28 files changed, 767 insertions, 180 deletions

diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
index 65d0b72777ea..29dc0c89d4af 100644
--- a/arch/x86/kernel/acpi/boot.c
+++ b/arch/x86/kernel/acpi/boot.c
@@ -538,9 +538,10 @@ static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu)
         struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
         union acpi_object *obj;
         struct acpi_madt_local_apic *lapic;
-        cpumask_t tmp_map, new_map;
+        cpumask_var_t tmp_map, new_map;
         u8 physid;
         int cpu;
+        int retval = -ENOMEM;
 
         if (ACPI_FAILURE(acpi_evaluate_object(handle, "_MAT", NULL, &buffer)))
                 return -EINVAL;
@@ -569,23 +570,37 @@ static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu)
         buffer.length = ACPI_ALLOCATE_BUFFER;
         buffer.pointer = NULL;
 
-        tmp_map = cpu_present_map;
+        if (!alloc_cpumask_var(&tmp_map, GFP_KERNEL))
+                goto out;
+
+        if (!alloc_cpumask_var(&new_map, GFP_KERNEL))
+                goto free_tmp_map;
+
+        cpumask_copy(tmp_map, cpu_present_mask);
         acpi_register_lapic(physid, lapic->lapic_flags & ACPI_MADT_ENABLED);
 
         /*
          * If mp_register_lapic successfully generates a new logical cpu
          * number, then the following will get us exactly what was mapped
          */
-        cpus_andnot(new_map, cpu_present_map, tmp_map);
-        if (cpus_empty(new_map)) {
+        cpumask_andnot(new_map, cpu_present_mask, tmp_map);
+        if (cpumask_empty(new_map)) {
                 printk ("Unable to map lapic to logical cpu number\n");
-                return -EINVAL;
+                retval = -EINVAL;
+                goto free_new_map;
         }
 
-        cpu = first_cpu(new_map);
+        cpu = cpumask_first(new_map);
 
         *pcpu = cpu;
-        return 0;
+        retval = 0;
+
+free_new_map:
+        free_cpumask_var(new_map);
+free_tmp_map:
+        free_cpumask_var(tmp_map);
+out:
+        return retval;
 }
 
 /* wrapper to silence section mismatch warning */
@@ -598,7 +613,7 @@ EXPORT_SYMBOL(acpi_map_lsapic);
 int acpi_unmap_lsapic(int cpu)
 {
         per_cpu(x86_cpu_to_apicid, cpu) = -1;
-        cpu_clear(cpu, cpu_present_map);
+        set_cpu_present(cpu, false);
         num_processors--;
 
         return (0);
diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c
index 2e2da717b350..5113c080f0c4 100644
--- a/arch/x86/kernel/amd_iommu.c
+++ b/arch/x86/kernel/amd_iommu.c
@@ -20,8 +20,12 @@
 #include <linux/pci.h>
 #include <linux/gfp.h>
 #include <linux/bitops.h>
+#include <linux/debugfs.h>
 #include <linux/scatterlist.h>
 #include <linux/iommu-helper.h>
+#ifdef CONFIG_IOMMU_API
+#include <linux/iommu.h>
+#endif
 #include <asm/proto.h>
 #include <asm/iommu.h>
 #include <asm/gart.h>
@@ -38,6 +42,10 @@ static DEFINE_RWLOCK(amd_iommu_devtable_lock);
 static LIST_HEAD(iommu_pd_list);
 static DEFINE_SPINLOCK(iommu_pd_list_lock);
 
+#ifdef CONFIG_IOMMU_API
+static struct iommu_ops amd_iommu_ops;
+#endif
+
 /*
  * general struct to manage commands send to an IOMMU
  */
@@ -47,6 +55,68 @@ struct iommu_cmd {
 
 static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
                              struct unity_map_entry *e);
+static struct dma_ops_domain *find_protection_domain(u16 devid);
+
+
+#ifdef CONFIG_AMD_IOMMU_STATS
+
+/*
+ * Initialization code for statistics collection
+ */
+
+DECLARE_STATS_COUNTER(compl_wait);
+DECLARE_STATS_COUNTER(cnt_map_single);
+DECLARE_STATS_COUNTER(cnt_unmap_single);
+DECLARE_STATS_COUNTER(cnt_map_sg);
+DECLARE_STATS_COUNTER(cnt_unmap_sg);
+DECLARE_STATS_COUNTER(cnt_alloc_coherent);
+DECLARE_STATS_COUNTER(cnt_free_coherent);
+DECLARE_STATS_COUNTER(cross_page);
+DECLARE_STATS_COUNTER(domain_flush_single);
+DECLARE_STATS_COUNTER(domain_flush_all);
+DECLARE_STATS_COUNTER(alloced_io_mem);
+DECLARE_STATS_COUNTER(total_map_requests);
+
+static struct dentry *stats_dir;
+static struct dentry *de_isolate;
+static struct dentry *de_fflush;
+
+static void amd_iommu_stats_add(struct __iommu_counter *cnt)
+{
+        if (stats_dir == NULL)
+                return;
+
+        cnt->dent = debugfs_create_u64(cnt->name, 0444, stats_dir,
+                                       &cnt->value);
+}
+
+static void amd_iommu_stats_init(void)
+{
+        stats_dir = debugfs_create_dir("amd-iommu", NULL);
+        if (stats_dir == NULL)
+                return;
+
+        de_isolate = debugfs_create_bool("isolation", 0444, stats_dir,
+                                         (u32 *)&amd_iommu_isolate);
+
+        de_fflush  = debugfs_create_bool("fullflush", 0444, stats_dir,
+                                         (u32 *)&amd_iommu_unmap_flush);
+
+        amd_iommu_stats_add(&compl_wait);
+        amd_iommu_stats_add(&cnt_map_single);
+        amd_iommu_stats_add(&cnt_unmap_single);
+        amd_iommu_stats_add(&cnt_map_sg);
+        amd_iommu_stats_add(&cnt_unmap_sg);
+        amd_iommu_stats_add(&cnt_alloc_coherent);
+        amd_iommu_stats_add(&cnt_free_coherent);
+        amd_iommu_stats_add(&cross_page);
+        amd_iommu_stats_add(&domain_flush_single);
+        amd_iommu_stats_add(&domain_flush_all);
+        amd_iommu_stats_add(&alloced_io_mem);
+        amd_iommu_stats_add(&total_map_requests);
+}
+
+#endif
 
 /* returns !0 if the IOMMU is caching non-present entries in its TLB */
 static int iommu_has_npcache(struct amd_iommu *iommu)
@@ -189,13 +259,55 @@ static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
         spin_lock_irqsave(&iommu->lock, flags);
         ret = __iommu_queue_command(iommu, cmd);
         if (!ret)
-                iommu->need_sync = 1;
+                iommu->need_sync = true;
         spin_unlock_irqrestore(&iommu->lock, flags);
 
         return ret;
 }
 
 /*
+ * This function waits until an IOMMU has completed a completion
+ * wait command
+ */
+static void __iommu_wait_for_completion(struct amd_iommu *iommu)
+{
+        int ready = 0;
+        unsigned status = 0;
+        unsigned long i = 0;
+
+        INC_STATS_COUNTER(compl_wait);
+
+        while (!ready && (i < EXIT_LOOP_COUNT)) {
+                ++i;
+                /* wait for the bit to become one */
+                status = readl(iommu->mmio_base + MMIO_STATUS_OFFSET);
+                ready = status & MMIO_STATUS_COM_WAIT_INT_MASK;
+        }
+
+        /* set bit back to zero */
+        status &= ~MMIO_STATUS_COM_WAIT_INT_MASK;
+        writel(status, iommu->mmio_base + MMIO_STATUS_OFFSET);
+
+        if (unlikely(i == EXIT_LOOP_COUNT))
+                panic("AMD IOMMU: Completion wait loop failed\n");
+}
+
+/*
+ * This function queues a completion wait command into the command
+ * buffer of an IOMMU
+ */
+static int __iommu_completion_wait(struct amd_iommu *iommu)
+{
+        struct iommu_cmd cmd;
+
+         memset(&cmd, 0, sizeof(cmd));
+         cmd.data[0] = CMD_COMPL_WAIT_INT_MASK;
+         CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT);
+
+         return __iommu_queue_command(iommu, &cmd);
+}
+
+/*
  * This function is called whenever we need to ensure that the IOMMU has
  * completed execution of all commands we sent. It sends a
  * COMPLETION_WAIT command and waits for it to finish. The IOMMU informs
@@ -204,40 +316,22 @@ static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
  */
 static int iommu_completion_wait(struct amd_iommu *iommu)
 {
-        int ret = 0, ready = 0;
-        unsigned status = 0;
-        struct iommu_cmd cmd;
-        unsigned long flags, i = 0;
-
-        memset(&cmd, 0, sizeof(cmd));
-        cmd.data[0] = CMD_COMPL_WAIT_INT_MASK;
-        CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT);
+        int ret = 0;
+        unsigned long flags;
 
         spin_lock_irqsave(&iommu->lock, flags);
 
         if (!iommu->need_sync)
                 goto out;
 
-        iommu->need_sync = 0;
+        ret = __iommu_completion_wait(iommu);
 
-        ret = __iommu_queue_command(iommu, &cmd);
+        iommu->need_sync = false;
 
         if (ret)
                 goto out;
 
-        while (!ready && (i < EXIT_LOOP_COUNT)) {
-                ++i;
-                /* wait for the bit to become one */
-                status = readl(iommu->mmio_base + MMIO_STATUS_OFFSET);
-                ready = status & MMIO_STATUS_COM_WAIT_INT_MASK;
-        }
-
-        /* set bit back to zero */
-        status &= ~MMIO_STATUS_COM_WAIT_INT_MASK;
-        writel(status, iommu->mmio_base + MMIO_STATUS_OFFSET);
-
-        if (unlikely(i == EXIT_LOOP_COUNT))
-                panic("AMD IOMMU: Completion wait loop failed\n");
+        __iommu_wait_for_completion(iommu);
 
 out:
         spin_unlock_irqrestore(&iommu->lock, flags);
@@ -264,6 +358,21 @@ static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
         return ret;
 }
 
+static void __iommu_build_inv_iommu_pages(struct iommu_cmd *cmd, u64 address,
+                                          u16 domid, int pde, int s)
+{
+        memset(cmd, 0, sizeof(*cmd));
+        address &= PAGE_MASK;
+        CMD_SET_TYPE(cmd, CMD_INV_IOMMU_PAGES);
+        cmd->data[1] |= domid;
+        cmd->data[2] = lower_32_bits(address);
+        cmd->data[3] = upper_32_bits(address);
+        if (s) /* size bit - we flush more than one 4kb page */
+                cmd->data[2] |= CMD_INV_IOMMU_PAGES_SIZE_MASK;
+        if (pde) /* PDE bit - we wan't flush everything not only the PTEs */
+                cmd->data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK;
+}
+
 /*
  * Generic command send function for invalidaing TLB entries
  */
@@ -273,16 +382,7 @@ static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
         struct iommu_cmd cmd;
         int ret;
 
-        memset(&cmd, 0, sizeof(cmd));
-        address &= PAGE_MASK;
-        CMD_SET_TYPE(&cmd, CMD_INV_IOMMU_PAGES);
-        cmd.data[1] |= domid;
-        cmd.data[2] = lower_32_bits(address);
-        cmd.data[3] = upper_32_bits(address);
-        if (s) /* size bit - we flush more than one 4kb page */
-                cmd.data[2] |= CMD_INV_IOMMU_PAGES_SIZE_MASK;
-        if (pde) /* PDE bit - we wan't flush everything not only the PTEs */
-                cmd.data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK;
+        __iommu_build_inv_iommu_pages(&cmd, address, domid, pde, s);
 
         ret = iommu_queue_command(iommu, &cmd);
 
@@ -321,9 +421,35 @@ static void iommu_flush_tlb(struct amd_iommu *iommu, u16 domid)
 {
         u64 address = CMD_INV_IOMMU_ALL_PAGES_ADDRESS;
 
+        INC_STATS_COUNTER(domain_flush_single);
+
         iommu_queue_inv_iommu_pages(iommu, address, domid, 0, 1);
 }
 
+/*
+ * This function is used to flush the IO/TLB for a given protection domain
+ * on every IOMMU in the system
+ */
+static void iommu_flush_domain(u16 domid)
+{
+        unsigned long flags;
+        struct amd_iommu *iommu;
+        struct iommu_cmd cmd;
+
+        INC_STATS_COUNTER(domain_flush_all);
+
+        __iommu_build_inv_iommu_pages(&cmd, CMD_INV_IOMMU_ALL_PAGES_ADDRESS,
+                                      domid, 1, 1);
+
+        list_for_each_entry(iommu, &amd_iommu_list, list) {
+                spin_lock_irqsave(&iommu->lock, flags);
+                __iommu_queue_command(iommu, &cmd);
+                __iommu_completion_wait(iommu);
+                __iommu_wait_for_completion(iommu);
+                spin_unlock_irqrestore(&iommu->lock, flags);
+        }
+}
+
 /****************************************************************************
  *
  * The functions below are used the create the page table mappings for
@@ -338,10 +464,10 @@ static void iommu_flush_tlb(struct amd_iommu *iommu, u16 domid)
  * supporting all features of AMD IOMMU page tables like level skipping
  * and full 64 bit address spaces.
  */
-static int iommu_map(struct protection_domain *dom,
-                     unsigned long bus_addr,
-                     unsigned long phys_addr,
-                     int prot)
+static int iommu_map_page(struct protection_domain *dom,
+                          unsigned long bus_addr,
+                          unsigned long phys_addr,
+                          int prot)
 {
         u64 __pte, *pte, *page;
 
@@ -388,6 +514,28 @@ static int iommu_map(struct protection_domain *dom,
         return 0;
 }
 
+static void iommu_unmap_page(struct protection_domain *dom,
+                             unsigned long bus_addr)
+{
+        u64 *pte;
+
+        pte = &dom->pt_root[IOMMU_PTE_L2_INDEX(bus_addr)];
+
+        if (!IOMMU_PTE_PRESENT(*pte))
+                return;
+
+        pte = IOMMU_PTE_PAGE(*pte);
+        pte = &pte[IOMMU_PTE_L1_INDEX(bus_addr)];
+
+        if (!IOMMU_PTE_PRESENT(*pte))
+                return;
+
+        pte = IOMMU_PTE_PAGE(*pte);
+        pte = &pte[IOMMU_PTE_L1_INDEX(bus_addr)];
+
+        *pte = 0;
+}
+
 /*
  * This function checks if a specific unity mapping entry is needed for
  * this specific IOMMU.
@@ -440,7 +588,7 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
 
         for (addr = e->address_start; addr < e->address_end;
              addr += PAGE_SIZE) {
-                ret = iommu_map(&dma_dom->domain, addr, addr, e->prot);
+                ret = iommu_map_page(&dma_dom->domain, addr, addr, e->prot);
                 if (ret)
                         return ret;
                 /*
@@ -571,6 +719,16 @@ static u16 domain_id_alloc(void)
         return id;
 }
 
+static void domain_id_free(int id)
+{
+        unsigned long flags;
+
+        write_lock_irqsave(&amd_iommu_devtable_lock, flags);
+        if (id > 0 && id < MAX_DOMAIN_ID)
+                __clear_bit(id, amd_iommu_pd_alloc_bitmap);
+        write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);
+}
+
 /*
  * Used to reserve address ranges in the aperture (e.g. for exclusion
  * ranges.
@@ -587,12 +745,12 @@ static void dma_ops_reserve_addresses(struct dma_ops_domain *dom,
         iommu_area_reserve(dom->bitmap, start_page, pages);
 }
 
-static void dma_ops_free_pagetable(struct dma_ops_domain *dma_dom)
+static void free_pagetable(struct protection_domain *domain)
 {
         int i, j;
         u64 *p1, *p2, *p3;
 
-        p1 = dma_dom->domain.pt_root;
+        p1 = domain->pt_root;
 
         if (!p1)
                 return;
@@ -613,6 +771,8 @@ static void dma_ops_free_pagetable(struct dma_ops_domain *dma_dom)
         }
 
         free_page((unsigned long)p1);
+
+        domain->pt_root = NULL;
 }
 
 /*
@@ -624,7 +784,7 @@ static void dma_ops_domain_free(struct dma_ops_domain *dom)
         if (!dom)
                 return;
 
-        dma_ops_free_pagetable(dom);
+        free_pagetable(&dom->domain);
 
         kfree(dom->pte_pages);
 
@@ -663,6 +823,7 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
                 goto free_dma_dom;
         dma_dom->domain.mode = PAGE_MODE_3_LEVEL;
         dma_dom->domain.pt_root = (void *)get_zeroed_page(GFP_KERNEL);
+        dma_dom->domain.flags = PD_DMA_OPS_MASK;
         dma_dom->domain.priv = dma_dom;
         if (!dma_dom->domain.pt_root)
                 goto free_dma_dom;
@@ -725,6 +886,15 @@ free_dma_dom:
 }
 
 /*
+ * little helper function to check whether a given protection domain is a
+ * dma_ops domain
+ */
+static bool dma_ops_domain(struct protection_domain *domain)
+{
+        return domain->flags & PD_DMA_OPS_MASK;
+}
+
+/*
  * Find out the protection domain structure for a given PCI device. This
  * will give us the pointer to the page table root for example.
  */
@@ -744,14 +914,15 @@ static struct protection_domain *domain_for_device(u16 devid)
  * If a device is not yet associated with a domain, this function does
  * assigns it visible for the hardware
  */
-static void set_device_domain(struct amd_iommu *iommu,
-                              struct protection_domain *domain,
-                              u16 devid)
+static void attach_device(struct amd_iommu *iommu,
+                          struct protection_domain *domain,
+                          u16 devid)
 {
         unsigned long flags;
-
         u64 pte_root = virt_to_phys(domain->pt_root);
 
+        domain->dev_cnt += 1;
+
         pte_root |= (domain->mode & DEV_ENTRY_MODE_MASK)
                     << DEV_ENTRY_MODE_SHIFT;
         pte_root |= IOMMU_PTE_IR | IOMMU_PTE_IW | IOMMU_PTE_P | IOMMU_PTE_TV;
@@ -767,6 +938,116 @@ static void set_device_domain(struct amd_iommu *iommu,
         iommu_queue_inv_dev_entry(iommu, devid);
 }
 
+/*
+ * Removes a device from a protection domain (unlocked)
+ */
+static void __detach_device(struct protection_domain *domain, u16 devid)
+{
+
+        /* lock domain */
+        spin_lock(&domain->lock);
+
+        /* remove domain from the lookup table */
+        amd_iommu_pd_table[devid] = NULL;
+
+        /* remove entry from the device table seen by the hardware */
+        amd_iommu_dev_table[devid].data[0] = IOMMU_PTE_P | IOMMU_PTE_TV;
+        amd_iommu_dev_table[devid].data[1] = 0;
+        amd_iommu_dev_table[devid].data[2] = 0;
+
+        /* decrease reference counter */
+        domain->dev_cnt -= 1;
+
+        /* ready */
+        spin_unlock(&domain->lock);
+}
+
+/*
+ * Removes a device from a protection domain (with devtable_lock held)
+ */
+static void detach_device(struct protection_domain *domain, u16 devid)
+{
+        unsigned long flags;
+
+        /* lock device table */
+        write_lock_irqsave(&amd_iommu_devtable_lock, flags);
+        __detach_device(domain, devid);
+        write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);
+}
+
+static int device_change_notifier(struct notifier_block *nb,
+                                  unsigned long action, void *data)
+{
+        struct device *dev = data;
+        struct pci_dev *pdev = to_pci_dev(dev);
+        u16 devid = calc_devid(pdev->bus->number, pdev->devfn);
+        struct protection_domain *domain;
+        struct dma_ops_domain *dma_domain;
+        struct amd_iommu *iommu;
+        int order = amd_iommu_aperture_order;
+        unsigned long flags;
+
+        if (devid > amd_iommu_last_bdf)
+                goto out;
+
+        devid = amd_iommu_alias_table[devid];
+
+        iommu = amd_iommu_rlookup_table[devid];
+        if (iommu == NULL)
+                goto out;
+
+        domain = domain_for_device(devid);
+
+        if (domain && !dma_ops_domain(domain))
+                WARN_ONCE(1, "AMD IOMMU WARNING: device %s already bound "
+                          "to a non-dma-ops domain\n", dev_name(dev));
+
+        switch (action) {
+        case BUS_NOTIFY_BOUND_DRIVER:
+                if (domain)
+                        goto out;
+                dma_domain = find_protection_domain(devid);
+                if (!dma_domain)
+                        dma_domain = iommu->default_dom;
+                attach_device(iommu, &dma_domain->domain, devid);
+                printk(KERN_INFO "AMD IOMMU: Using protection domain %d for "
+                       "device %s\n", dma_domain->domain.id, dev_name(dev));
+                break;
+        case BUS_NOTIFY_UNBIND_DRIVER:
+                if (!domain)
+                        goto out;
+                detach_device(domain, devid);
+                break;
+        case BUS_NOTIFY_ADD_DEVICE:
+                /* allocate a protection domain if a device is added */
+                dma_domain = find_protection_domain(devid);
+                if (dma_domain)
+                        goto out;
+                dma_domain = dma_ops_domain_alloc(iommu, order);
+                if (!dma_domain)
+                        goto out;
+                dma_domain->target_dev = devid;
+
+                spin_lock_irqsave(&iommu_pd_list_lock, flags);
+                list_add_tail(&dma_domain->list, &iommu_pd_list);
+                spin_unlock_irqrestore(&iommu_pd_list_lock, flags);
+
+                break;
+        default:
+                goto out;
+        }
+
+        iommu_queue_inv_dev_entry(iommu, devid);
+        iommu_completion_wait(iommu);
+
+out:
+        return 0;
+}
+
+struct notifier_block device_nb = {
+        .notifier_call = device_change_notifier,
+};
+
 /*****************************************************************************
  *
  * The next functions belong to the dma_ops mapping/unmapping code.
@@ -802,7 +1083,6 @@ static struct dma_ops_domain *find_protection_domain(u16 devid)
         list_for_each_entry(entry, &iommu_pd_list, list) {
                 if (entry->target_dev == devid) {
                         ret = entry;
-                        list_del(&ret->list);
                         break;
                 }
         }
@@ -853,14 +1133,13 @@ static int get_device_resources(struct device *dev,
                 if (!dma_dom)
                         dma_dom = (*iommu)->default_dom;
                 *domain = &dma_dom->domain;
-                set_device_domain(*iommu, *domain, *bdf);
+                attach_device(*iommu, *domain, *bdf);
                 printk(KERN_INFO "AMD IOMMU: Using protection domain %d for "
-                                "device ", (*domain)->id);
-                print_devid(_bdf, 1);
+                                "device %s\n", (*domain)->id, dev_name(dev));
         }
 
         if (domain_for_device(_bdf) == NULL)
-                set_device_domain(*iommu, *domain, _bdf);
+                attach_device(*iommu, *domain, _bdf);
 
         return 1;
 }
@@ -946,6 +1225,11 @@ static dma_addr_t __map_single(struct device *dev,
         pages = iommu_num_pages(paddr, size, PAGE_SIZE);
         paddr &= PAGE_MASK;
 
+        INC_STATS_COUNTER(total_map_requests);
+
+        if (pages > 1)
+                INC_STATS_COUNTER(cross_page);
+
         if (align)
                 align_mask = (1UL << get_order(size)) - 1;
 
@@ -962,6 +1246,8 @@ static dma_addr_t __map_single(struct device *dev,
         }
         address += offset;
 
+        ADD_STATS_COUNTER(alloced_io_mem, size);
+
         if (unlikely(dma_dom->need_flush && !amd_iommu_unmap_flush)) {
                 iommu_flush_tlb(iommu, dma_dom->domain.id);
                 dma_dom->need_flush = false;
@@ -998,6 +1284,8 @@ static void __unmap_single(struct amd_iommu *iommu,
                 start += PAGE_SIZE;
         }
 
+        SUB_STATS_COUNTER(alloced_io_mem, size);
+
         dma_ops_free_addresses(dma_dom, dma_addr, pages);
 
         if (amd_iommu_unmap_flush || dma_dom->need_flush) {
@@ -1019,6 +1307,8 @@ static dma_addr_t map_single(struct device *dev, phys_addr_t paddr,
         dma_addr_t addr;
         u64 dma_mask;
 
+        INC_STATS_COUNTER(cnt_map_single);
+
         if (!check_device(dev))
                 return bad_dma_address;
 
@@ -1030,6 +1320,9 @@ static dma_addr_t map_single(struct device *dev, phys_addr_t paddr,
                 /* device not handled by any AMD IOMMU */
                 return (dma_addr_t)paddr;
 
+        if (!dma_ops_domain(domain))
+                return bad_dma_address;
+
         spin_lock_irqsave(&domain->lock, flags);
         addr = __map_single(dev, iommu, domain->priv, paddr, size, dir, false,
                             dma_mask);
@@ -1055,11 +1348,16 @@ static void unmap_single(struct device *dev, dma_addr_t dma_addr,
         struct protection_domain *domain;
         u16 devid;
 
+        INC_STATS_COUNTER(cnt_unmap_single);
+
         if (!check_device(dev) ||
             !get_device_resources(dev, &iommu, &domain, &devid))
                 /* device not handled by any AMD IOMMU */
                 return;
 
+        if (!dma_ops_domain(domain))
+                return;
+
         spin_lock_irqsave(&domain->lock, flags);
 
         __unmap_single(iommu, domain->priv, dma_addr, size, dir);
@@ -1104,6 +1402,8 @@ static int map_sg(struct device *dev, struct scatterlist *sglist,
         int mapped_elems = 0;
         u64 dma_mask;
 
+        INC_STATS_COUNTER(cnt_map_sg);
+
         if (!check_device(dev))
                 return 0;
 
@@ -1114,6 +1414,9 @@ static int map_sg(struct device *dev, struct scatterlist *sglist,
         if (!iommu || !domain)
                 return map_sg_no_iommu(dev, sglist, nelems, dir);
 
+        if (!dma_ops_domain(domain))
+                return 0;
+
         spin_lock_irqsave(&domain->lock, flags);
 
         for_each_sg(sglist, s, nelems, i) {
@@ -1163,10 +1466,15 @@ static void unmap_sg(struct device *dev, struct scatterlist *sglist,
         u16 devid;
         int i;
 
+        INC_STATS_COUNTER(cnt_unmap_sg);
+
         if (!check_device(dev) ||
             !get_device_resources(dev, &iommu, &domain, &devid))
                 return;
 
+        if (!dma_ops_domain(domain))
+                return;
+
         spin_lock_irqsave(&domain->lock, flags);
 
         for_each_sg(sglist, s, nelems, i) {
@@ -1194,6 +1502,8 @@ static void *alloc_coherent(struct device *dev, size_t size,
         phys_addr_t paddr;
         u64 dma_mask = dev->coherent_dma_mask;
 
+        INC_STATS_COUNTER(cnt_alloc_coherent);
+
         if (!check_device(dev))
                 return NULL;
 
@@ -1212,6 +1522,9 @@ static void *alloc_coherent(struct device *dev, size_t size,
                 return virt_addr;
         }
 
+        if (!dma_ops_domain(domain))
+                goto out_free;
+
         if (!dma_mask)
                 dma_mask = *dev->dma_mask;
 
@@ -1220,18 +1533,20 @@ static void *alloc_coherent(struct device *dev, size_t size,
         *dma_addr = __map_single(dev, iommu, domain->priv, paddr,
                                  size, DMA_BIDIRECTIONAL, true, dma_mask);
 
-        if (*dma_addr == bad_dma_address) {
-                free_pages((unsigned long)virt_addr, get_order(size));
-                virt_addr = NULL;
-                goto out;
-        }
+        if (*dma_addr == bad_dma_address)
+                goto out_free;
 
         iommu_completion_wait(iommu);
 
-out:
         spin_unlock_irqrestore(&domain->lock, flags);
 
         return virt_addr;
+
+out_free:
+
+        free_pages((unsigned long)virt_addr, get_order(size));
+
+        return NULL;
 }
 
 /*
@@ -1245,6 +1560,8 @@ static void free_coherent(struct device *dev, size_t size,
         struct protection_domain *domain;
         u16 devid;
 
+        INC_STATS_COUNTER(cnt_free_coherent);
+
         if (!check_device(dev))
                 return;
 
@@ -1253,6 +1570,9 @@ static void free_coherent(struct device *dev, size_t size,
         if (!iommu || !domain)
                 goto free_mem;
 
+        if (!dma_ops_domain(domain))
+                goto free_mem;
+
         spin_lock_irqsave(&domain->lock, flags);
 
         __unmap_single(iommu, domain->priv, dma_addr, size, DMA_BIDIRECTIONAL);
@@ -1296,7 +1616,7 @@ static int amd_iommu_dma_supported(struct device *dev, u64 mask)
  * we don't need to preallocate the protection domains anymore.
  * For now we have to.
  */
-void prealloc_protection_domains(void)
+static void prealloc_protection_domains(void)
 {
         struct pci_dev *dev = NULL;
         struct dma_ops_domain *dma_dom;
@@ -1305,7 +1625,7 @@ void prealloc_protection_domains(void)
         u16 devid;
 
         while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-                devid = (dev->bus->number << 8) | dev->devfn;
+                devid = calc_devid(dev->bus->number, dev->devfn);
                 if (devid > amd_iommu_last_bdf)
                         continue;
                 devid = amd_iommu_alias_table[devid];
@@ -1352,6 +1672,7 @@ int __init amd_iommu_init_dma_ops(void)
                 iommu->default_dom = dma_ops_domain_alloc(iommu, order);
                 if (iommu->default_dom == NULL)
                         return -ENOMEM;
+                iommu->default_dom->domain.flags |= PD_DEFAULT_MASK;
                 ret = iommu_init_unity_mappings(iommu);
                 if (ret)
                         goto free_domains;
@@ -1375,6 +1696,12 @@ int __init amd_iommu_init_dma_ops(void)
         /* Make the driver finally visible to the drivers */
         dma_ops = &amd_iommu_dma_ops;
 
+        register_iommu(&amd_iommu_ops);
+
+        bus_register_notifier(&pci_bus_type, &device_nb);
+
+        amd_iommu_stats_init();
+
         return 0;
 
 free_domains:
@@ -1386,3 +1713,224 @@ free_domains:
 
         return ret;
 }
+
+/*****************************************************************************
+ *
+ * The following functions belong to the exported interface of AMD IOMMU
+ *
+ * This interface allows access to lower level functions of the IOMMU
+ * like protection domain handling and assignement of devices to domains
+ * which is not possible with the dma_ops interface.
+ *
+ *****************************************************************************/
+
+static void cleanup_domain(struct protection_domain *domain)
+{
+        unsigned long flags;
+        u16 devid;
+
+        write_lock_irqsave(&amd_iommu_devtable_lock, flags);
+
+        for (devid = 0; devid <= amd_iommu_last_bdf; ++devid)
+                if (amd_iommu_pd_table[devid] == domain)
+                        __detach_device(domain, devid);
+
+        write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);
+}
+
+static int amd_iommu_domain_init(struct iommu_domain *dom)
+{
+        struct protection_domain *domain;
+
+        domain = kzalloc(sizeof(*domain), GFP_KERNEL);
+        if (!domain)
+                return -ENOMEM;
+
+        spin_lock_init(&domain->lock);
+        domain->mode = PAGE_MODE_3_LEVEL;
+        domain->id = domain_id_alloc();
+        if (!domain->id)
+                goto out_free;
+        domain->pt_root = (void *)get_zeroed_page(GFP_KERNEL);
+        if (!domain->pt_root)
+                goto out_free;
+
+        dom->priv = domain;
+
+        return 0;
+
+out_free:
+        kfree(domain);
+
+        return -ENOMEM;
+}
+
+static void amd_iommu_domain_destroy(struct iommu_domain *dom)
+{
+        struct protection_domain *domain = dom->priv;
+
+        if (!domain)
+                return;
+
+        if (domain->dev_cnt > 0)
+                cleanup_domain(domain);
+
+        BUG_ON(domain->dev_cnt != 0);
+
+        free_pagetable(domain);
+
+        domain_id_free(domain->id);
+
+        kfree(domain);
+
+        dom->priv = NULL;
+}
+
+static void amd_iommu_detach_device(struct iommu_domain *dom,
+                                    struct device *dev)
+{
+        struct protection_domain *domain = dom->priv;
+        struct amd_iommu *iommu;
+        struct pci_dev *pdev;
+        u16 devid;
+
+        if (dev->bus != &pci_bus_type)
+                return;
+
+        pdev = to_pci_dev(dev);
+
+        devid = calc_devid(pdev->bus->number, pdev->devfn);
+
+        if (devid > 0)
+                detach_device(domain, devid);
+
+        iommu = amd_iommu_rlookup_table[devid];
+        if (!iommu)
+                return;
+
+        iommu_queue_inv_dev_entry(iommu, devid);
+        iommu_completion_wait(iommu);
+}
+
+static int amd_iommu_attach_device(struct iommu_domain *dom,
+                                   struct device *dev)
+{
+        struct protection_domain *domain = dom->priv;
+        struct protection_domain *old_domain;
+        struct amd_iommu *iommu;
+        struct pci_dev *pdev;
+        u16 devid;
+
+        if (dev->bus != &pci_bus_type)
+                return -EINVAL;
+
+        pdev = to_pci_dev(dev);
+
+        devid = calc_devid(pdev->bus->number, pdev->devfn);
+
+        if (devid >= amd_iommu_last_bdf ||
+                        devid != amd_iommu_alias_table[devid])
+                return -EINVAL;
+
+        iommu = amd_iommu_rlookup_table[devid];
+        if (!iommu)
+                return -EINVAL;
+
+        old_domain = domain_for_device(devid);
+        if (old_domain)
+                return -EBUSY;
+
+        attach_device(iommu, domain, devid);
+
+        iommu_completion_wait(iommu);
+
+        return 0;
+}
+
+static int amd_iommu_map_range(struct iommu_domain *dom,
+                               unsigned long iova, phys_addr_t paddr,
+                               size_t size, int iommu_prot)
+{
+        struct protection_domain *domain = dom->priv;
+        unsigned long i,  npages = iommu_num_pages(paddr, size, PAGE_SIZE);
+        int prot = 0;
+        int ret;
+
+        if (iommu_prot & IOMMU_READ)
+                prot |= IOMMU_PROT_IR;
+        if (iommu_prot & IOMMU_WRITE)
+                prot |= IOMMU_PROT_IW;
+
+        iova  &= PAGE_MASK;
+        paddr &= PAGE_MASK;
+
+        for (i = 0; i < npages; ++i) {
+                ret = iommu_map_page(domain, iova, paddr, prot);
+                if (ret)
+                        return ret;
+
+                iova  += PAGE_SIZE;
+                paddr += PAGE_SIZE;
+        }
+
+        return 0;
+}
+
+static void amd_iommu_unmap_range(struct iommu_domain *dom,
+                                  unsigned long iova, size_t size)
+{
+
+        struct protection_domain *domain = dom->priv;
+        unsigned long i,  npages = iommu_num_pages(iova, size, PAGE_SIZE);
+
+        iova  &= PAGE_MASK;
+
+        for (i = 0; i < npages; ++i) {
+                iommu_unmap_page(domain, iova);
+                iova  += PAGE_SIZE;
+        }
+
+        iommu_flush_domain(domain->id);
+}
+
+static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom,
+                                          unsigned long iova)
+{
+        struct protection_domain *domain = dom->priv;
+        unsigned long offset = iova & ~PAGE_MASK;
+        phys_addr_t paddr;
+        u64 *pte;
+
+        pte = &domain->pt_root[IOMMU_PTE_L2_INDEX(iova)];
+
+        if (!IOMMU_PTE_PRESENT(*pte))
+                return 0;
+
+        pte = IOMMU_PTE_PAGE(*pte);
+        pte = &pte[IOMMU_PTE_L1_INDEX(iova)];
+
+        if (!IOMMU_PTE_PRESENT(*pte))
+                return 0;
+
+        pte = IOMMU_PTE_PAGE(*pte);
+        pte = &pte[IOMMU_PTE_L0_INDEX(iova)];
+
+        if (!IOMMU_PTE_PRESENT(*pte))
+                return 0;
+
+        paddr  = *pte & IOMMU_PAGE_MASK;
+        paddr |= offset;
+
+        return paddr;
+}
+
+static struct iommu_ops amd_iommu_ops = {
+        .domain_init = amd_iommu_domain_init,
+        .domain_destroy = amd_iommu_domain_destroy,
+        .attach_dev = amd_iommu_attach_device,
+        .detach_dev = amd_iommu_detach_device,
+        .map = amd_iommu_map_range,
+        .unmap = amd_iommu_unmap_range,
+        .iova_to_phys = amd_iommu_iova_to_phys,
+};
+
diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c
index c625800c55ca..42c33cebf00f 100644
--- a/arch/x86/kernel/amd_iommu_init.c
+++ b/arch/x86/kernel/amd_iommu_init.c
@@ -122,7 +122,8 @@ u16 amd_iommu_last_bdf;			/* largest PCI device id we have
 LIST_HEAD(amd_iommu_unity_map);         /* a list of required unity mappings
                                            we find in ACPI */
 unsigned amd_iommu_aperture_order = 26; /* size of aperture in power of 2 */
-int amd_iommu_isolate = 1;              /* if 1, device isolation is enabled */
+bool amd_iommu_isolate = true;          /* if true, device isolation is
+                                           enabled */
 bool amd_iommu_unmap_flush;             /* if true, flush on every unmap */
 
 LIST_HEAD(amd_iommu_list);              /* list of all AMD IOMMUs in the
@@ -243,20 +244,16 @@ static void __init iommu_feature_disable(struct amd_iommu *iommu, u8 bit)
 }
 
 /* Function to enable the hardware */
-void __init iommu_enable(struct amd_iommu *iommu)
+static void __init iommu_enable(struct amd_iommu *iommu)
 {
-        printk(KERN_INFO "AMD IOMMU: Enabling IOMMU "
-               "at %02x:%02x.%x cap 0x%hx\n",
-               iommu->dev->bus->number,
-               PCI_SLOT(iommu->dev->devfn),
-               PCI_FUNC(iommu->dev->devfn),
-               iommu->cap_ptr);
+        printk(KERN_INFO "AMD IOMMU: Enabling IOMMU at %s cap 0x%hx\n",
+               dev_name(&iommu->dev->dev), iommu->cap_ptr);
 
         iommu_feature_enable(iommu, CONTROL_IOMMU_EN);
 }
 
 /* Function to enable IOMMU event logging and event interrupts */
-void __init iommu_enable_event_logging(struct amd_iommu *iommu)
+static void __init iommu_enable_event_logging(struct amd_iommu *iommu)
 {
         iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN);
         iommu_feature_enable(iommu, CONTROL_EVT_INT_EN);
@@ -1218,9 +1215,9 @@ static int __init parse_amd_iommu_options(char *str)
 {
         for (; *str; ++str) {
                 if (strncmp(str, "isolate", 7) == 0)
-                        amd_iommu_isolate = 1;
+                        amd_iommu_isolate = true;
                 if (strncmp(str, "share", 5) == 0)
-                        amd_iommu_isolate = 0;
+                        amd_iommu_isolate = false;
                 if (strncmp(str, "fullflush", 9) == 0)
                         amd_iommu_unmap_flush = true;
         }
diff --git a/arch/x86/kernel/apic.c b/arch/x86/kernel/apic.c
index 6b7f824db160..b13d3c4dbd42 100644
--- a/arch/x86/kernel/apic.c
+++ b/arch/x86/kernel/apic.c
@@ -98,8 +98,8 @@ __setup("apicpmtimer", setup_apicpmtimer);
 #ifdef HAVE_X2APIC
 int x2apic;
 /* x2apic enabled before OS handover */
-int x2apic_preenabled;
-int disable_x2apic;
+static int x2apic_preenabled;
+static int disable_x2apic;
 static __init int setup_nox2apic(char *str)
 {
         disable_x2apic = 1;
@@ -140,7 +140,7 @@ static int lapic_next_event(unsigned long delta,
                             struct clock_event_device *evt);
 static void lapic_timer_setup(enum clock_event_mode mode,
                               struct clock_event_device *evt);
-static void lapic_timer_broadcast(const cpumask_t *mask);
+static void lapic_timer_broadcast(const struct cpumask *mask);
 static void apic_pm_activate(void);
 
 /*
@@ -226,7 +226,7 @@ void xapic_icr_write(u32 low, u32 id)
         apic_write(APIC_ICR, low);
 }
 
-u64 xapic_icr_read(void)
+static u64 xapic_icr_read(void)
 {
         u32 icr1, icr2;
 
@@ -266,7 +266,7 @@ void x2apic_icr_write(u32 low, u32 id)
         wrmsrl(APIC_BASE_MSR + (APIC_ICR >> 4), ((__u64) id) << 32 | low);
 }
 
-u64 x2apic_icr_read(void)
+static u64 x2apic_icr_read(void)
 {
         unsigned long val;
 
@@ -453,7 +453,7 @@ static void lapic_timer_setup(enum clock_event_mode mode,
 /*
  * Local APIC timer broadcast function
  */
-static void lapic_timer_broadcast(const cpumask_t *mask)
+static void lapic_timer_broadcast(const struct cpumask *mask)
 {
 #ifdef CONFIG_SMP
         send_IPI_mask(mask, LOCAL_TIMER_VECTOR);
diff --git a/arch/x86/kernel/bios_uv.c b/arch/x86/kernel/bios_uv.c
index 2a0a2a3cac26..f63882728d91 100644
--- a/arch/x86/kernel/bios_uv.c
+++ b/arch/x86/kernel/bios_uv.c
@@ -25,7 +25,7 @@
 #include <asm/uv/bios.h>
 #include <asm/uv/uv_hub.h>
 
-struct uv_systab uv_systab;
+static struct uv_systab uv_systab;
 
 s64 uv_bios_call(enum uv_bios_cmd which, u64 a1, u64 a2, u64 a3, u64 a4, u64 a5)
 {
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 42e0853030cb..3f95a40f718a 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -355,7 +355,7 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c)
                 printk(KERN_INFO  "CPU: Hyper-Threading is disabled\n");
         } else if (smp_num_siblings > 1) {
 
-                if (smp_num_siblings > NR_CPUS) {
+                if (smp_num_siblings > nr_cpu_ids) {
                         printk(KERN_WARNING "CPU: Unsupported number of siblings %d",
                                         smp_num_siblings);
                         smp_num_siblings = 1;
diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
index 88ea02dcb622..28102ad1a363 100644
--- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -517,6 +517,17 @@ acpi_cpufreq_guess_freq(struct acpi_cpufreq_data *data, unsigned int cpu)
         }
 }
 
+static void free_acpi_perf_data(void)
+{
+        unsigned int i;
+
+        /* Freeing a NULL pointer is OK, and alloc_percpu zeroes. */
+        for_each_possible_cpu(i)
+                free_cpumask_var(per_cpu_ptr(acpi_perf_data, i)
+                                 ->shared_cpu_map);
+        free_percpu(acpi_perf_data);
+}
+
 /*
  * acpi_cpufreq_early_init - initialize ACPI P-States library
  *
@@ -527,6 +538,7 @@ acpi_cpufreq_guess_freq(struct acpi_cpufreq_data *data, unsigned int cpu)
  */
 static int __init acpi_cpufreq_early_init(void)
 {
+        unsigned int i;
         dprintk("acpi_cpufreq_early_init\n");
 
         acpi_perf_data = alloc_percpu(struct acpi_processor_performance);
@@ -534,6 +546,16 @@ static int __init acpi_cpufreq_early_init(void)
                 dprintk("Memory allocation error for acpi_perf_data.\n");
                 return -ENOMEM;
         }
+        for_each_possible_cpu(i) {
+                if (!alloc_cpumask_var_node(
+                        &per_cpu_ptr(acpi_perf_data, i)->shared_cpu_map,
+                        GFP_KERNEL, cpu_to_node(i))) {
+
+                        /* Freeing a NULL pointer is OK: alloc_percpu zeroes. */
+                        free_acpi_perf_data();
+                        return -ENOMEM;
+                }
+        }
 
         /* Do initialization in ACPI core */
         acpi_processor_preregister_performance(acpi_perf_data);
@@ -604,9 +626,9 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy)
          */
         if (policy->shared_type == CPUFREQ_SHARED_TYPE_ALL ||
             policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) {
-                policy->cpus = perf->shared_cpu_map;
+                cpumask_copy(&policy->cpus, perf->shared_cpu_map);
         }
-        policy->related_cpus = perf->shared_cpu_map;
+        cpumask_copy(&policy->related_cpus, perf->shared_cpu_map);
 
 #ifdef CONFIG_SMP
         dmi_check_system(sw_any_bug_dmi_table);
@@ -795,7 +817,7 @@ static int __init acpi_cpufreq_init(void)
 
         ret = cpufreq_register_driver(&acpi_cpufreq_driver);
         if (ret)
-                free_percpu(acpi_perf_data);
+                free_acpi_perf_data();
 
         return ret;
 }
diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k7.c b/arch/x86/kernel/cpu/cpufreq/powernow-k7.c
index 7c7d56b43136..1b446d79a8fd 100644
--- a/arch/x86/kernel/cpu/cpufreq/powernow-k7.c
+++ b/arch/x86/kernel/cpu/cpufreq/powernow-k7.c
@@ -310,6 +310,12 @@ static int powernow_acpi_init(void)
                 goto err0;
         }
 
+        if (!alloc_cpumask_var(&acpi_processor_perf->shared_cpu_map,
+                                                                GFP_KERNEL)) {
+                retval = -ENOMEM;
+                goto err05;
+        }
+
         if (acpi_processor_register_performance(acpi_processor_perf, 0)) {
                 retval = -EIO;
                 goto err1;
@@ -412,6 +418,8 @@ static int powernow_acpi_init(void)
 err2:
         acpi_processor_unregister_performance(acpi_processor_perf, 0);
 err1:
+        free_cpumask_var(acpi_processor_perf->shared_cpu_map);
+err05:
         kfree(acpi_processor_perf);
 err0:
         printk(KERN_WARNING PFX "ACPI perflib can not be used in this platform\n");
@@ -652,6 +660,7 @@ static int powernow_cpu_exit (struct cpufreq_policy *policy) {
 #ifdef CONFIG_X86_POWERNOW_K7_ACPI
         if (acpi_processor_perf) {
                 acpi_processor_unregister_performance(acpi_processor_perf, 0);
+                free_cpumask_var(acpi_processor_perf->shared_cpu_map);
                 kfree(acpi_processor_perf);
         }
 #endif
diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
index 7f05f44b97e9..c3c9adbaa26f 100644
--- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
+++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
@@ -766,7 +766,7 @@ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned
 static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
 {
         struct cpufreq_frequency_table *powernow_table;
-        int ret_val;
+        int ret_val = -ENODEV;
 
         if (acpi_processor_register_performance(&data->acpi_data, data->cpu)) {
                 dprintk("register performance failed: bad ACPI data\n");
@@ -815,6 +815,13 @@ static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data)
         /* notify BIOS that we exist */
         acpi_processor_notify_smm(THIS_MODULE);
 
+        if (!alloc_cpumask_var(&data->acpi_data.shared_cpu_map, GFP_KERNEL)) {
+                printk(KERN_ERR PFX
+                                "unable to alloc powernow_k8_data cpumask\n");
+                ret_val = -ENOMEM;
+                goto err_out_mem;
+        }
+
         return 0;
 
 err_out_mem:
@@ -826,7 +833,7 @@ err_out:
         /* data->acpi_data.state_count informs us at ->exit() whether ACPI was used */
         data->acpi_data.state_count = 0;
 
-        return -ENODEV;
+        return ret_val;
 }
 
 static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table)
@@ -929,6 +936,7 @@ static void powernow_k8_cpu_exit_acpi(struct powernow_k8_data *data)
 {
         if (data->acpi_data.state_count)
                 acpi_processor_unregister_performance(&data->acpi_data, data->cpu);
+        free_cpumask_var(data->acpi_data.shared_cpu_map);
 }
 
 #else
@@ -1134,7 +1142,8 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
         data->cpu = pol->cpu;
         data->currpstate = HW_PSTATE_INVALID;
 
-        if (powernow_k8_cpu_init_acpi(data)) {
+        rc = powernow_k8_cpu_init_acpi(data);
+        if (rc) {
                 /*
                  * Use the PSB BIOS structure. This is only availabe on
                  * an UP version, and is deprecated by AMD.
@@ -1152,20 +1161,17 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
                                "ACPI maintainers and complain to your BIOS "
                                "vendor.\n");
 #endif
-                        kfree(data);
-                        return -ENODEV;
+                        goto err_out;
                 }
                 if (pol->cpu != 0) {
                         printk(KERN_ERR FW_BUG PFX "No ACPI _PSS objects for "
                                "CPU other than CPU0. Complain to your BIOS "
                                "vendor.\n");
-                        kfree(data);
-                        return -ENODEV;
+                        goto err_out;
                 }
                 rc = find_psb_table(data);
                 if (rc) {
-                        kfree(data);
-                        return -ENODEV;
+                        goto err_out;
                 }
         }
 
diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index c6ecda64f5f1..48533d77be78 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -534,7 +534,7 @@ static void __cpuinit free_cache_attributes(unsigned int cpu)
         per_cpu(cpuid4_info, cpu) = NULL;
 }
 
-static void get_cpu_leaves(void *_retval)
+static void __cpuinit get_cpu_leaves(void *_retval)
 {
         int j, *retval = _retval, cpu = smp_processor_id();
 
diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index d6ec7ec30274..d259e5d2e054 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -824,16 +824,14 @@ static int enable_mtrr_cleanup __initdata =
 
 static int __init disable_mtrr_cleanup_setup(char *str)
 {
-        if (enable_mtrr_cleanup != -1)
-                enable_mtrr_cleanup = 0;
+        enable_mtrr_cleanup = 0;
         return 0;
 }
 early_param("disable_mtrr_cleanup", disable_mtrr_cleanup_setup);
 
 static int __init enable_mtrr_cleanup_setup(char *str)
 {
-        if (enable_mtrr_cleanup != -1)
-                enable_mtrr_cleanup = 1;
+        enable_mtrr_cleanup = 1;
         return 0;
 }
 early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup);
diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c
index 72cefd1e649b..2ac1f0c2beb3 100644
--- a/arch/x86/kernel/cpuid.c
+++ b/arch/x86/kernel/cpuid.c
@@ -39,10 +39,10 @@
 #include <linux/device.h>
 #include <linux/cpu.h>
 #include <linux/notifier.h>
+#include <linux/uaccess.h>
 
 #include <asm/processor.h>
 #include <asm/msr.h>
-#include <asm/uaccess.h>
 #include <asm/system.h>
 
 static struct class *cpuid_class;
@@ -82,7 +82,7 @@ static loff_t cpuid_seek(struct file *file, loff_t offset, int orig)
 }
 
 static ssize_t cpuid_read(struct file *file, char __user *buf,
-                          size_t count, loff_t * ppos)
+                          size_t count, loff_t *ppos)
 {
         char __user *tmp = buf;
         struct cpuid_regs cmd;
@@ -117,11 +117,11 @@ static int cpuid_open(struct inode *inode, struct file *file)
         unsigned int cpu;
         struct cpuinfo_x86 *c;
         int ret = 0;
-        
+
         lock_kernel();
 
         cpu = iminor(file->f_path.dentry->d_inode);
-        if (cpu >= NR_CPUS || !cpu_online(cpu)) {
+        if (cpu >= nr_cpu_ids || !cpu_online(cpu)) {
                 ret = -ENXIO;   /* No such CPU */
                 goto out;
         }
diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c
index 23b138e31e9c..504ad198e4ad 100644
--- a/arch/x86/kernel/early_printk.c
+++ b/arch/x86/kernel/early_printk.c
@@ -886,7 +886,7 @@ asmlinkage void early_printk(const char *fmt, ...)
         va_list ap;
 
         va_start(ap, fmt);
-        n = vscnprintf(buf, 512, fmt, ap);
+        n = vscnprintf(buf, sizeof(buf), fmt, ap);
         early_console->write(early_console, buf, n);
         va_end(ap);
 }
diff --git a/arch/x86/kernel/genx2apic_phys.c b/arch/x86/kernel/genx2apic_phys.c
index 62895cf315ff..21bcc0e098ba 100644
--- a/arch/x86/kernel/genx2apic_phys.c
+++ b/arch/x86/kernel/genx2apic_phys.c
@@ -161,12 +161,12 @@ static unsigned int phys_pkg_id(int index_msb)
         return current_cpu_data.initial_apicid >> index_msb;
 }
 
-void x2apic_send_IPI_self(int vector)
+static void x2apic_send_IPI_self(int vector)
 {
         apic_write(APIC_SELF_IPI, vector);
 }
 
-void init_x2apic_ldr(void)
+static void init_x2apic_ldr(void)
 {
         return;
 }
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index 388e05a5fc17..b9a4d8c4b935 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -27,7 +27,7 @@
 #include <asm/trampoline.h>
 
 /* boot cpu pda */
-static struct x8664_pda _boot_cpu_pda __read_mostly;
+static struct x8664_pda _boot_cpu_pda;
 
 #ifdef CONFIG_SMP
 /*
diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c
index 62ecfc991e1e..3639442aa7a4 100644
--- a/arch/x86/kernel/io_apic.c
+++ b/arch/x86/kernel/io_apic.c
@@ -214,11 +214,11 @@ static struct irq_cfg *get_one_free_irq_cfg(int cpu)
 
         cfg = kzalloc_node(sizeof(*cfg), GFP_ATOMIC, node);
         if (cfg) {
-                /* FIXME: needs alloc_cpumask_var_node() */
-                if (!alloc_cpumask_var(&cfg->domain, GFP_ATOMIC)) {
+                if (!alloc_cpumask_var_node(&cfg->domain, GFP_ATOMIC, node)) {
                         kfree(cfg);
                         cfg = NULL;
-                } else if (!alloc_cpumask_var(&cfg->old_domain, GFP_ATOMIC)) {
+                } else if (!alloc_cpumask_var_node(&cfg->old_domain,
+                                                          GFP_ATOMIC, node)) {
                         free_cpumask_var(cfg->domain);
                         kfree(cfg);
                         cfg = NULL;
@@ -706,7 +706,7 @@ static void __unmask_IO_APIC_irq(struct irq_cfg *cfg)
 }
 
 #ifdef CONFIG_X86_64
-void io_apic_sync(struct irq_pin_list *entry)
+static void io_apic_sync(struct irq_pin_list *entry)
 {
         /*
          * Synchronize the IO-APIC and the CPU by doing
diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c
index eee32b43fee3..71f1d99a635d 100644
--- a/arch/x86/kernel/ldt.c
+++ b/arch/x86/kernel/ldt.c
@@ -12,8 +12,8 @@
 #include <linux/mm.h>
 #include <linux/smp.h>
 #include <linux/vmalloc.h>
+#include <linux/uaccess.h>
 
-#include <asm/uaccess.h>
 #include <asm/system.h>
 #include <asm/ldt.h>
 #include <asm/desc.h>
@@ -93,7 +93,7 @@ static inline int copy_ldt(mm_context_t *new, mm_context_t *old)
         if (err < 0)
                 return err;
 
-        for(i = 0; i < old->size; i++)
+        for (i = 0; i < old->size; i++)
                 write_ldt_entry(new->ldt, i, old->ldt + i * LDT_ENTRY_SIZE);
         return 0;
 }
diff --git a/arch/x86/kernel/mmconf-fam10h_64.c b/arch/x86/kernel/mmconf-fam10h_64.c
index efc2f361fe85..666e43df51f9 100644
--- a/arch/x86/kernel/mmconf-fam10h_64.c
+++ b/arch/x86/kernel/mmconf-fam10h_64.c
@@ -13,8 +13,7 @@
 #include <asm/msr.h>
 #include <asm/acpi.h>
 #include <asm/mmconfig.h>
-
-#include "../pci/pci.h"
+#include <asm/pci_x86.h>
 
 struct pci_hostbridge_probe {
         u32 bus;
diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c
index 45e3b69808ba..c5c5b8df1dbc 100644
--- a/arch/x86/kernel/mpparse.c
+++ b/arch/x86/kernel/mpparse.c
@@ -16,14 +16,14 @@
 #include <linux/bitops.h>
 #include <linux/acpi.h>
 #include <linux/module.h>
+#include <linux/smp.h>
+#include <linux/acpi.h>
 
-#include <asm/smp.h>
 #include <asm/mtrr.h>
 #include <asm/mpspec.h>
 #include <asm/pgalloc.h>
 #include <asm/io_apic.h>
 #include <asm/proto.h>
-#include <asm/acpi.h>
 #include <asm/bios_ebda.h>
 #include <asm/e820.h>
 #include <asm/trampoline.h>
@@ -95,8 +95,8 @@ static void __init MP_bus_info(struct mpc_config_bus *m)
 #endif
 
         if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA) - 1) == 0) {
-                 set_bit(m->mpc_busid, mp_bus_not_pci);
-#if defined(CONFIG_EISA) || defined (CONFIG_MCA)
+                set_bit(m->mpc_busid, mp_bus_not_pci);
+#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
                 mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
 #endif
         } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) {
@@ -104,7 +104,7 @@ static void __init MP_bus_info(struct mpc_config_bus *m)
                         x86_quirks->mpc_oem_pci_bus(m);
 
                 clear_bit(m->mpc_busid, mp_bus_not_pci);
-#if defined(CONFIG_EISA) || defined (CONFIG_MCA)
+#if defined(CONFIG_EISA) || defined(CONFIG_MCA)
                 mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI;
         } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA) - 1) == 0) {
                 mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA;
diff --git a/arch/x86/kernel/msr.c b/arch/x86/kernel/msr.c
index 82a7c7ed6d45..726266695b2c 100644
--- a/arch/x86/kernel/msr.c
+++ b/arch/x86/kernel/msr.c
@@ -136,7 +136,7 @@ static int msr_open(struct inode *inode, struct file *file)
         lock_kernel();
         cpu = iminor(file->f_path.dentry->d_inode);
 
-        if (cpu >= NR_CPUS || !cpu_online(cpu)) {
+        if (cpu >= nr_cpu_ids || !cpu_online(cpu)) {
                 ret = -ENXIO;   /* No such CPU */
                 goto out;
         }
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c
index 8bd1bf9622a7..45a09ccdc214 100644
--- a/arch/x86/kernel/nmi.c
+++ b/arch/x86/kernel/nmi.c
@@ -26,11 +26,10 @@
 #include <linux/kernel_stat.h>
 #include <linux/kdebug.h>
 #include <linux/smp.h>
+#include <linux/nmi.h>
 
 #include <asm/i8259.h>
 #include <asm/io_apic.h>
-#include <asm/smp.h>
-#include <asm/nmi.h>
 #include <asm/proto.h>
 #include <asm/timer.h>
 
diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
index a35eaa379ff6..00c2bcd41463 100644
--- a/arch/x86/kernel/pci-gart_64.c
+++ b/arch/x86/kernel/pci-gart_64.c
@@ -52,7 +52,7 @@ static u32 *iommu_gatt_base;		/* Remapping table */
  * to trigger bugs with some popular PCI cards, in particular 3ware (but
  * has been also also seen with Qlogic at least).
  */
-int iommu_fullflush = 1;
+static int iommu_fullflush = 1;
 
 /* Allocation bitmap for the remapping area: */
 static DEFINE_SPINLOCK(iommu_bitmap_lock);
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index 39643b1df061..2b46eb41643b 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -12,6 +12,7 @@
 #include <asm/proto.h>
 #include <asm/reboot_fixups.h>
 #include <asm/reboot.h>
+#include <asm/pci_x86.h>
 #include <asm/virtext.h>
 
 #ifdef CONFIG_X86_32
@@ -24,7 +25,6 @@
 
 #include <mach_ipi.h>
 
-
 /*
  * Power off function, if any
  */
@@ -501,7 +501,7 @@ void native_machine_shutdown(void)
 
 #ifdef CONFIG_X86_32
         /* See if there has been given a command line override */
-        if ((reboot_cpu != -1) && (reboot_cpu < NR_CPUS) &&
+        if ((reboot_cpu != -1) && (reboot_cpu < nr_cpu_ids) &&
                 cpu_online(reboot_cpu))
                 reboot_cpu_id = reboot_cpu;
 #endif
@@ -511,7 +511,7 @@ void native_machine_shutdown(void)
                 reboot_cpu_id = smp_processor_id();
 
         /* Make certain I only run on the appropriate processor */
-        set_cpus_allowed_ptr(current, &cpumask_of_cpu(reboot_cpu_id));
+        set_cpus_allowed_ptr(current, cpumask_of(reboot_cpu_id));
 
         /* O.K Now that I'm on the appropriate processor,
          * stop all of the others.
diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c
index 0b63b08e7530..a4b619c33106 100644
--- a/arch/x86/kernel/setup_percpu.c
+++ b/arch/x86/kernel/setup_percpu.c
@@ -153,12 +153,10 @@ void __init setup_per_cpu_areas(void)
         align = max_t(unsigned long, PAGE_SIZE, align);
         size = roundup(old_size, align);
 
-        printk(KERN_INFO
-                "NR_CPUS:%d nr_cpumask_bits:%d nr_cpu_ids:%d nr_node_ids:%d\n",
+        pr_info("NR_CPUS:%d nr_cpumask_bits:%d nr_cpu_ids:%d nr_node_ids:%d\n",
                 NR_CPUS, nr_cpumask_bits, nr_cpu_ids, nr_node_ids);
 
-        printk(KERN_INFO "PERCPU: Allocating %zd bytes of per cpu data\n",
-                          size);
+        pr_info("PERCPU: Allocating %zd bytes of per cpu data\n", size);
 
         for_each_possible_cpu(cpu) {
 #ifndef CONFIG_NEED_MULTIPLE_NODES
@@ -169,22 +167,15 @@ void __init setup_per_cpu_areas(void)
                 if (!node_online(node) || !NODE_DATA(node)) {
                         ptr = __alloc_bootmem(size, align,
                                          __pa(MAX_DMA_ADDRESS));
-                        printk(KERN_INFO
-                               "cpu %d has no node %d or node-local memory\n",
+                        pr_info("cpu %d has no node %d or node-local memory\n",
                                 cpu, node);
-                        if (ptr)
-                                printk(KERN_DEBUG
-                                        "per cpu data for cpu%d at %016lx\n",
-                                         cpu, __pa(ptr));
-                }
-                else {
+                        pr_debug("per cpu data for cpu%d at %016lx\n",
+                                 cpu, __pa(ptr));
+                } else {
                         ptr = __alloc_bootmem_node(NODE_DATA(node), size, align,
                                                         __pa(MAX_DMA_ADDRESS));
-                        if (ptr)
-                                printk(KERN_DEBUG
-                                        "per cpu data for cpu%d on node%d "
-                                        "at %016lx\n",
-                                        cpu, node, __pa(ptr));
+                        pr_debug("per cpu data for cpu%d on node%d at %016lx\n",
+                                cpu, node, __pa(ptr));
                 }
 #endif
                 per_cpu_offset(cpu) = ptr - __per_cpu_start;
@@ -339,25 +330,25 @@ static const cpumask_t cpu_mask_none;
 /*
  * Returns a pointer to the bitmask of CPUs on Node 'node'.
  */
-const cpumask_t *_node_to_cpumask_ptr(int node)
+const cpumask_t *cpumask_of_node(int node)
 {
         if (node_to_cpumask_map == NULL) {
                 printk(KERN_WARNING
-                        "_node_to_cpumask_ptr(%d): no node_to_cpumask_map!\n",
+                        "cpumask_of_node(%d): no node_to_cpumask_map!\n",
                         node);
                 dump_stack();
                 return (const cpumask_t *)&cpu_online_map;
         }
         if (node >= nr_node_ids) {
                 printk(KERN_WARNING
-                        "_node_to_cpumask_ptr(%d): node > nr_node_ids(%d)\n",
+                        "cpumask_of_node(%d): node > nr_node_ids(%d)\n",
                         node, nr_node_ids);
                 dump_stack();
                 return &cpu_mask_none;
         }
         return &node_to_cpumask_map[node];
 }
-EXPORT_SYMBOL(_node_to_cpumask_ptr);
+EXPORT_SYMBOL(cpumask_of_node);
 
 /*
  * Returns a bitmask of CPUs on Node 'node'.
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 31869bf5fabd..6bd4d9b73870 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -496,7 +496,7 @@ void __cpuinit set_cpu_sibling_map(int cpu)
 }
 
 /* maps the cpu to the sched domain representing multi-core */
-cpumask_t cpu_coregroup_map(int cpu)
+const struct cpumask *cpu_coregroup_mask(int cpu)
 {
         struct cpuinfo_x86 *c = &cpu_data(cpu);
         /*
@@ -504,9 +504,14 @@ cpumask_t cpu_coregroup_map(int cpu)
          * And for power savings, we return cpu_core_map
          */
         if (sched_mc_power_savings || sched_smt_power_savings)
-                return per_cpu(cpu_core_map, cpu);
+                return &per_cpu(cpu_core_map, cpu);
         else
-                return c->llc_shared_map;
+                return &c->llc_shared_map;
+}
+
+cpumask_t cpu_coregroup_map(int cpu)
+{
+        return *cpu_coregroup_mask(cpu);
 }
 
 static void impress_friends(void)
@@ -1149,7 +1154,7 @@ static void __init smp_cpu_index_default(void)
         for_each_possible_cpu(i) {
                 c = &cpu_data(i);
                 /* mark all to hotplug */
-                c->cpu_index = NR_CPUS;
+                c->cpu_index = nr_cpu_ids;
         }
 }
 
@@ -1293,6 +1298,8 @@ __init void prefill_possible_map(void)
         else
                 possible = setup_possible_cpus;
 
+        total_cpus = max_t(int, possible, num_processors + disabled_cpus);
+
         if (possible > CONFIG_NR_CPUS) {
                 printk(KERN_WARNING
                         "%d Processors exceeds NR_CPUS limit of %d\n",
diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c
index 6a00e5faaa74..f885023167e0 100644
--- a/arch/x86/kernel/tlb_uv.c
+++ b/arch/x86/kernel/tlb_uv.c
@@ -582,7 +582,6 @@ static int __init uv_ptc_init(void)
 static struct bau_control * __init uv_table_bases_init(int blade, int node)
 {
         int i;
-        int *ip;
         struct bau_msg_status *msp;
         struct bau_control *bau_tabp;
 
@@ -599,13 +598,6 @@ static struct bau_control * __init uv_table_bases_init(int blade, int node)
                 bau_cpubits_clear(&msp->seen_by, (int)
                                   uv_blade_nr_possible_cpus(blade));
 
-        bau_tabp->watching =
-            kmalloc_node(sizeof(int) * DEST_NUM_RESOURCES, GFP_KERNEL, node);
-        BUG_ON(!bau_tabp->watching);
-
-        for (i = 0, ip = bau_tabp->watching; i < DEST_Q_SIZE; i++, ip++)
-                *ip = 0;
-
         uv_bau_table_bases[blade] = bau_tabp;
 
         return bau_tabp;
@@ -628,7 +620,6 @@ uv_table_bases_finish(int blade, int node, int cur_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;
-                bcp->watching           = bau_tablesp->watching;
                 bcp->msg_statuses       = bau_tablesp->msg_statuses;
                 bcp->descriptor_base    = adp;
         }
diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
index 2d1f4c7e4052..ce6650eb64e9 100644
--- a/arch/x86/kernel/traps.c
+++ b/arch/x86/kernel/traps.c
@@ -292,8 +292,10 @@ dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code)
         tsk->thread.error_code = error_code;
         tsk->thread.trap_no = 8;
 
-        /* This is always a kernel trap and never fixable (and thus must
-           never return). */
+        /*
+         * This is always a kernel trap and never fixable (and thus must
+         * never return).
+         */
         for (;;)
                 die(str, regs, error_code);
 }
@@ -520,9 +522,11 @@ dotraplinkage void __kprobes do_int3(struct pt_regs *regs, long error_code)
 }
 
 #ifdef CONFIG_X86_64
-/* Help handler running on IST stack to switch back to user stack
-   for scheduling or signal handling. The actual stack switch is done in
-   entry.S */
+/*
+ * Help handler running on IST stack to switch back to user stack
+ * for scheduling or signal handling. The actual stack switch is done in
+ * entry.S
+ */
 asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
 {
         struct pt_regs *regs = eregs;
@@ -532,8 +536,10 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs)
         /* Exception from user space */
         else if (user_mode(eregs))
                 regs = task_pt_regs(current);
-        /* Exception from kernel and interrupts are enabled. Move to
-           kernel process stack. */
+        /*
+         * Exception from kernel and interrupts are enabled. Move to
+         * kernel process stack.
+         */
         else if (eregs->flags & X86_EFLAGS_IF)
                 regs = (struct pt_regs *)(eregs->sp -= sizeof(struct pt_regs));
         if (eregs != regs)
@@ -685,12 +691,7 @@ void math_error(void __user *ip)
         cwd = get_fpu_cwd(task);
         swd = get_fpu_swd(task);
 
-        err = swd & ~cwd & 0x3f;
-
-#ifdef CONFIG_X86_32
-        if (!err)
-                return;
-#endif
+        err = swd & ~cwd;
 
         if (err & 0x001) {      /* Invalid op */
                 /*
@@ -708,7 +709,11 @@ void math_error(void __user *ip)
         } else if (err & 0x020) { /* Precision */
                 info.si_code = FPE_FLTRES;
         } else {
-                info.si_code = __SI_FAULT|SI_KERNEL; /* WTF? */
+                /*
+                 * If we're using IRQ 13, or supposedly even some trap 16
+                 * implementations, it's possible we get a spurious trap...
+                 */
+                return;         /* Spurious trap, no error */
         }
         force_sig_info(SIGFPE, &info, task);
 }
diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c
index 15c3e6999182..2b54fe002e94 100644
--- a/arch/x86/kernel/xsave.c
+++ b/arch/x86/kernel/xsave.c
@@ -159,7 +159,7 @@ int save_i387_xstate(void __user *buf)
  * Restore the extended state if present. Otherwise, restore the FP/SSE
  * state.
  */
-int restore_user_xstate(void __user *buf)
+static int restore_user_xstate(void __user *buf)
 {
         struct _fpx_sw_bytes fx_sw_user;
         u64 mask;