {
+ case sigp_status_stored:
+ /* Check for stopped and check stop state */
+ if (status & 0x50)
+ return 1;
+ break;
+ default:
+ break;
}
+ return 0;
}
-/*
- * 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)
+static inline int cpu_stopped(int cpu)
{
- 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;
+ return raw_cpu_stopped(cpu_logical_map(cpu));
}
-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)
+void smp_switch_to_ipl_cpu(void (*func)(void *), void *data)
{
- preempt_disable();
- __smp_call_function_map(func, info, nonatomic, wait,
- cpumask_of_cpu(cpu));
- preempt_enable();
- return 0;
+ struct _lowcore *lc, *current_lc;
+ struct stack_frame *sf;
+ struct pt_regs *regs;
+ unsigned long sp;
+
+ if (smp_processor_id() == 0)
+ func(data);
+ __load_psw_mask(PSW_BASE_BITS | PSW_DEFAULT_KEY);
+ /* Disable lowcore protection */
+ __ctl_clear_bit(0, 28);
+ current_lc = lowcore_ptr[smp_processor_id()];
+ lc = lowcore_ptr[0];
+ if (!lc)
+ lc = current_lc;
+ lc->restart_psw.mask = PSW_BASE_BITS | PSW_DEFAULT_KEY;
+ lc->restart_psw.addr = PSW_ADDR_AMODE | (unsigned long) smp_restart_cpu;
+ if (!cpu_online(0))
+ smp_switch_to_cpu(func, data, 0, stap(), __cpu_logical_map[0]);
+ while (sigp(0, sigp_stop_and_store_status) == sigp_busy)
+ cpu_relax();
+ sp = lc->panic_stack;
+ sp -= sizeof(struct pt_regs);
+ regs = (struct pt_regs *) sp;
+ memcpy(®s->gprs, ¤t_lc->gpregs_save_area, sizeof(regs->gprs));
+ regs->psw = lc->psw_save_area;
+ sp -= STACK_FRAME_OVERHEAD;
+ sf = (struct stack_frame *) sp;
+ sf->back_chain = regs->gprs[15];
+ smp_switch_to_cpu(func, data, sp, stap(), __cpu_logical_map[0]);
}
-EXPORT_SYMBOL(smp_call_function_single);
void smp_send_stop(void)
{
/* 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;
+ trace_hardirqs_off();
/* stop all processors */
for_each_online_cpu(cpu) {
if (cpu == smp_processor_id())
continue;
do {
- rc = signal_processor(cpu, sigp_stop);
+ rc = sigp(cpu, sigp_stop);
} while (rc == sigp_busy);
- while (!smp_cpu_not_running(cpu))
+ while (!cpu_stopped(cpu))
cpu_relax();
}
}
bits = xchg(&S390_lowcore.ext_call_fast, 0);
if (test_bit(ec_call_function, &bits))
- do_call_function();
+ generic_smp_call_function_interrupt();
+
+ if (test_bit(ec_call_function_single, &bits))
+ generic_smp_call_function_single_interrupt();
}
/*
* 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)
+static void smp_ext_bitcall(int cpu, int 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)
+ while (sigp(cpu, sigp_emergency_signal) == sigp_busy)
udelay(10);
}
+void arch_send_call_function_ipi_mask(const struct cpumask *mask)
+{
+ int cpu;
+
+ for_each_cpu(cpu, mask)
+ smp_ext_bitcall(cpu, ec_call_function);
+}
+
+void arch_send_call_function_single_ipi(int cpu)
+{
+ smp_ext_bitcall(cpu, ec_call_function_single);
+}
+
#ifndef CONFIG_64BIT
/*
* this function sends a 'purge tlb' signal to another CPU.
*/
-void smp_ptlb_callback(void *info)
+static void smp_ptlb_callback(void *info)
{
__tlb_flush_local();
}
void smp_ptlb_all(void)
{
- on_each_cpu(smp_ptlb_callback, NULL, 0, 1);
+ on_each_cpu(smp_ptlb_callback, NULL, 1);
}
EXPORT_SYMBOL(smp_ptlb_all);
#endif /* ! CONFIG_64BIT */
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);
+ on_each_cpu(smp_ctl_bit_callback, &parms, 1);
}
EXPORT_SYMBOL(smp_ctl_set_bit);
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);
+ on_each_cpu(smp_ctl_bit_callback, &parms, 1);
}
EXPORT_SYMBOL(smp_ctl_clear_bit);
-/*
- * In early ipl state a temp. logically cpu number is needed, so the sigp
- * functions can be used to sense other cpus. Since NR_CPUS is >= 2 on
- * CONFIG_SMP and the ipl cpu is logical cpu 0, it must be 1.
- */
-#define CPU_INIT_NO 1
-
-#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")));
+#ifdef CONFIG_ZFCPDUMP
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);
+ pr_warning("CPU %i exceeds the maximum %i and is excluded from "
+ "the dump\n", cpu, NR_CPUS - 1);
return;
}
- zfcpdump_save_areas[cpu] = kmalloc(sizeof(union save_area), GFP_KERNEL);
- __cpu_logical_map[CPU_INIT_NO] = (__u16) phy_cpu;
- while (signal_processor(CPU_INIT_NO, sigp_stop_and_store_status) ==
- sigp_busy)
+ zfcpdump_save_areas[cpu] = kmalloc(sizeof(struct save_area), GFP_KERNEL);
+ while (raw_sigp(phy_cpu, 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
+ memcpy_real(zfcpdump_save_areas[cpu],
+ (void *)(unsigned long) store_prefix() + SAVE_AREA_BASE,
+ sizeof(struct save_area));
}
-union save_area *zfcpdump_save_areas[NR_CPUS + 1];
+struct 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 */
-
-static int cpu_stopped(int cpu)
-{
- __u32 status;
-
- /* Check for stopped state */
- if (signal_processor_ps(&status, 0, cpu, sigp_sense) ==
- sigp_status_stored) {
- if (status & 0x40)
- return 1;
- }
- return 0;
-}
+#endif /* CONFIG_ZFCPDUMP */
static int cpu_known(int cpu_id)
{
{
int cpu_id, logical_cpu;
- logical_cpu = first_cpu(avail);
- if (logical_cpu == NR_CPUS)
+ logical_cpu = cpumask_first(&avail);
+ if (logical_cpu >= nr_cpu_ids)
return 0;
- for (cpu_id = 0; cpu_id <= 65535; cpu_id++) {
+ for (cpu_id = 0; cpu_id <= MAX_CPU_ADDRESS; cpu_id++) {
if (cpu_known(cpu_id))
continue;
__cpu_logical_map[logical_cpu] = cpu_id;
+ smp_cpu_polarization[logical_cpu] = POLARIZATION_UNKNWN;
if (!cpu_stopped(logical_cpu))
continue;
cpu_set(logical_cpu, cpu_present_map);
smp_cpu_state[logical_cpu] = CPU_STATE_CONFIGURED;
- logical_cpu = next_cpu(logical_cpu, avail);
- if (logical_cpu == NR_CPUS)
+ logical_cpu = cpumask_next(logical_cpu, &avail);
+ if (logical_cpu >= nr_cpu_ids)
break;
}
return 0;
int cpu_id, logical_cpu, cpu;
int rc;
- logical_cpu = first_cpu(avail);
- if (logical_cpu == NR_CPUS)
+ logical_cpu = cpumask_first(&avail);
+ if (logical_cpu >= nr_cpu_ids)
return 0;
info = kmalloc(sizeof(*info), GFP_KERNEL);
if (!info)
if (cpu_known(cpu_id))
continue;
__cpu_logical_map[logical_cpu] = cpu_id;
+ smp_cpu_polarization[logical_cpu] = POLARIZATION_UNKNWN;
cpu_set(logical_cpu, cpu_present_map);
if (cpu >= info->configured)
smp_cpu_state[logical_cpu] = CPU_STATE_STANDBY;
else
smp_cpu_state[logical_cpu] = CPU_STATE_CONFIGURED;
- logical_cpu = next_cpu(logical_cpu, avail);
- if (logical_cpu == NR_CPUS)
+ logical_cpu = cpumask_next(logical_cpu, &avail);
+ if (logical_cpu >= nr_cpu_ids)
break;
}
out:
return rc;
}
-static int smp_rescan_cpus(void)
+static int __smp_rescan_cpus(void)
{
cpumask_t avail;
c_cpus = 1;
s_cpus = 0;
- boot_cpu_addr = S390_lowcore.cpu_data.cpu_addr;
+ boot_cpu_addr = __cpu_logical_map[0];
info = kmalloc(sizeof(*info), GFP_KERNEL);
if (!info)
panic("smp_detect_cpus failed to allocate memory\n");
/* Use sigp detection algorithm if sclp doesn't work. */
if (sclp_get_cpu_info(info)) {
smp_use_sigp_detection = 1;
- for (cpu = 0; cpu <= 65535; cpu++) {
+ for (cpu = 0; cpu <= MAX_CPU_ADDRESS; cpu++) {
if (cpu == boot_cpu_addr)
continue;
- __cpu_logical_map[CPU_INIT_NO] = cpu;
- if (!cpu_stopped(CPU_INIT_NO))
+ if (!raw_cpu_stopped(cpu))
continue;
smp_get_save_area(c_cpus, cpu);
c_cpus++;
cpu_addr = info->cpu[cpu].address;
if (cpu_addr == boot_cpu_addr)
continue;
- __cpu_logical_map[CPU_INIT_NO] = cpu_addr;
- if (!cpu_stopped(CPU_INIT_NO)) {
+ if (!raw_cpu_stopped(cpu_addr)) {
s_cpus++;
continue;
}
}
out:
kfree(info);
- printk(KERN_INFO "CPUs: %d configured, %d standby\n", c_cpus, s_cpus);
- lock_cpu_hotplug();
- smp_rescan_cpus();
- unlock_cpu_hotplug();
+ pr_info("%d configured CPUs, %d standby CPUs\n", c_cpus, s_cpus);
+ get_online_cpus();
+ __smp_rescan_cpus();
+ put_online_cpus();
}
/*
preempt_disable();
/* Enable TOD clock interrupts on the secondary cpu. */
init_cpu_timer();
-#ifdef CONFIG_VIRT_TIMER
/* Enable cpu timer interrupts on the secondary cpu. */
init_cpu_vtimer();
-#endif
/* Enable pfault pseudo page faults on this cpu. */
pfault_init();
+ /* call cpu notifiers */
+ notify_cpu_starting(smp_processor_id());
/* Mark this cpu as online */
+ ipi_call_lock();
cpu_set(smp_processor_id(), cpu_online_map);
+ ipi_call_unlock();
/* Switch on interrupts */
local_irq_enable();
/* Print info about this processor */
- print_cpu_info(&S390_lowcore.cpu_data);
+ print_cpu_info();
/* cpu_idle will call schedule for us */
cpu_idle();
return 0;
if (IS_ERR(p))
panic("failed fork for CPU %u: %li", cpu, PTR_ERR(p));
current_set[cpu] = p;
- spin_lock_init(&(&per_cpu(s390_idle, cpu))->lock);
+}
+
+static int __cpuinit smp_alloc_lowcore(int cpu)
+{
+ unsigned long async_stack, panic_stack;
+ struct _lowcore *lowcore;
+
+ lowcore = (void *) __get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER);
+ if (!lowcore)
+ return -ENOMEM;
+ async_stack = __get_free_pages(GFP_KERNEL, ASYNC_ORDER);
+ panic_stack = __get_free_page(GFP_KERNEL);
+ if (!panic_stack || !async_stack)
+ goto out;
+ memcpy(lowcore, &S390_lowcore, 512);
+ memset((char *)lowcore + 512, 0, sizeof(*lowcore) - 512);
+ lowcore->async_stack = async_stack + ASYNC_SIZE;
+ lowcore->panic_stack = panic_stack + PAGE_SIZE;
+
+#ifndef CONFIG_64BIT
+ if (MACHINE_HAS_IEEE) {
+ unsigned long save_area;
+
+ save_area = get_zeroed_page(GFP_KERNEL);
+ if (!save_area)
+ goto out;
+ lowcore->extended_save_area_addr = (u32) save_area;
+ }
+#else
+ if (vdso_alloc_per_cpu(cpu, lowcore))
+ goto out;
+#endif
+ lowcore_ptr[cpu] = lowcore;
+ return 0;
+
+out:
+ free_page(panic_stack);
+ free_pages(async_stack, ASYNC_ORDER);
+ free_pages((unsigned long) lowcore, LC_ORDER);
+ return -ENOMEM;
+}
+
+static void smp_free_lowcore(int cpu)
+{
+ struct _lowcore *lowcore;
+
+ lowcore = lowcore_ptr[cpu];
+#ifndef CONFIG_64BIT
+ if (MACHINE_HAS_IEEE)
+ free_page((unsigned long) lowcore->extended_save_area_addr);
+#else
+ vdso_free_per_cpu(cpu, lowcore);
+#endif
+ free_page(lowcore->panic_stack - PAGE_SIZE);
+ free_pages(lowcore->async_stack - ASYNC_SIZE, ASYNC_ORDER);
+ free_pages((unsigned long) lowcore, LC_ORDER);
+ lowcore_ptr[cpu] = NULL;
}
/* Upping and downing of CPUs */
-int __cpu_up(unsigned int cpu)
+int __cpuinit __cpu_up(unsigned int cpu)
{
- struct task_struct *idle;
struct _lowcore *cpu_lowcore;
+ struct task_struct *idle;
struct stack_frame *sf;
- sigp_ccode ccode;
+ u32 lowcore;
+ int ccode;
if (smp_cpu_state[cpu] != CPU_STATE_CONFIGURED)
return -EIO;
-
- ccode = signal_processor_p((__u32)(unsigned long)(lowcore_ptr[cpu]),
- cpu, sigp_set_prefix);
- if (ccode) {
- printk("sigp_set_prefix failed for cpu %d "
- "with condition code %d\n",
- (int) cpu, (int) ccode);
- return -EIO;
- }
+ if (smp_alloc_lowcore(cpu))
+ return -ENOMEM;
+ do {
+ ccode = sigp(cpu, sigp_initial_cpu_reset);
+ if (ccode == sigp_busy)
+ udelay(10);
+ if (ccode == sigp_not_operational)
+ goto err_out;
+ } while (ccode == sigp_busy);
+
+ lowcore = (u32)(unsigned long)lowcore_ptr[cpu];
+ while (sigp_p(lowcore, cpu, sigp_set_prefix) == sigp_busy)
+ udelay(10);
idle = current_set[cpu];
cpu_lowcore = lowcore_ptr[cpu];
cpu_lowcore->kernel_stack = (unsigned long)
task_stack_page(idle) + THREAD_SIZE;
+ cpu_lowcore->thread_info = (unsigned long) task_thread_info(idle);
sf = (struct stack_frame *) (cpu_lowcore->kernel_stack
- sizeof(struct pt_regs)
- sizeof(struct stack_frame));
memset(sf, 0, sizeof(struct stack_frame));
sf->gprs[9] = (unsigned long) sf;
cpu_lowcore->save_area[15] = (unsigned long) sf;
- __ctl_store(cpu_lowcore->cregs_save_area[0], 0, 15);
+ __ctl_store(cpu_lowcore->cregs_save_area, 0, 15);
asm volatile(
" stam 0,15,0(%0)"
: : "a" (&cpu_lowcore->access_regs_save_area) : "memory");
cpu_lowcore->percpu_offset = __per_cpu_offset[cpu];
cpu_lowcore->current_task = (unsigned long) idle;
- cpu_lowcore->cpu_data.cpu_nr = cpu;
+ cpu_lowcore->cpu_nr = cpu;
+ cpu_lowcore->kernel_asce = S390_lowcore.kernel_asce;
+ cpu_lowcore->machine_flags = S390_lowcore.machine_flags;
+ cpu_lowcore->ftrace_func = S390_lowcore.ftrace_func;
eieio();
- while (signal_processor(cpu, sigp_restart) == sigp_busy)
+ while (sigp(cpu, sigp_restart) == sigp_busy)
udelay(10);
while (!cpu_online(cpu))
cpu_relax();
return 0;
+
+err_out:
+ smp_free_lowcore(cpu);
+ return -EIO;
}
static int __init setup_possible_cpus(char *s)
int pcpus, cpu;
pcpus = simple_strtoul(s, NULL, 0);
- cpu_possible_map = cpumask_of_cpu(0);
- for (cpu = 1; cpu < pcpus && cpu < NR_CPUS; cpu++)
- cpu_set(cpu, cpu_possible_map);
+ init_cpu_possible(cpumask_of(0));
+ for (cpu = 1; cpu < pcpus && cpu < nr_cpu_ids; cpu++)
+ set_cpu_possible(cpu, true);
return 0;
}
early_param("possible_cpus", setup_possible_cpus);
void __cpu_die(unsigned int cpu)
{
/* Wait until target cpu is down */
- while (!smp_cpu_not_running(cpu))
+ while (!cpu_stopped(cpu))
cpu_relax();
- printk(KERN_INFO "Processor %d spun down\n", cpu);
+ while (sigp_p(0, cpu, sigp_set_prefix) == sigp_busy)
+ udelay(10);
+ smp_free_lowcore(cpu);
+ pr_info("Processor %d stopped\n", cpu);
}
void cpu_die(void)
{
idle_task_exit();
- signal_processor(smp_processor_id(), sigp_stop);
- BUG();
+ while (sigp(smp_processor_id(), sigp_stop) == sigp_busy)
+ cpu_relax();
for (;;);
}
#endif /* CONFIG_HOTPLUG_CPU */
-/*
- * Cycle through the processors and setup structures.
- */
-
void __init smp_prepare_cpus(unsigned int max_cpus)
{
- unsigned long stack;
+#ifndef CONFIG_64BIT
+ unsigned long save_area = 0;
+#endif
+ unsigned long async_stack, panic_stack;
+ struct _lowcore *lowcore;
unsigned int cpu;
- int i;
smp_detect_cpus();
/* request the 0x1201 emergency signal external interrupt */
if (register_external_interrupt(0x1201, do_ext_call_interrupt) != 0)
panic("Couldn't request external interrupt 0x1201");
- memset(lowcore_ptr, 0, sizeof(lowcore_ptr));
- /*
- * Initialize prefix pages and stacks for all possible cpus
- */
- print_cpu_info(&S390_lowcore.cpu_data);
-
- for_each_possible_cpu(i) {
- lowcore_ptr[i] = (struct _lowcore *)
- __get_free_pages(GFP_KERNEL | GFP_DMA,
- sizeof(void*) == 8 ? 1 : 0);
- stack = __get_free_pages(GFP_KERNEL, ASYNC_ORDER);
- if (!lowcore_ptr[i] || !stack)
- panic("smp_boot_cpus failed to allocate memory\n");
-
- *(lowcore_ptr[i]) = S390_lowcore;
- lowcore_ptr[i]->async_stack = stack + ASYNC_SIZE;
- stack = __get_free_pages(GFP_KERNEL, 0);
- if (!stack)
- panic("smp_boot_cpus failed to allocate memory\n");
- lowcore_ptr[i]->panic_stack = stack + PAGE_SIZE;
+ print_cpu_info();
+
+ /* Reallocate current lowcore, but keep its contents. */
+ lowcore = (void *) __get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER);
+ panic_stack = __get_free_page(GFP_KERNEL);
+ async_stack = __get_free_pages(GFP_KERNEL, ASYNC_ORDER);
+ BUG_ON(!lowcore || !panic_stack || !async_stack);
#ifndef CONFIG_64BIT
- if (MACHINE_HAS_IEEE) {
- lowcore_ptr[i]->extended_save_area_addr =
- (__u32) __get_free_pages(GFP_KERNEL, 0);
- if (!lowcore_ptr[i]->extended_save_area_addr)
- panic("smp_boot_cpus failed to "
- "allocate memory\n");
- }
+ if (MACHINE_HAS_IEEE)
+ save_area = get_zeroed_page(GFP_KERNEL);
#endif
- }
+ local_irq_disable();
+ local_mcck_disable();
+ lowcore_ptr[smp_processor_id()] = lowcore;
+ *lowcore = S390_lowcore;
+ lowcore->panic_stack = panic_stack + PAGE_SIZE;
+ lowcore->async_stack = async_stack + ASYNC_SIZE;
#ifndef CONFIG_64BIT
if (MACHINE_HAS_IEEE)
- ctl_set_bit(14, 29); /* enable extended save area */
+ lowcore->extended_save_area_addr = (u32) save_area;
+#endif
+ set_prefix((u32)(unsigned long) lowcore);
+ local_mcck_enable();
+ local_irq_enable();
+#ifdef CONFIG_64BIT
+ if (vdso_alloc_per_cpu(smp_processor_id(), &S390_lowcore))
+ BUG();
#endif
- set_prefix((u32)(unsigned long) lowcore_ptr[smp_processor_id()]);
-
for_each_possible_cpu(cpu)
if (cpu != smp_processor_id())
smp_create_idle(cpu);
S390_lowcore.percpu_offset = __per_cpu_offset[0];
current_set[0] = current;
smp_cpu_state[0] = CPU_STATE_CONFIGURED;
- spin_lock_init(&(&__get_cpu_var(s390_idle))->lock);
+ smp_cpu_polarization[0] = POLARIZATION_UNKNWN;
}
void __init smp_cpus_done(unsigned int max_cpus)
{
}
+void __init smp_setup_processor_id(void)
+{
+ S390_lowcore.cpu_nr = 0;
+ __cpu_logical_map[0] = stap();
+}
+
/*
* the frequency of the profiling timer can be changed
* by writing a multiplier value into /proc/profile.
}
#ifdef CONFIG_HOTPLUG_CPU
-static ssize_t cpu_configure_show(struct sys_device *dev, char *buf)
+static ssize_t cpu_configure_show(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
{
ssize_t count;
return count;
}
-static ssize_t cpu_configure_store(struct sys_device *dev, const char *buf,
- size_t count)
+static ssize_t cpu_configure_store(struct sys_device *dev,
+ struct sysdev_attribute *attr,
+ const char *buf, size_t count)
{
int cpu = dev->id;
int val, rc;
if (val != 0 && val != 1)
return -EINVAL;
+ get_online_cpus();
mutex_lock(&smp_cpu_state_mutex);
- lock_cpu_hotplug();
rc = -EBUSY;
- if (cpu_online(cpu))
+ /* disallow configuration changes of online cpus and cpu 0 */
+ if (cpu_online(cpu) || cpu == 0)
goto out;
rc = 0;
switch (val) {
case 0:
if (smp_cpu_state[cpu] == CPU_STATE_CONFIGURED) {
rc = sclp_cpu_deconfigure(__cpu_logical_map[cpu]);
- if (!rc)
+ if (!rc) {
smp_cpu_state[cpu] = CPU_STATE_STANDBY;
+ smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN;
+ }
}
break;
case 1:
if (smp_cpu_state[cpu] == CPU_STATE_STANDBY) {
rc = sclp_cpu_configure(__cpu_logical_map[cpu]);
- if (!rc)
+ if (!rc) {
smp_cpu_state[cpu] = CPU_STATE_CONFIGURED;
+ smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN;
+ }
}
break;
default:
break;
}
out:
- unlock_cpu_hotplug();
mutex_unlock(&smp_cpu_state_mutex);
+ put_online_cpus();
return rc ? rc : count;
}
static SYSDEV_ATTR(configure, 0644, cpu_configure_show, cpu_configure_store);
#endif /* CONFIG_HOTPLUG_CPU */
-static ssize_t show_cpu_address(struct sys_device *dev, char *buf)
+static ssize_t cpu_polarization_show(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
+{
+ int cpu = dev->id;
+ ssize_t count;
+
+ mutex_lock(&smp_cpu_state_mutex);
+ switch (smp_cpu_polarization[cpu]) {
+ case POLARIZATION_HRZ:
+ count = sprintf(buf, "horizontal\n");
+ break;
+ case POLARIZATION_VL:
+ count = sprintf(buf, "vertical:low\n");
+ break;
+ case POLARIZATION_VM:
+ count = sprintf(buf, "vertical:medium\n");
+ break;
+ case POLARIZATION_VH:
+ count = sprintf(buf, "vertical:high\n");
+ break;
+ default:
+ count = sprintf(buf, "unknown\n");
+ break;
+ }
+ mutex_unlock(&smp_cpu_state_mutex);
+ return count;
+}
+static SYSDEV_ATTR(polarization, 0444, cpu_polarization_show, NULL);
+
+static ssize_t show_cpu_address(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
{
return sprintf(buf, "%d\n", __cpu_logical_map[dev->id]);
}
&attr_configure.attr,
#endif
&attr_address.attr,
+ &attr_polarization.attr,
NULL,
};
.attrs = cpu_common_attrs,
};
-static ssize_t show_capability(struct sys_device *dev, char *buf)
+static ssize_t show_capability(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
{
unsigned int capability;
int rc;
}
static SYSDEV_ATTR(capability, 0444, show_capability, NULL);
-static ssize_t show_idle_count(struct sys_device *dev, char *buf)
+static ssize_t show_idle_count(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
{
struct s390_idle_data *idle;
unsigned long long idle_count;
+ unsigned int sequence;
idle = &per_cpu(s390_idle, dev->id);
- spin_lock_irq(&idle->lock);
+repeat:
+ sequence = idle->sequence;
+ smp_rmb();
+ if (sequence & 1)
+ goto repeat;
idle_count = idle->idle_count;
- spin_unlock_irq(&idle->lock);
+ if (idle->idle_enter)
+ idle_count++;
+ smp_rmb();
+ if (idle->sequence != sequence)
+ goto repeat;
return sprintf(buf, "%llu\n", idle_count);
}
static SYSDEV_ATTR(idle_count, 0444, show_idle_count, NULL);
-static ssize_t show_idle_time(struct sys_device *dev, char *buf)
+static ssize_t show_idle_time(struct sys_device *dev,
+ struct sysdev_attribute *attr, char *buf)
{
struct s390_idle_data *idle;
- unsigned long long new_time;
+ unsigned long long now, idle_time, idle_enter;
+ unsigned int sequence;
idle = &per_cpu(s390_idle, dev->id);
- spin_lock_irq(&idle->lock);
- if (idle->in_idle) {
- new_time = get_clock();
- idle->idle_time += new_time - idle->idle_enter;
- idle->idle_enter = new_time;
- }
- new_time = idle->idle_time;
- spin_unlock_irq(&idle->lock);
- return sprintf(buf, "%llu\n", new_time >> 12);
+ now = get_clock();
+repeat:
+ sequence = idle->sequence;
+ smp_rmb();
+ if (sequence & 1)
+ goto repeat;
+ idle_time = idle->idle_time;
+ idle_enter = idle->idle_enter;
+ if (idle_enter != 0ULL && idle_enter < now)
+ idle_time += now - idle_enter;
+ smp_rmb();
+ if (idle->sequence != sequence)
+ goto repeat;
+ return sprintf(buf, "%llu\n", idle_time >> 12);
}
static SYSDEV_ATTR(idle_time_us, 0444, show_idle_time, NULL);
struct cpu *c = &per_cpu(cpu_devices, cpu);
struct sys_device *s = &c->sysdev;
struct s390_idle_data *idle;
+ int err = 0;
switch (action) {
case CPU_ONLINE:
case CPU_ONLINE_FROZEN:
idle = &per_cpu(s390_idle, cpu);
- spin_lock_irq(&idle->lock);
- idle->idle_enter = 0;
- idle->idle_time = 0;
- idle->idle_count = 0;
- spin_unlock_irq(&idle->lock);
- if (sysfs_create_group(&s->kobj, &cpu_online_attr_group))
- return NOTIFY_BAD;
+ memset(idle, 0, sizeof(struct s390_idle_data));
+ err = sysfs_create_group(&s->kobj, &cpu_online_attr_group);
break;
case CPU_DEAD:
case CPU_DEAD_FROZEN:
sysfs_remove_group(&s->kobj, &cpu_online_attr_group);
break;
}
- return NOTIFY_OK;
+ return notifier_from_errno(err);
}
static struct notifier_block __cpuinitdata smp_cpu_nb = {
.notifier_call = smp_cpu_notify,
};
-static int smp_add_present_cpu(int cpu)
+static int __devinit smp_add_present_cpu(int cpu)
{
struct cpu *c = &per_cpu(cpu_devices, cpu);
struct sys_device *s = &c->sysdev;
}
#ifdef CONFIG_HOTPLUG_CPU
-static ssize_t rescan_store(struct sys_device *dev, const char *buf,
- size_t count)
+
+int __ref smp_rescan_cpus(void)
{
cpumask_t newcpus;
int cpu;
int rc;
+ get_online_cpus();
mutex_lock(&smp_cpu_state_mutex);
- lock_cpu_hotplug();
newcpus = cpu_present_map;
- rc = smp_rescan_cpus();
+ rc = __smp_rescan_cpus();
if (rc)
goto out;
cpus_andnot(newcpus, cpu_present_map, newcpus);
}
rc = 0;
out:
- unlock_cpu_hotplug();
mutex_unlock(&smp_cpu_state_mutex);
+ put_online_cpus();
+ if (!cpus_empty(newcpus))
+ topology_schedule_update();
+ return rc;
+}
+
+static ssize_t __ref rescan_store(struct sysdev_class *class,
+ struct sysdev_class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ int rc;
+
+ rc = smp_rescan_cpus();
return rc ? rc : count;
}
-static SYSDEV_ATTR(rescan, 0200, NULL, rescan_store);
+static SYSDEV_CLASS_ATTR(rescan, 0200, NULL, rescan_store);
#endif /* CONFIG_HOTPLUG_CPU */
+static ssize_t dispatching_show(struct sysdev_class *class,
+ struct sysdev_class_attribute *attr,
+ char *buf)
+{
+ ssize_t count;
+
+ mutex_lock(&smp_cpu_state_mutex);
+ count = sprintf(buf, "%d\n", cpu_management);
+ mutex_unlock(&smp_cpu_state_mutex);
+ return count;
+}
+
+static ssize_t dispatching_store(struct sysdev_class *dev,
+ struct sysdev_class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ int val, rc;
+ char delim;
+
+ if (sscanf(buf, "%d %c", &val, &delim) != 1)
+ return -EINVAL;
+ if (val != 0 && val != 1)
+ return -EINVAL;
+ rc = 0;
+ get_online_cpus();
+ mutex_lock(&smp_cpu_state_mutex);
+ if (cpu_management == val)
+ goto out;
+ rc = topology_set_cpu_management(val);
+ if (!rc)
+ cpu_management = val;
+out:
+ mutex_unlock(&smp_cpu_state_mutex);
+ put_online_cpus();
+ return rc ? rc : count;
+}
+static SYSDEV_CLASS_ATTR(dispatching, 0644, dispatching_show,
+ dispatching_store);
+
static int __init topology_init(void)
{
int cpu;
register_cpu_notifier(&smp_cpu_nb);
#ifdef CONFIG_HOTPLUG_CPU
- rc = sysfs_create_file(&cpu_sysdev_class.kset.kobj,
- &attr_rescan.attr);
+ rc = sysdev_class_create_file(&cpu_sysdev_class, &attr_rescan);
if (rc)
return rc;
#endif
+ rc = sysdev_class_create_file(&cpu_sysdev_class, &attr_dispatching);
+ if (rc)
+ return rc;
for_each_present_cpu(cpu) {
rc = smp_add_present_cpu(cpu);
if (rc)
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff