1 files changed, 127 insertions, 103 deletions

diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 7985c5b3f916..7ece815ea637 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -52,6 +52,7 @@
 #include <asm/desc.h>
 #include <asm/nmi.h>
 #include <asm/irq.h>
+#include <asm/idle.h>
 #include <asm/smp.h>
 #include <asm/trampoline.h>
 #include <asm/cpu.h>
@@ -88,7 +89,7 @@ static DEFINE_PER_CPU(struct task_struct *, idle_thread_array);
 #define get_idle_for_cpu(x)      (per_cpu(idle_thread_array, x))
 #define set_idle_for_cpu(x, p)   (per_cpu(idle_thread_array, x) = (p))
 #else
-struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ;
+static struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ;
 #define get_idle_for_cpu(x)      (idle_thread_array[(x)])
 #define set_idle_for_cpu(x, p)   (idle_thread_array[(x)] = (p))
 #endif
@@ -123,13 +124,12 @@ EXPORT_PER_CPU_SYMBOL(cpu_info);
 
 static atomic_t init_deasserted;
 
-static int boot_cpu_logical_apicid;
 
 /* representing cpus for which sibling maps can be computed */
 static cpumask_t cpu_sibling_setup_map;
 
 /* Set if we find a B stepping CPU */
-int __cpuinitdata smp_b_stepping;
+static int __cpuinitdata smp_b_stepping;
 
 #if defined(CONFIG_NUMA) && defined(CONFIG_X86_32)
 
@@ -165,6 +165,8 @@ static void unmap_cpu_to_node(int cpu)
 #endif
 
 #ifdef CONFIG_X86_32
+static int boot_cpu_logical_apicid;
+
 u8 cpu_2_logical_apicid[NR_CPUS] __read_mostly =
                                         { [0 ... NR_CPUS-1] = BAD_APICID };
 
@@ -210,7 +212,7 @@ static void __cpuinit smp_callin(void)
         /*
          * (This works even if the APIC is not enabled.)
          */
-        phys_id = GET_APIC_ID(read_apic_id());
+        phys_id = read_apic_id();
         cpuid = smp_processor_id();
         if (cpu_isset(cpuid, cpu_callin_map)) {
                 panic("%s: phys CPU#%d, CPU#%d already present??\n", __func__,
@@ -257,6 +259,7 @@ static void __cpuinit smp_callin(void)
         end_local_APIC_setup();
         map_cpu_to_logical_apicid();
 
+        notify_cpu_starting(cpuid);
         /*
          * Get our bogomips.
          *
@@ -279,6 +282,8 @@ static void __cpuinit smp_callin(void)
         cpu_set(cpuid, cpu_callin_map);
 }
 
+static int __cpuinitdata unsafe_smp;
+
 /*
  * Activate a secondary processor.
  */
@@ -331,14 +336,17 @@ static void __cpuinit start_secondary(void *unused)
          * does not change while we are assigning vectors to cpus.  Holding
          * this lock ensures we don't half assign or remove an irq from a cpu.
          */
-        ipi_call_lock_irq();
+        ipi_call_lock();
         lock_vector_lock();
         __setup_vector_irq(smp_processor_id());
         cpu_set(smp_processor_id(), cpu_online_map);
         unlock_vector_lock();
-        ipi_call_unlock_irq();
+        ipi_call_unlock();
         per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE;
 
+        /* enable local interrupts */
+        local_irq_enable();
+
         setup_secondary_clock();
 
         wmb();
@@ -391,7 +399,7 @@ static void __cpuinit smp_apply_quirks(struct cpuinfo_x86 *c)
                                 goto valid_k7;
 
                 /* If we get here, not a certified SMP capable AMD system. */
-                add_taint(TAINT_UNSAFE_SMP);
+                unsafe_smp = 1;
         }
 
 valid_k7:
@@ -408,12 +416,10 @@ static void __cpuinit smp_checks(void)
          * Don't taint if we are running SMP kernel on a single non-MP
          * approved Athlon
          */
-        if (tainted & TAINT_UNSAFE_SMP) {
-                if (num_online_cpus())
-                        printk(KERN_INFO "WARNING: This combination of AMD"
-                                "processors is not suitable for SMP.\n");
-                else
-                        tainted &= ~TAINT_UNSAFE_SMP;
+        if (unsafe_smp && num_online_cpus() > 1) {
+                printk(KERN_INFO "WARNING: This combination of AMD"
+                        "processors is not suitable for SMP.\n");
+                add_taint(TAINT_UNSAFE_SMP);
         }
 }
 
@@ -537,10 +543,10 @@ static inline void __inquire_remote_apic(int apicid)
         int timeout;
         u32 status;
 
-        printk(KERN_INFO "Inquiring remote APIC #%d...\n", apicid);
+        printk(KERN_INFO "Inquiring remote APIC 0x%x...\n", apicid);
 
         for (i = 0; i < ARRAY_SIZE(regs); i++) {
-                printk(KERN_INFO "... APIC #%d %s: ", apicid, names[i]);
+                printk(KERN_INFO "... APIC 0x%x %s: ", apicid, names[i]);
 
                 /*
                  * Wait for idle.
@@ -550,8 +556,7 @@ static inline void __inquire_remote_apic(int apicid)
                         printk(KERN_CONT
                                "a previous APIC delivery may have failed\n");
 
-                apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(apicid));
-                apic_write(APIC_ICR, APIC_DM_REMRD | regs[i]);
+                apic_icr_write(APIC_DM_REMRD | regs[i], apicid);
 
                 timeout = 0;
                 do {
@@ -583,11 +588,9 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip)
         int maxlvt;
 
         /* Target chip */
-        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(logical_apicid));
-
         /* Boot on the stack */
         /* Kick the second */
-        apic_write(APIC_ICR, APIC_DM_NMI | APIC_DEST_LOGICAL);
+        apic_icr_write(APIC_DM_NMI | APIC_DEST_LOGICAL, logical_apicid);
 
         pr_debug("Waiting for send to finish...\n");
         send_status = safe_apic_wait_icr_idle();
@@ -596,10 +599,12 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip)
          * Give the other CPU some time to accept the IPI.
          */
         udelay(200);
-        maxlvt = lapic_get_maxlvt();
-        if (maxlvt > 3)                 /* Due to the Pentium erratum 3AP.  */
-                apic_write(APIC_ESR, 0);
-        accept_status = (apic_read(APIC_ESR) & 0xEF);
+        if (APIC_INTEGRATED(apic_version[phys_apicid])) {
+                maxlvt = lapic_get_maxlvt();
+                if (maxlvt > 3)                 /* Due to the Pentium erratum 3AP.  */
+                        apic_write(APIC_ESR, 0);
+                accept_status = (apic_read(APIC_ESR) & 0xEF);
+        }
         pr_debug("NMI sent.\n");
 
         if (send_status)
@@ -640,13 +645,11 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
         /*
          * Turn INIT on target chip
          */
-        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
-
         /*
          * Send IPI
          */
-        apic_write(APIC_ICR,
-                   APIC_INT_LEVELTRIG | APIC_INT_ASSERT | APIC_DM_INIT);
+        apic_icr_write(APIC_INT_LEVELTRIG | APIC_INT_ASSERT | APIC_DM_INIT,
+                       phys_apicid);
 
         pr_debug("Waiting for send to finish...\n");
         send_status = safe_apic_wait_icr_idle();
@@ -656,10 +659,8 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
         pr_debug("Deasserting INIT.\n");
 
         /* Target chip */
-        apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
-
         /* Send IPI */
-        apic_write(APIC_ICR, APIC_INT_LEVELTRIG | APIC_DM_INIT);
+        apic_icr_write(APIC_INT_LEVELTRIG | APIC_DM_INIT, phys_apicid);
 
         pr_debug("Waiting for send to finish...\n");
         send_status = safe_apic_wait_icr_idle();
@@ -702,11 +703,10 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip)
                  */
 
                 /* Target chip */
-                apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid));
-
                 /* Boot on the stack */
                 /* Kick the second */
-                apic_write(APIC_ICR, APIC_DM_STARTUP | (start_eip >> 12));
+                apic_icr_write(APIC_DM_STARTUP | (start_eip >> 12),
+                               phys_apicid);
 
                 /*
                  * Give the other CPU some time to accept the IPI.
@@ -874,7 +874,7 @@ do_rest:
         start_ip = setup_trampoline();
 
         /* So we see what's up   */
-        printk(KERN_INFO "Booting processor %d/%d ip %lx\n",
+        printk(KERN_INFO "Booting processor %d APIC 0x%x ip 0x%lx\n",
                           cpu, apicid, start_ip);
 
         /*
@@ -1175,10 +1175,17 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus)
          * Setup boot CPU information
          */
         smp_store_cpu_info(0); /* Final full version of the data */
+#ifdef CONFIG_X86_32
         boot_cpu_logical_apicid = logical_smp_processor_id();
+#endif
         current_thread_info()->cpu = 0;  /* needed? */
         set_cpu_sibling_map(0);
 
+#ifdef CONFIG_X86_64
+        enable_IR_x2apic();
+        setup_apic_routing();
+#endif
+
         if (smp_sanity_check(max_cpus) < 0) {
                 printk(KERN_INFO "SMP disabled\n");
                 disable_smp();
@@ -1186,9 +1193,9 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus)
         }
 
         preempt_disable();
-        if (GET_APIC_ID(read_apic_id()) != boot_cpu_physical_apicid) {
+        if (read_apic_id() != boot_cpu_physical_apicid) {
                 panic("Boot APIC ID in local APIC unexpected (%d vs %d)",
-                     GET_APIC_ID(read_apic_id()), boot_cpu_physical_apicid);
+                     read_apic_id(), boot_cpu_physical_apicid);
                 /* Or can we switch back to PIC here? */
         }
         preempt_enable();
@@ -1254,39 +1261,6 @@ void __init native_smp_cpus_done(unsigned int max_cpus)
         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_nr(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_nr(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);
-}
-
-static int additional_cpus __initdata = -1;
-
-static __init int setup_additional_cpus(char *s)
-{
-        return s && get_option(&s, &additional_cpus) ? 0 : -EINVAL;
-}
-early_param("additional_cpus", setup_additional_cpus);
-
 /*
  * cpu_possible_map should be static, it cannot change as cpu's
  * are onlined, or offlined. The reason is per-cpu data-structures
@@ -1306,24 +1280,13 @@ early_param("additional_cpus", setup_additional_cpus);
  */
 __init void prefill_possible_map(void)
 {
-        int i;
-        int possible;
+        int i, possible;
 
         /* no processor from mptable or madt */
         if (!num_processors)
                 num_processors = 1;
 
-#ifdef CONFIG_HOTPLUG_CPU
-        if (additional_cpus == -1) {
-                if (disabled_cpus > 0)
-                        additional_cpus = disabled_cpus;
-                else
-                        additional_cpus = 0;
-        }
-#else
-        additional_cpus = 0;
-#endif
-        possible = num_processors + additional_cpus;
+        possible = num_processors + disabled_cpus;
         if (possible > NR_CPUS)
                 possible = NR_CPUS;
 
@@ -1336,6 +1299,31 @@ __init void prefill_possible_map(void)
         nr_cpu_ids = possible;
 }
 
+#ifdef CONFIG_HOTPLUG_CPU
+
+static void remove_siblinginfo(int cpu)
+{
+        int sibling;
+        struct cpuinfo_x86 *c = &cpu_data(cpu);
+
+        for_each_cpu_mask_nr(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_nr(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);
+}
+
 static void __ref remove_cpu_from_maps(int cpu)
 {
         cpu_clear(cpu, cpu_online_map);
@@ -1346,25 +1334,9 @@ static void __ref remove_cpu_from_maps(int cpu)
         numa_remove_cpu(cpu);
 }
 
-int __cpu_disable(void)
+void cpu_disable_common(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
@@ -1382,10 +1354,32 @@ int __cpu_disable(void)
         remove_cpu_from_maps(cpu);
         unlock_vector_lock();
         fixup_irqs(cpu_online_map);
+}
+
+int native_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();
+
+        cpu_disable_common();
         return 0;
 }
 
-void __cpu_die(unsigned int cpu)
+void native_cpu_die(unsigned int cpu)
 {
         /* We don't do anything here: idle task is faking death itself. */
         unsigned int i;
@@ -1402,15 +1396,45 @@ void __cpu_die(unsigned int cpu)
         }
         printk(KERN_ERR "CPU %u didn't die...\n", cpu);
 }
+
+void play_dead_common(void)
+{
+        idle_task_exit();
+        reset_lazy_tlbstate();
+        irq_ctx_exit(raw_smp_processor_id());
+        c1e_remove_cpu(raw_smp_processor_id());
+
+        mb();
+        /* Ack it */
+        __get_cpu_var(cpu_state) = CPU_DEAD;
+
+        /*
+         * With physical CPU hotplug, we should halt the cpu
+         */
+        local_irq_disable();
+}
+
+void native_play_dead(void)
+{
+        play_dead_common();
+        wbinvd_halt();
+}
+
 #else /* ... !CONFIG_HOTPLUG_CPU */
-int __cpu_disable(void)
+int native_cpu_disable(void)
 {
         return -ENOSYS;
 }
 
-void __cpu_die(unsigned int cpu)
+void native_cpu_die(unsigned int cpu)
 {
         /* We said "no" in __cpu_disable */
         BUG();
 }
+
+void native_play_dead(void)
+{
+        BUG();
+}
+
 #endif