#include <linux/delay.h>
#include <linux/efi.h>
#include <linux/bitops.h>
+#include <linux/kexec.h>
#include <asm/atomic.h>
#include <asm/current.h>
#include <asm/mca.h>
/*
- * Structure and data for smp_call_function(). This is designed to minimise static memory
- * requirements. It also looks cleaner.
+ * Note: alignment of 4 entries/cacheline was empirically determined
+ * to be a good tradeoff between hot cachelines & spreading the array
+ * across too many cacheline.
*/
-static __cacheline_aligned DEFINE_SPINLOCK(call_lock);
+static struct local_tlb_flush_counts {
+ unsigned int count;
+} __attribute__((__aligned__(32))) local_tlb_flush_counts[NR_CPUS];
-struct call_data_struct {
- void (*func) (void *info);
- void *info;
- long wait;
- atomic_t started;
- atomic_t finished;
-};
-
-static volatile struct call_data_struct *call_data;
+static DEFINE_PER_CPU(unsigned short, shadow_flush_counts[NR_CPUS]) ____cacheline_aligned;
#define IPI_CALL_FUNC 0
#define IPI_CPU_STOP 1
+#define IPI_CALL_FUNC_SINGLE 2
+#define IPI_KDUMP_CPU_STOP 3
/* This needs to be cacheline aligned because it is written to by *other* CPUs. */
-static DEFINE_PER_CPU(u64, ipi_operation) ____cacheline_aligned;
+static DEFINE_PER_CPU_SHARED_ALIGNED(u64, ipi_operation);
extern void cpu_halt (void);
-void
-lock_ipi_calllock(void)
-{
- spin_lock_irq(&call_lock);
-}
-
-void
-unlock_ipi_calllock(void)
-{
- spin_unlock_irq(&call_lock);
-}
-
static void
-stop_this_cpu (void)
+stop_this_cpu(void)
{
/*
* Remove this CPU:
}
irqreturn_t
-handle_IPI (int irq, void *dev_id, struct pt_regs *regs)
+handle_IPI (int irq, void *dev_id)
{
int this_cpu = get_cpu();
unsigned long *pending_ipis = &__ia64_per_cpu_var(ipi_operation);
ops &= ~(1 << which);
switch (which) {
- case IPI_CALL_FUNC:
- {
- struct call_data_struct *data;
- void (*func)(void *info);
- void *info;
- int wait;
-
- /* release the 'pointer lock' */
- data = (struct call_data_struct *) call_data;
- func = data->func;
- info = data->info;
- wait = data->wait;
-
- mb();
- atomic_inc(&data->started);
- /*
- * At this point the structure may be gone unless
- * wait is true.
- */
- (*func)(info);
-
- /* Notify the sending CPU that the task is done. */
- mb();
- if (wait)
- atomic_inc(&data->finished);
- }
- break;
-
- case IPI_CPU_STOP:
+ case IPI_CPU_STOP:
stop_this_cpu();
break;
-
- default:
- printk(KERN_CRIT "Unknown IPI on CPU %d: %lu\n", this_cpu, which);
+ case IPI_CALL_FUNC:
+ generic_smp_call_function_interrupt();
+ break;
+ case IPI_CALL_FUNC_SINGLE:
+ generic_smp_call_function_single_interrupt();
+ break;
+#ifdef CONFIG_KEXEC
+ case IPI_KDUMP_CPU_STOP:
+ unw_init_running(kdump_cpu_freeze, NULL);
+ break;
+#endif
+ default:
+ printk(KERN_CRIT "Unknown IPI on CPU %d: %lu\n",
+ this_cpu, which);
break;
}
} while (ops);
return IRQ_HANDLED;
}
+
+
/*
- * Called with preeemption disabled.
+ * Called with preemption disabled.
*/
static inline void
send_IPI_single (int dest_cpu, int op)
}
/*
- * Called with preeemption disabled.
+ * Called with preemption disabled.
*/
static inline void
send_IPI_allbutself (int op)
{
unsigned int i;
- for (i = 0; i < NR_CPUS; i++) {
- if (cpu_online(i) && i != smp_processor_id())
+ for_each_online_cpu(i) {
+ if (i != smp_processor_id())
send_IPI_single(i, op);
}
}
/*
- * Called with preeemption disabled.
+ * Called with preemption disabled.
+ */
+static inline void
+send_IPI_mask(cpumask_t mask, int op)
+{
+ unsigned int cpu;
+
+ for_each_cpu_mask(cpu, mask) {
+ send_IPI_single(cpu, op);
+ }
+}
+
+/*
+ * Called with preemption disabled.
*/
static inline void
send_IPI_all (int op)
{
int i;
- for (i = 0; i < NR_CPUS; i++)
- if (cpu_online(i))
- send_IPI_single(i, op);
+ for_each_online_cpu(i) {
+ send_IPI_single(i, op);
+ }
}
/*
- * Called with preeemption disabled.
+ * Called with preemption disabled.
*/
static inline void
send_IPI_self (int op)
send_IPI_single(smp_processor_id(), op);
}
+#ifdef CONFIG_KEXEC
+void
+kdump_smp_send_stop(void)
+{
+ send_IPI_allbutself(IPI_KDUMP_CPU_STOP);
+}
+
+void
+kdump_smp_send_init(void)
+{
+ unsigned int cpu, self_cpu;
+ self_cpu = smp_processor_id();
+ for_each_online_cpu(cpu) {
+ if (cpu != self_cpu) {
+ if(kdump_status[cpu] == 0)
+ platform_send_ipi(cpu, 0, IA64_IPI_DM_INIT, 0);
+ }
+ }
+}
+#endif
/*
- * Called with preeemption disabled.
+ * Called with preemption disabled.
*/
void
smp_send_reschedule (int cpu)
platform_send_ipi(cpu, IA64_IPI_RESCHEDULE, IA64_IPI_DM_INT, 0);
}
+/*
+ * Called with preemption disabled.
+ */
+static void
+smp_send_local_flush_tlb (int cpu)
+{
+ platform_send_ipi(cpu, IA64_IPI_LOCAL_TLB_FLUSH, IA64_IPI_DM_INT, 0);
+}
+
+void
+smp_local_flush_tlb(void)
+{
+ /*
+ * Use atomic ops. Otherwise, the load/increment/store sequence from
+ * a "++" operation can have the line stolen between the load & store.
+ * The overhead of the atomic op in negligible in this case & offers
+ * significant benefit for the brief periods where lots of cpus
+ * are simultaneously flushing TLBs.
+ */
+ ia64_fetchadd(1, &local_tlb_flush_counts[smp_processor_id()].count, acq);
+ local_flush_tlb_all();
+}
+
+#define FLUSH_DELAY 5 /* Usec backoff to eliminate excessive cacheline bouncing */
+
+void
+smp_flush_tlb_cpumask(cpumask_t xcpumask)
+{
+ unsigned short *counts = __ia64_per_cpu_var(shadow_flush_counts);
+ cpumask_t cpumask = xcpumask;
+ int mycpu, cpu, flush_mycpu = 0;
+
+ preempt_disable();
+ mycpu = smp_processor_id();
+
+ for_each_cpu_mask(cpu, cpumask)
+ counts[cpu] = local_tlb_flush_counts[cpu].count & 0xffff;
+
+ mb();
+ for_each_cpu_mask(cpu, cpumask) {
+ if (cpu == mycpu)
+ flush_mycpu = 1;
+ else
+ smp_send_local_flush_tlb(cpu);
+ }
+
+ if (flush_mycpu)
+ smp_local_flush_tlb();
+
+ for_each_cpu_mask(cpu, cpumask)
+ while(counts[cpu] == (local_tlb_flush_counts[cpu].count & 0xffff))
+ udelay(FLUSH_DELAY);
+
+ preempt_enable();
+}
+
void
smp_flush_tlb_all (void)
{
- on_each_cpu((void (*)(void *))local_flush_tlb_all, NULL, 1, 1);
+ on_each_cpu((void (*)(void *))local_flush_tlb_all, NULL, 1);
}
void
smp_flush_tlb_mm (struct mm_struct *mm)
{
+ preempt_disable();
/* this happens for the common case of a single-threaded fork(): */
if (likely(mm == current->active_mm && atomic_read(&mm->mm_users) == 1))
{
local_finish_flush_tlb_mm(mm);
+ preempt_enable();
return;
}
+ preempt_enable();
/*
* We could optimize this further by using mm->cpu_vm_mask to track which CPUs
* have been running in the address space. It's not clear that this is worth the
* anyhow, and once a CPU is interrupted, the cost of local_flush_tlb_all() is
* rather trivial.
*/
- on_each_cpu((void (*)(void *))local_finish_flush_tlb_mm, mm, 1, 1);
+ on_each_cpu((void (*)(void *))local_finish_flush_tlb_mm, mm, 1);
}
-/*
- * Run a function on another CPU
- * <func> The function to run. This must be fast and non-blocking.
- * <info> An arbitrary pointer to pass to the function.
- * <nonatomic> Currently unused.
- * <wait> If true, wait until function has completed on other CPUs.
- * [RETURNS] 0 on success, else a negative status code.
- *
- * Does not return until the remote CPU is nearly ready to execute <func>
- * or is or has executed.
- */
-
-int
-smp_call_function_single (int cpuid, void (*func) (void *info), void *info, int nonatomic,
- int wait)
+void arch_send_call_function_single_ipi(int cpu)
{
- struct call_data_struct data;
- int cpus = 1;
- int me = get_cpu(); /* prevent preemption and reschedule on another processor */
-
- if (cpuid == me) {
- printk(KERN_INFO "%s: trying to call self\n", __FUNCTION__);
- put_cpu();
- return -EBUSY;
- }
-
- data.func = func;
- data.info = info;
- atomic_set(&data.started, 0);
- data.wait = wait;
- if (wait)
- atomic_set(&data.finished, 0);
-
- spin_lock_bh(&call_lock);
-
- call_data = &data;
- mb(); /* ensure store to call_data precedes setting of IPI_CALL_FUNC */
- send_IPI_single(cpuid, IPI_CALL_FUNC);
-
- /* Wait for response */
- while (atomic_read(&data.started) != cpus)
- cpu_relax();
-
- if (wait)
- while (atomic_read(&data.finished) != cpus)
- cpu_relax();
- call_data = NULL;
-
- spin_unlock_bh(&call_lock);
- put_cpu();
- return 0;
+ send_IPI_single(cpu, IPI_CALL_FUNC_SINGLE);
}
-EXPORT_SYMBOL(smp_call_function_single);
-/*
- * this function sends a 'generic call function' IPI to all other CPUs
- * in the system.
- */
-
-/*
- * [SUMMARY] Run a function on all other CPUs.
- * <func> The function to run. This must be fast and non-blocking.
- * <info> An arbitrary pointer to pass to the function.
- * <nonatomic> currently unused.
- * <wait> If true, wait (atomically) until function has completed on other CPUs.
- * [RETURNS] 0 on success, else a negative status code.
- *
- * Does not return until remote CPUs are nearly ready to execute <func> or are or have
- * executed.
- *
- * You must not call this function with disabled interrupts or from a
- * hardware interrupt handler or from a bottom half handler.
- */
-int
-smp_call_function (void (*func) (void *info), void *info, int nonatomic, int wait)
+void arch_send_call_function_ipi(cpumask_t mask)
{
- struct call_data_struct data;
- int cpus = num_online_cpus()-1;
-
- if (!cpus)
- return 0;
-
- /* Can deadlock when called with interrupts disabled */
- WARN_ON(irqs_disabled());
-
- data.func = func;
- data.info = info;
- atomic_set(&data.started, 0);
- data.wait = wait;
- if (wait)
- atomic_set(&data.finished, 0);
-
- spin_lock(&call_lock);
-
- call_data = &data;
- mb(); /* ensure store to call_data precedes setting of IPI_CALL_FUNC */
- send_IPI_allbutself(IPI_CALL_FUNC);
-
- /* Wait for response */
- while (atomic_read(&data.started) != cpus)
- cpu_relax();
-
- if (wait)
- while (atomic_read(&data.finished) != cpus)
- cpu_relax();
- call_data = NULL;
-
- spin_unlock(&call_lock);
- return 0;
+ send_IPI_mask(mask, IPI_CALL_FUNC);
}
-EXPORT_SYMBOL(smp_call_function);
/*
* this function calls the 'stop' function on all other CPUs in the system.
send_IPI_allbutself(IPI_CPU_STOP);
}
-int __init
+int
setup_profiling_timer (unsigned int multiplier)
{
return -EINVAL;
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff