12 files changed, 194 insertions, 110 deletions

diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 1248318436e8..de1a50af807b 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -549,7 +549,8 @@ void __init uv_system_init(void)
         unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size;
         int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val;
         int max_pnode = 0;
-        unsigned long mmr_base, present;
+        unsigned long mmr_base, present, paddr;
+        unsigned short pnode_mask;
 
         map_low_mmrs();
 
@@ -592,6 +593,7 @@ void __init uv_system_init(void)
                 }
         }
 
+        pnode_mask = (1 << n_val) - 1;
         node_id.v = uv_read_local_mmr(UVH_NODE_ID);
         gnode_upper = (((unsigned long)node_id.s.node_id) &
                        ~((1 << n_val) - 1)) << m_val;
@@ -615,7 +617,7 @@ void __init uv_system_init(void)
                 uv_cpu_hub_info(cpu)->numa_blade_id = blade;
                 uv_cpu_hub_info(cpu)->blade_processor_id = lcpu;
                 uv_cpu_hub_info(cpu)->pnode = pnode;
-                uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) - 1;
+                uv_cpu_hub_info(cpu)->pnode_mask = pnode_mask;
                 uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1;
                 uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper;
                 uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base;
@@ -631,6 +633,16 @@ void __init uv_system_init(void)
                         lcpu, blade);
         }
 
+        /* Add blade/pnode info for nodes without cpus */
+        for_each_online_node(nid) {
+                if (uv_node_to_blade[nid] >= 0)
+                        continue;
+                paddr = node_start_pfn(nid) << PAGE_SHIFT;
+                pnode = (paddr >> m_val) & pnode_mask;
+                blade = boot_pnode_to_blade(pnode);
+                uv_node_to_blade[nid] = blade;
+        }
+
         map_gru_high(max_pnode);
         map_mmr_high(max_pnode);
         map_config_high(max_pnode);
diff --git a/arch/x86/kernel/bios_uv.c b/arch/x86/kernel/bios_uv.c
index f63882728d91..63a88e1f987d 100644
--- a/arch/x86/kernel/bios_uv.c
+++ b/arch/x86/kernel/bios_uv.c
@@ -182,7 +182,8 @@ void uv_bios_init(void)
         memcpy(&uv_systab, tab, sizeof(struct uv_systab));
         iounmap(tab);
 
-        printk(KERN_INFO "EFI UV System Table Revision %d\n", tab->revision);
+        printk(KERN_INFO "EFI UV System Table Revision %d\n",
+                                        uv_systab.revision);
 }
 #else   /* !CONFIG_EFI */
 
diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
index 837c2c4cc203..208ecf6643df 100644
--- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -65,14 +65,18 @@ enum {
 struct acpi_cpufreq_data {
         struct acpi_processor_performance *acpi_data;
         struct cpufreq_frequency_table *freq_table;
-        unsigned int max_freq;
         unsigned int resume;
         unsigned int cpu_feature;
-        u64 saved_aperf, saved_mperf;
 };
 
 static DEFINE_PER_CPU(struct acpi_cpufreq_data *, drv_data);
 
+struct acpi_msr_data {
+        u64 saved_aperf, saved_mperf;
+};
+
+static DEFINE_PER_CPU(struct acpi_msr_data, msr_data);
+
 DEFINE_TRACE(power_mark);
 
 /* acpi_perf_data is a pointer to percpu data. */
@@ -204,7 +208,13 @@ static void drv_read(struct drv_cmd *cmd)
 
 static void drv_write(struct drv_cmd *cmd)
 {
+        int this_cpu;
+
+        this_cpu = get_cpu();
+        if (cpumask_test_cpu(this_cpu, cmd->mask))
+                do_drv_write(cmd);
         smp_call_function_many(cmd->mask, do_drv_write, cmd, 1);
+        put_cpu();
 }
 
 static u32 get_cur_val(const struct cpumask *mask)
@@ -281,11 +291,11 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy,
                 return 0;
 
         cur.aperf.whole = readin.aperf.whole -
-                                per_cpu(drv_data, cpu)->saved_aperf;
+                                per_cpu(msr_data, cpu).saved_aperf;
         cur.mperf.whole = readin.mperf.whole -
-                                per_cpu(drv_data, cpu)->saved_mperf;
-        per_cpu(drv_data, cpu)->saved_aperf = readin.aperf.whole;
-        per_cpu(drv_data, cpu)->saved_mperf = readin.mperf.whole;
+                                per_cpu(msr_data, cpu).saved_mperf;
+        per_cpu(msr_data, cpu).saved_aperf = readin.aperf.whole;
+        per_cpu(msr_data, cpu).saved_mperf = readin.mperf.whole;
 
 #ifdef __i386__
         /*
@@ -329,7 +339,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy,
 
 #endif
 
-        retval = per_cpu(drv_data, policy->cpu)->max_freq * perf_percent / 100;
+        retval = (policy->cpuinfo.max_freq * perf_percent) / 100;
 
         return retval;
 }
@@ -682,16 +692,11 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy)
         /* Check for high latency (>20uS) from buggy BIOSes, like on T42 */
         if (perf->control_register.space_id == ACPI_ADR_SPACE_FIXED_HARDWARE &&
             policy->cpuinfo.transition_latency > 20 * 1000) {
-                static int print_once;
                 policy->cpuinfo.transition_latency = 20 * 1000;
-                if (!print_once) {
-                        print_once = 1;
-                        printk(KERN_INFO "Capping off P-state tranision latency"
-                                " at 20 uS\n");
-                }
+                        printk_once(KERN_INFO "Capping off P-state tranision"
+                                    " latency at 20 uS\n");
         }
 
-        data->max_freq = perf->states[0].core_frequency * 1000;
         /* table init */
         for (i = 0; i < perf->state_count; i++) {
                 if (i > 0 && perf->states[i].core_frequency >=
@@ -710,6 +715,9 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy)
         if (result)
                 goto err_freqfree;
 
+        if (perf->states[0].core_frequency * 1000 != policy->cpuinfo.max_freq)
+                printk(KERN_WARNING FW_WARN "P-state 0 is not max freq\n");
+
         switch (perf->control_register.space_id) {
         case ACPI_ADR_SPACE_SYSTEM_IO:
                 /* Current speed is unknown and not detectable by IO port */
diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c
index 648b3a2a3a44..3f0019e0a229 100644
--- a/arch/x86/kernel/hpet.c
+++ b/arch/x86/kernel/hpet.c
@@ -722,7 +722,7 @@ static int hpet_cpuhp_notify(struct notifier_block *n,
 /*
  * Clock source related code
  */
-static cycle_t read_hpet(void)
+static cycle_t read_hpet(struct clocksource *cs)
 {
         return (cycle_t)hpet_readl(HPET_COUNTER);
 }
@@ -756,7 +756,7 @@ static int hpet_clocksource_register(void)
         hpet_restart_counter();
 
         /* Verify whether hpet counter works */
-        t1 = read_hpet();
+        t1 = hpet_readl(HPET_COUNTER);
         rdtscll(start);
 
         /*
@@ -770,7 +770,7 @@ static int hpet_clocksource_register(void)
                 rdtscll(now);
         } while ((now - start) < 200000UL);
 
-        if (t1 == read_hpet()) {
+        if (t1 == hpet_readl(HPET_COUNTER)) {
                 printk(KERN_WARNING
                        "HPET counter not counting. HPET disabled\n");
                 return -ENODEV;
diff --git a/arch/x86/kernel/i8253.c b/arch/x86/kernel/i8253.c
index 3475440baa54..c2e0bb0890d4 100644
--- a/arch/x86/kernel/i8253.c
+++ b/arch/x86/kernel/i8253.c
@@ -129,7 +129,7 @@ void __init setup_pit_timer(void)
  * to just read by itself. So use jiffies to emulate a free
  * running counter:
  */
-static cycle_t pit_read(void)
+static cycle_t pit_read(struct clocksource *cs)
 {
         static int old_count;
         static u32 old_jifs;
diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c
index 137f2e8132df..223af43f1526 100644
--- a/arch/x86/kernel/kvmclock.c
+++ b/arch/x86/kernel/kvmclock.c
@@ -77,6 +77,11 @@ static cycle_t kvm_clock_read(void)
         return ret;
 }
 
+static cycle_t kvm_clock_get_cycles(struct clocksource *cs)
+{
+        return kvm_clock_read();
+}
+
 /*
  * If we don't do that, there is the possibility that the guest
  * will calibrate under heavy load - thus, getting a lower lpj -
@@ -107,7 +112,7 @@ static void kvm_get_preset_lpj(void)
 
 static struct clocksource kvm_clock = {
         .name = "kvm-clock",
-        .read = kvm_clock_read,
+        .read = kvm_clock_get_cycles,
         .rating = 400,
         .mask = CLOCKSOURCE_MASK(64),
         .mult = 1 << KVM_SCALE,
diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/microcode_core.c
index a0f3851ef310..98c470c069d1 100644
--- a/arch/x86/kernel/microcode_core.c
+++ b/arch/x86/kernel/microcode_core.c
@@ -108,40 +108,29 @@ struct ucode_cpu_info		ucode_cpu_info[NR_CPUS];
 EXPORT_SYMBOL_GPL(ucode_cpu_info);
 
 #ifdef CONFIG_MICROCODE_OLD_INTERFACE
-struct update_for_cpu {
-        const void __user       *buf;
-        size_t                  size;
-};
-
-static long update_for_cpu(void *_ufc)
-{
-        struct update_for_cpu *ufc = _ufc;
-        int error;
-
-        error = microcode_ops->request_microcode_user(smp_processor_id(),
-                                                      ufc->buf, ufc->size);
-        if (error < 0)
-                return error;
-        if (!error)
-                microcode_ops->apply_microcode(smp_processor_id());
-        return error;
-}
-
 static int do_microcode_update(const void __user *buf, size_t size)
 {
+        cpumask_t old;
         int error = 0;
         int cpu;
-        struct update_for_cpu ufc = { .buf = buf, .size = size };
+
+        old = current->cpus_allowed;
 
         for_each_online_cpu(cpu) {
                 struct ucode_cpu_info *uci = ucode_cpu_info + cpu;
 
                 if (!uci->valid)
                         continue;
-                error = work_on_cpu(cpu, update_for_cpu, &ufc);
+
+                set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu));
+                error = microcode_ops->request_microcode_user(cpu, buf, size);
                 if (error < 0)
-                        break;
+                        goto out;
+                if (!error)
+                        microcode_ops->apply_microcode(cpu);
         }
+out:
+        set_cpus_allowed_ptr(current, &old);
         return error;
 }
 
@@ -391,8 +380,6 @@ static int mc_sysdev_add(struct sys_device *sys_dev)
                 return err;
 
         err = microcode_init_cpu(cpu);
-        if (err)
-                sysfs_remove_group(&sys_dev->kobj, &mc_attr_group);
 
         return err;
 }
diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c
index 34f12e9996ed..221a3853e268 100644
--- a/arch/x86/kernel/pci-swiotlb.c
+++ b/arch/x86/kernel/pci-swiotlb.c
@@ -50,7 +50,7 @@ static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size,
         return swiotlb_alloc_coherent(hwdev, size, dma_handle, flags);
 }
 
-struct dma_map_ops swiotlb_dma_ops = {
+static struct dma_map_ops swiotlb_dma_ops = {
         .mapping_error = swiotlb_dma_mapping_error,
         .alloc_coherent = x86_swiotlb_alloc_coherent,
         .free_coherent = swiotlb_free_coherent,
diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c
index deb5ebb32c3b..ed0c33761e6d 100644
--- a/arch/x86/kernel/tlb_uv.c
+++ b/arch/x86/kernel/tlb_uv.c
@@ -25,6 +25,8 @@ static int			uv_bau_retry_limit __read_mostly;
 
 /* position of pnode (which is nasid>>1): */
 static int                      uv_nshift __read_mostly;
+/* base pnode in this partition */
+static int                      uv_partition_base_pnode __read_mostly;
 
 static unsigned long            uv_mmask __read_mostly;
 
@@ -32,6 +34,34 @@ static DEFINE_PER_CPU(struct ptc_stats, ptcstats);
 static DEFINE_PER_CPU(struct bau_control, bau_control);
 
 /*
+ * Determine the first node on a blade.
+ */
+static int __init blade_to_first_node(int blade)
+{
+        int node, b;
+
+        for_each_online_node(node) {
+                b = uv_node_to_blade_id(node);
+                if (blade == b)
+                        return node;
+        }
+        return -1; /* shouldn't happen */
+}
+
+/*
+ * Determine the apicid of the first cpu on a blade.
+ */
+static int __init blade_to_first_apicid(int blade)
+{
+        int cpu;
+
+        for_each_present_cpu(cpu)
+                if (blade == uv_cpu_to_blade_id(cpu))
+                        return per_cpu(x86_cpu_to_apicid, cpu);
+        return -1;
+}
+
+/*
  * Free a software acknowledge hardware resource by clearing its Pending
  * bit. This will return a reply to the sender.
  * If the message has timed out, a reply has already been sent by the
@@ -67,7 +97,7 @@ static void uv_bau_process_message(struct bau_payload_queue_entry *msg,
         msp = __get_cpu_var(bau_control).msg_statuses + msg_slot;
         cpu = uv_blade_processor_id();
         msg->number_of_cpus =
-            uv_blade_nr_online_cpus(uv_node_to_blade_id(numa_node_id()));
+                uv_blade_nr_online_cpus(uv_node_to_blade_id(numa_node_id()));
         this_cpu_mask = 1UL << cpu;
         if (msp->seen_by.bits & this_cpu_mask)
                 return;
@@ -215,14 +245,14 @@ static int uv_wait_completion(struct bau_desc *bau_desc,
  * Returns @flush_mask if some remote flushing remains to be done. The
  * mask will have some bits still set.
  */
-const struct cpumask *uv_flush_send_and_wait(int cpu, int this_blade,
+const struct cpumask *uv_flush_send_and_wait(int cpu, int this_pnode,
                                              struct bau_desc *bau_desc,
                                              struct cpumask *flush_mask)
 {
         int completion_status = 0;
         int right_shift;
         int tries = 0;
-        int blade;
+        int pnode;
         int bit;
         unsigned long mmr_offset;
         unsigned long index;
@@ -265,8 +295,8 @@ const struct cpumask *uv_flush_send_and_wait(int cpu, int this_blade,
          * use the IPI method of shootdown on them.
          */
         for_each_cpu(bit, flush_mask) {
-                blade = uv_cpu_to_blade_id(bit);
-                if (blade == this_blade)
+                pnode = uv_cpu_to_pnode(bit);
+                if (pnode == this_pnode)
                         continue;
                 cpumask_clear_cpu(bit, flush_mask);
         }
@@ -309,16 +339,16 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
         struct cpumask *flush_mask = __get_cpu_var(uv_flush_tlb_mask);
         int i;
         int bit;
-        int blade;
+        int pnode;
         int uv_cpu;
-        int this_blade;
+        int this_pnode;
         int locals = 0;
         struct bau_desc *bau_desc;
 
         cpumask_andnot(flush_mask, cpumask, cpumask_of(cpu));
 
         uv_cpu = uv_blade_processor_id();
-        this_blade = uv_numa_blade_id();
+        this_pnode = uv_hub_info->pnode;
         bau_desc = __get_cpu_var(bau_control).descriptor_base;
         bau_desc += UV_ITEMS_PER_DESCRIPTOR * uv_cpu;
 
@@ -326,13 +356,14 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
 
         i = 0;
         for_each_cpu(bit, flush_mask) {
-                blade = uv_cpu_to_blade_id(bit);
-                BUG_ON(blade > (UV_DISTRIBUTION_SIZE - 1));
-                if (blade == this_blade) {
+                pnode = uv_cpu_to_pnode(bit);
+                BUG_ON(pnode > (UV_DISTRIBUTION_SIZE - 1));
+                if (pnode == this_pnode) {
                         locals++;
                         continue;
                 }
-                bau_node_set(blade, &bau_desc->distribution);
+                bau_node_set(pnode - uv_partition_base_pnode,
+                                &bau_desc->distribution);
                 i++;
         }
         if (i == 0) {
@@ -350,7 +381,7 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
         bau_desc->payload.address = va;
         bau_desc->payload.sending_cpu = cpu;
 
-        return uv_flush_send_and_wait(uv_cpu, this_blade, bau_desc, flush_mask);
+        return uv_flush_send_and_wait(uv_cpu, this_pnode, bau_desc, flush_mask);
 }
 
 /*
@@ -418,24 +449,58 @@ void uv_bau_message_interrupt(struct pt_regs *regs)
         set_irq_regs(old_regs);
 }
 
+/*
+ * uv_enable_timeouts
+ *
+ * Each target blade (i.e. blades that have cpu's) needs to have
+ * shootdown message timeouts enabled.  The timeout does not cause
+ * an interrupt, but causes an error message to be returned to
+ * the sender.
+ */
 static void uv_enable_timeouts(void)
 {
-        int i;
         int blade;
-        int last_blade;
+        int nblades;
         int pnode;
-        int cur_cpu = 0;
-        unsigned long apicid;
+        unsigned long mmr_image;
 
-        last_blade = -1;
-        for_each_online_node(i) {
-                blade = uv_node_to_blade_id(i);
-                if (blade == last_blade)
+        nblades = uv_num_possible_blades();
+
+        for (blade = 0; blade < nblades; blade++) {
+                if (!uv_blade_nr_possible_cpus(blade))
                         continue;
-                last_blade = blade;
-                apicid = per_cpu(x86_cpu_to_apicid, cur_cpu);
+
                 pnode = uv_blade_to_pnode(blade);
-                cur_cpu += uv_blade_nr_possible_cpus(i);
+                mmr_image =
+                    uv_read_global_mmr64(pnode, UVH_LB_BAU_MISC_CONTROL);
+                /*
+                 * Set the timeout period and then lock it in, in three
+                 * steps; captures and locks in the period.
+                 *
+                 * To program the period, the SOFT_ACK_MODE must be off.
+                 */
+                mmr_image &= ~((unsigned long)1 <<
+                               UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT);
+                uv_write_global_mmr64
+                    (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
+                /*
+                 * Set the 4-bit period.
+                 */
+                mmr_image &= ~((unsigned long)0xf <<
+                        UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT);
+                mmr_image |= (UV_INTD_SOFT_ACK_TIMEOUT_PERIOD <<
+                             UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT);
+                uv_write_global_mmr64
+                    (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
+                /*
+                 * Subsequent reversals of the timebase bit (3) cause an
+                 * immediate timeout of one or all INTD resources as
+                 * indicated in bits 2:0 (7 causes all of them to timeout).
+                 */
+                mmr_image |= ((unsigned long)1 <<
+                              UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT);
+                uv_write_global_mmr64
+                    (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
         }
 }
 
@@ -482,8 +547,7 @@ static int uv_ptc_seq_show(struct seq_file *file, void *data)
                            stat->requestee, stat->onetlb, stat->alltlb,
                            stat->s_retry, stat->d_retry, stat->ptc_i);
                 seq_printf(file, "%lx %ld %ld %ld %ld %ld %ld\n",
-                           uv_read_global_mmr64(uv_blade_to_pnode
-                                        (uv_cpu_to_blade_id(cpu)),
+                           uv_read_global_mmr64(uv_cpu_to_pnode(cpu),
                                         UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE),
                            stat->sflush, stat->dflush,
                            stat->retriesok, stat->nomsg,
@@ -617,16 +681,18 @@ static struct bau_control * __init uv_table_bases_init(int blade, int node)
  * finish the initialization of the per-blade control structures
  */
 static void __init
-uv_table_bases_finish(int blade, int node, int cur_cpu,
+uv_table_bases_finish(int blade,
                       struct bau_control *bau_tablesp,
                       struct bau_desc *adp)
 {
         struct bau_control *bcp;
-        int i;
+        int cpu;
 
-        for (i = cur_cpu; i < cur_cpu + uv_blade_nr_possible_cpus(blade); i++) {
-                bcp = (struct bau_control *)&per_cpu(bau_control, i);
+        for_each_present_cpu(cpu) {
+                if (blade != uv_cpu_to_blade_id(cpu))
+                        continue;
 
+                bcp = (struct bau_control *)&per_cpu(bau_control, cpu);
                 bcp->bau_msg_head       = bau_tablesp->va_queue_first;
                 bcp->va_queue_first     = bau_tablesp->va_queue_first;
                 bcp->va_queue_last      = bau_tablesp->va_queue_last;
@@ -649,11 +715,10 @@ uv_activation_descriptor_init(int node, int pnode)
         struct bau_desc *adp;
         struct bau_desc *ad2;
 
-        adp = (struct bau_desc *)
-            kmalloc_node(16384, GFP_KERNEL, node);
+        adp = (struct bau_desc *)kmalloc_node(16384, GFP_KERNEL, node);
         BUG_ON(!adp);
 
-        pa = __pa((unsigned long)adp);
+        pa = uv_gpa(adp); /* need the real nasid*/
         n = pa >> uv_nshift;
         m = pa & uv_mmask;
 
@@ -667,8 +732,12 @@ uv_activation_descriptor_init(int node, int pnode)
         for (i = 0, ad2 = adp; i < UV_ACTIVATION_DESCRIPTOR_SIZE; i++, ad2++) {
                 memset(ad2, 0, sizeof(struct bau_desc));
                 ad2->header.sw_ack_flag = 1;
-                ad2->header.base_dest_nodeid =
-                    uv_blade_to_pnode(uv_cpu_to_blade_id(0));
+                /*
+                 * base_dest_nodeid is the first node in the partition, so
+                 * the bit map will indicate partition-relative node numbers.
+                 * note that base_dest_nodeid is actually a nasid.
+                 */
+                ad2->header.base_dest_nodeid = uv_partition_base_pnode << 1;
                 ad2->header.command = UV_NET_ENDPOINT_INTD;
                 ad2->header.int_both = 1;
                 /*
@@ -686,6 +755,8 @@ static struct bau_payload_queue_entry * __init
 uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
 {
         struct bau_payload_queue_entry *pqp;
+        unsigned long pa;
+        int pn;
         char *cp;
 
         pqp = (struct bau_payload_queue_entry *) kmalloc_node(
@@ -696,10 +767,14 @@ uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
         cp = (char *)pqp + 31;
         pqp = (struct bau_payload_queue_entry *)(((unsigned long)cp >> 5) << 5);
         bau_tablesp->va_queue_first = pqp;
+        /*
+         * need the pnode of where the memory was really allocated
+         */
+        pa = uv_gpa(pqp);
+        pn = pa >> uv_nshift;
         uv_write_global_mmr64(pnode,
                               UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST,
-                              ((unsigned long)pnode <<
-                               UV_PAYLOADQ_PNODE_SHIFT) |
+                              ((unsigned long)pn << UV_PAYLOADQ_PNODE_SHIFT) |
                               uv_physnodeaddr(pqp));
         uv_write_global_mmr64(pnode, UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL,
                               uv_physnodeaddr(pqp));
@@ -715,8 +790,9 @@ uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
 /*
  * Initialization of each UV blade's structures
  */
-static int __init uv_init_blade(int blade, int node, int cur_cpu)
+static int __init uv_init_blade(int blade)
 {
+        int node;
         int pnode;
         unsigned long pa;
         unsigned long apicid;
@@ -724,16 +800,17 @@ static int __init uv_init_blade(int blade, int node, int cur_cpu)
         struct bau_payload_queue_entry *pqp;
         struct bau_control *bau_tablesp;
 
+        node = blade_to_first_node(blade);
         bau_tablesp = uv_table_bases_init(blade, node);
         pnode = uv_blade_to_pnode(blade);
         adp = uv_activation_descriptor_init(node, pnode);
         pqp = uv_payload_queue_init(node, pnode, bau_tablesp);
-        uv_table_bases_finish(blade, node, cur_cpu, bau_tablesp, adp);
+        uv_table_bases_finish(blade, bau_tablesp, adp);
         /*
          * the below initialization can't be in firmware because the
          * messaging IRQ will be determined by the OS
          */
-        apicid = per_cpu(x86_cpu_to_apicid, cur_cpu);
+        apicid = blade_to_first_apicid(blade);
         pa = uv_read_global_mmr64(pnode, UVH_BAU_DATA_CONFIG);
         if ((pa & 0xff) != UV_BAU_MESSAGE) {
                 uv_write_global_mmr64(pnode, UVH_BAU_DATA_CONFIG,
@@ -748,9 +825,7 @@ static int __init uv_init_blade(int blade, int node, int cur_cpu)
 static int __init uv_bau_init(void)
 {
         int blade;
-        int node;
         int nblades;
-        int last_blade;
         int cur_cpu;
 
         if (!is_uv_system())
@@ -763,29 +838,21 @@ static int __init uv_bau_init(void)
         uv_bau_retry_limit = 1;
         uv_nshift = uv_hub_info->n_val;
         uv_mmask = (1UL << uv_hub_info->n_val) - 1;
-        nblades = 0;
-        last_blade = -1;
-        cur_cpu = 0;
-        for_each_online_node(node) {
-                blade = uv_node_to_blade_id(node);
-                if (blade == last_blade)
-                        continue;
-                last_blade = blade;
-                nblades++;
-        }
+        nblades = uv_num_possible_blades();
+
         uv_bau_table_bases = (struct bau_control **)
             kmalloc(nblades * sizeof(struct bau_control *), GFP_KERNEL);
         BUG_ON(!uv_bau_table_bases);
 
-        last_blade = -1;
-        for_each_online_node(node) {
-                blade = uv_node_to_blade_id(node);
-                if (blade == last_blade)
-                        continue;
-                last_blade = blade;
-                uv_init_blade(blade, node, cur_cpu);
-                cur_cpu += uv_blade_nr_possible_cpus(blade);
-        }
+        uv_partition_base_pnode = 0x7fffffff;
+        for (blade = 0; blade < nblades; blade++)
+                if (uv_blade_nr_possible_cpus(blade) &&
+                        (uv_blade_to_pnode(blade) < uv_partition_base_pnode))
+                        uv_partition_base_pnode = uv_blade_to_pnode(blade);
+        for (blade = 0; blade < nblades; blade++)
+                if (uv_blade_nr_possible_cpus(blade))
+                        uv_init_blade(blade);
+
         alloc_intr_gate(UV_BAU_MESSAGE, uv_bau_message_intr1);
         uv_enable_timeouts();
 
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 7a567ebe6361..d57de05dc430 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -699,7 +699,7 @@ static struct clocksource clocksource_tsc;
  * code, which is necessary to support wrapping clocksources like pm
  * timer.
  */
-static cycle_t read_tsc(void)
+static cycle_t read_tsc(struct clocksource *cs)
 {
         cycle_t ret = (cycle_t)get_cycles();
 
diff --git a/arch/x86/kernel/uv_sysfs.c b/arch/x86/kernel/uv_sysfs.c
index 67f9b9dbf800..36afb98675a4 100644
--- a/arch/x86/kernel/uv_sysfs.c
+++ b/arch/x86/kernel/uv_sysfs.c
@@ -21,6 +21,7 @@
 
 #include <linux/sysdev.h>
 #include <asm/uv/bios.h>
+#include <asm/uv/uv.h>
 
 struct kobject *sgi_uv_kobj;
 
@@ -47,6 +48,9 @@ static int __init sgi_uv_sysfs_init(void)
 {
         unsigned long ret;
 
+        if (!is_uv_system())
+                return -ENODEV;
+
         if (!sgi_uv_kobj)
                 sgi_uv_kobj = kobject_create_and_add("sgi_uv", firmware_kobj);
         if (!sgi_uv_kobj) {
diff --git a/arch/x86/kernel/vmiclock_32.c b/arch/x86/kernel/vmiclock_32.c
index d303369a7bad..2b3eb82efeeb 100644
--- a/arch/x86/kernel/vmiclock_32.c
+++ b/arch/x86/kernel/vmiclock_32.c
@@ -283,7 +283,7 @@ void __devinit vmi_time_ap_init(void)
 /** vmi clocksource */
 static struct clocksource clocksource_vmi;
 
-static cycle_t read_real_cycles(void)
+static cycle_t read_real_cycles(struct clocksource *cs)
 {
         cycle_t ret = (cycle_t)vmi_timer_ops.get_cycle_counter(VMI_CYCLES_REAL);
         return max(ret, clocksource_vmi.cycle_last);