11 files changed, 189 insertions, 66 deletions

diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
index 93d4fbfdb724..a4815d316722 100644
--- a/arch/powerpc/mm/fault.c
+++ b/arch/powerpc/mm/fault.c
@@ -81,7 +81,8 @@ static int store_updates_sp(struct pt_regs *regs)
 }
 
 #if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE))
-static void do_dabr(struct pt_regs *regs, unsigned long error_code)
+static void do_dabr(struct pt_regs *regs, unsigned long address,
+                    unsigned long error_code)
 {
         siginfo_t info;
 
@@ -99,7 +100,7 @@ static void do_dabr(struct pt_regs *regs, unsigned long error_code)
         info.si_signo = SIGTRAP;
         info.si_errno = 0;
         info.si_code = TRAP_HWBKPT;
-        info.si_addr = (void __user *)regs->nip;
+        info.si_addr = (void __user *)address;
         force_sig_info(SIGTRAP, &info, current);
 }
 #endif /* !(CONFIG_4xx || CONFIG_BOOKE)*/
@@ -159,7 +160,7 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address,
 #if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE))
         if (error_code & DSISR_DABRMATCH) {
                 /* DABR match */
-                do_dabr(regs, error_code);
+                do_dabr(regs, address, error_code);
                 return 0;
         }
 #endif /* !(CONFIG_4xx || CONFIG_BOOKE)*/
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index a606504678bd..5bb433cbe41b 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -456,7 +456,7 @@ void __init htab_initialize(void)
 
         /* create bolted the linear mapping in the hash table */
         for (i=0; i < lmb.memory.cnt; i++) {
-                base = lmb.memory.region[i].base + KERNELBASE;
+                base = (unsigned long)__va(lmb.memory.region[i].base);
                 size = lmb.memory.region[i].size;
 
                 DBG("creating mapping for region: %lx : %lx\n", base, size);
@@ -498,8 +498,8 @@ void __init htab_initialize(void)
          * for either 4K or 16MB pages.
          */
         if (tce_alloc_start) {
-                tce_alloc_start += KERNELBASE;
-                tce_alloc_end += KERNELBASE;
+                tce_alloc_start = (unsigned long)__va(tce_alloc_start);
+                tce_alloc_end = (unsigned long)__va(tce_alloc_end);
 
                 if (base + size >= tce_alloc_start)
                         tce_alloc_start = base + size + 1;
@@ -644,6 +644,7 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap)
         DBG_LOW(" -> rc=%d\n", rc);
         return rc;
 }
+EXPORT_SYMBOL_GPL(hash_page);
 
 void hash_preload(struct mm_struct *mm, unsigned long ea,
                   unsigned long access, unsigned long trap)
diff --git a/arch/powerpc/mm/hugetlbpage.c b/arch/powerpc/mm/hugetlbpage.c
index 54131b877da3..b51bb28c054b 100644
--- a/arch/powerpc/mm/hugetlbpage.c
+++ b/arch/powerpc/mm/hugetlbpage.c
@@ -549,6 +549,17 @@ fail:
         return addr;
 }
 
+static int htlb_check_hinted_area(unsigned long addr, unsigned long len)
+{
+        struct vm_area_struct *vma;
+
+        vma = find_vma(current->mm, addr);
+        if (!vma || ((addr + len) <= vma->vm_start))
+                return 0;
+
+        return -ENOMEM;
+}
+
 static unsigned long htlb_get_low_area(unsigned long len, u16 segmask)
 {
         unsigned long addr = 0;
@@ -618,15 +629,28 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
         if (!cpu_has_feature(CPU_FTR_16M_PAGE))
                 return -EINVAL;
 
+        /* Paranoia, caller should have dealt with this */
+        BUG_ON((addr + len)  < addr);
+
         if (test_thread_flag(TIF_32BIT)) {
+                /* Paranoia, caller should have dealt with this */
+                BUG_ON((addr + len) > 0x100000000UL);
+
                 curareas = current->mm->context.low_htlb_areas;
 
-                /* First see if we can do the mapping in the existing
-                 * low areas */
+                /* First see if we can use the hint address */
+                if (addr && (htlb_check_hinted_area(addr, len) == 0)) {
+                        areamask = LOW_ESID_MASK(addr, len);
+                        if (open_low_hpage_areas(current->mm, areamask) == 0)
+                                return addr;
+                }
+
+                /* Next see if we can map in the existing low areas */
                 addr = htlb_get_low_area(len, curareas);
                 if (addr != -ENOMEM)
                         return addr;
 
+                /* Finally go looking for areas to open */
                 lastshift = 0;
                 for (areamask = LOW_ESID_MASK(0x100000000UL-len, len);
                      ! lastshift; areamask >>=1) {
@@ -641,12 +665,22 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
         } else {
                 curareas = current->mm->context.high_htlb_areas;
 
-                /* First see if we can do the mapping in the existing
-                 * high areas */
+                /* First see if we can use the hint address */
+                /* We discourage 64-bit processes from doing hugepage
+                 * mappings below 4GB (must use MAP_FIXED) */
+                if ((addr >= 0x100000000UL)
+                    && (htlb_check_hinted_area(addr, len) == 0)) {
+                        areamask = HTLB_AREA_MASK(addr, len);
+                        if (open_high_hpage_areas(current->mm, areamask) == 0)
+                                return addr;
+                }
+
+                /* Next see if we can map in the existing high areas */
                 addr = htlb_get_high_area(len, curareas);
                 if (addr != -ENOMEM)
                         return addr;
 
+                /* Finally go looking for areas to open */
                 lastshift = 0;
                 for (areamask = HTLB_AREA_MASK(TASK_SIZE_USER64-len, len);
                      ! lastshift; areamask >>=1) {
diff --git a/arch/powerpc/mm/imalloc.c b/arch/powerpc/mm/imalloc.c
index f9587bcc6a48..8b0c132bc163 100644
--- a/arch/powerpc/mm/imalloc.c
+++ b/arch/powerpc/mm/imalloc.c
@@ -107,6 +107,7 @@ static int im_region_status(unsigned long v_addr, unsigned long size,
                 if (v_addr < (unsigned long) tmp->addr + tmp->size)
                         break;
 
+        *vm = NULL;
         if (tmp) {
                 if (im_region_overlaps(v_addr, size, tmp))
                         return IM_REGION_OVERLAP;
@@ -127,7 +128,6 @@ static int im_region_status(unsigned long v_addr, unsigned long size,
                 }
         }
 
-        *vm = NULL;
         return IM_REGION_UNUSED;
 }
 
diff --git a/arch/powerpc/mm/init_32.c b/arch/powerpc/mm/init_32.c
index 7d4b8b5f0606..7d0d75c11848 100644
--- a/arch/powerpc/mm/init_32.c
+++ b/arch/powerpc/mm/init_32.c
@@ -188,6 +188,11 @@ void __init MMU_init(void)
 
         if (ppc_md.progress)
                 ppc_md.progress("MMU:exit", 0x211);
+
+        /* From now on, btext is no longer BAT mapped if it was at all */
+#ifdef CONFIG_BOOTX_TEXT
+        btext_unmap();
+#endif
 }
 
 /* This is only called until mem_init is done. */
diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c
index ed6ed2e30dac..15aac0d78dfa 100644
--- a/arch/powerpc/mm/mem.c
+++ b/arch/powerpc/mm/mem.c
@@ -114,19 +114,18 @@ void online_page(struct page *page)
         num_physpages++;
 }
 
-/*
- * This works only for the non-NUMA case.  Later, we'll need a lookup
- * to convert from real physical addresses to nid, that doesn't use
- * pfn_to_nid().
- */
 int __devinit add_memory(u64 start, u64 size)
 {
-        struct pglist_data *pgdata = NODE_DATA(0);
+        struct pglist_data *pgdata;
         struct zone *zone;
+        int nid;
         unsigned long start_pfn = start >> PAGE_SHIFT;
         unsigned long nr_pages = size >> PAGE_SHIFT;
 
-        start += KERNELBASE;
+        nid = hot_add_scn_to_nid(start);
+        pgdata = NODE_DATA(nid);
+
+        start = __va(start);
         create_section_mapping(start, start + size);
 
         /* this should work for most non-highmem platforms */
diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c
index ba7a3055a9fc..2863a912bcd0 100644
--- a/arch/powerpc/mm/numa.c
+++ b/arch/powerpc/mm/numa.c
@@ -37,6 +37,7 @@ EXPORT_SYMBOL(node_data);
 
 static bootmem_data_t __initdata plat_node_bdata[MAX_NUMNODES];
 static int min_common_depth;
+static int n_mem_addr_cells, n_mem_size_cells;
 
 /*
  * We need somewhere to store start/end/node for each region until we have
@@ -254,32 +255,20 @@ static int __init find_min_common_depth(void)
         return depth;
 }
 
-static int __init get_mem_addr_cells(void)
+static void __init get_n_mem_cells(int *n_addr_cells, int *n_size_cells)
 {
         struct device_node *memory = NULL;
-        int rc;
 
         memory = of_find_node_by_type(memory, "memory");
         if (!memory)
-                return 0; /* it won't matter */
+                panic("numa.c: No memory nodes found!");
 
-        rc = prom_n_addr_cells(memory);
-        return rc;
+        *n_addr_cells = prom_n_addr_cells(memory);
+        *n_size_cells = prom_n_size_cells(memory);
+        of_node_put(memory);
 }
 
-static int __init get_mem_size_cells(void)
-{
-        struct device_node *memory = NULL;
-        int rc;
-
-        memory = of_find_node_by_type(memory, "memory");
-        if (!memory)
-                return 0; /* it won't matter */
-        rc = prom_n_size_cells(memory);
-        return rc;
-}
-
-static unsigned long __init read_n_cells(int n, unsigned int **buf)
+static unsigned long __devinit read_n_cells(int n, unsigned int **buf)
 {
         unsigned long result = 0;
 
@@ -386,7 +375,6 @@ static int __init parse_numa_properties(void)
 {
         struct device_node *cpu = NULL;
         struct device_node *memory = NULL;
-        int addr_cells, size_cells;
         int max_domain;
         unsigned long i;
 
@@ -425,8 +413,7 @@ static int __init parse_numa_properties(void)
                 }
         }
 
-        addr_cells = get_mem_addr_cells();
-        size_cells = get_mem_size_cells();
+        get_n_mem_cells(&n_mem_addr_cells, &n_mem_size_cells);
         memory = NULL;
         while ((memory = of_find_node_by_type(memory, "memory")) != NULL) {
                 unsigned long start;
@@ -436,15 +423,21 @@ static int __init parse_numa_properties(void)
                 unsigned int *memcell_buf;
                 unsigned int len;
 
-                memcell_buf = (unsigned int *)get_property(memory, "reg", &len);
+                memcell_buf = (unsigned int *)get_property(memory,
+                        "linux,usable-memory", &len);
+                if (!memcell_buf || len <= 0)
+                        memcell_buf =
+                                (unsigned int *)get_property(memory, "reg",
+                                        &len);
                 if (!memcell_buf || len <= 0)
                         continue;
 
-                ranges = memory->n_addrs;
+                /* ranges in cell */
+                ranges = (len >> 2) / (n_mem_addr_cells + n_mem_size_cells);
 new_range:
                 /* these are order-sensitive, and modify the buffer pointer */
-                start = read_n_cells(addr_cells, &memcell_buf);
-                size = read_n_cells(size_cells, &memcell_buf);
+                start = read_n_cells(n_mem_addr_cells, &memcell_buf);
+                size = read_n_cells(n_mem_size_cells, &memcell_buf);
 
                 numa_domain = of_node_numa_domain(memory);
 
@@ -497,7 +490,41 @@ static void __init setup_nonnuma(void)
         node_set_online(0);
 }
 
-static void __init dump_numa_topology(void)
+void __init dump_numa_cpu_topology(void)
+{
+        unsigned int node;
+        unsigned int cpu, count;
+
+        if (min_common_depth == -1 || !numa_enabled)
+                return;
+
+        for_each_online_node(node) {
+                printk(KERN_INFO "Node %d CPUs:", node);
+
+                count = 0;
+                /*
+                 * If we used a CPU iterator here we would miss printing
+                 * the holes in the cpumap.
+                 */
+                for (cpu = 0; cpu < NR_CPUS; cpu++) {
+                        if (cpu_isset(cpu, numa_cpumask_lookup_table[node])) {
+                                if (count == 0)
+                                        printk(" %u", cpu);
+                                ++count;
+                        } else {
+                                if (count > 1)
+                                        printk("-%u", cpu - 1);
+                                count = 0;
+                        }
+                }
+
+                if (count > 1)
+                        printk("-%u", NR_CPUS - 1);
+                printk("\n");
+        }
+}
+
+static void __init dump_numa_memory_topology(void)
 {
         unsigned int node;
         unsigned int count;
@@ -529,7 +556,6 @@ static void __init dump_numa_topology(void)
                         printk("-0x%lx", i);
                 printk("\n");
         }
-        return;
 }
 
 /*
@@ -591,7 +617,7 @@ void __init do_init_bootmem(void)
         if (parse_numa_properties())
                 setup_nonnuma();
         else
-                dump_numa_topology();
+                dump_numa_memory_topology();
 
         register_cpu_notifier(&ppc64_numa_nb);
 
@@ -730,3 +756,60 @@ static int __init early_numa(char *p)
         return 0;
 }
 early_param("numa", early_numa);
+
+#ifdef CONFIG_MEMORY_HOTPLUG
+/*
+ * Find the node associated with a hot added memory section.  Section
+ * corresponds to a SPARSEMEM section, not an LMB.  It is assumed that
+ * sections are fully contained within a single LMB.
+ */
+int hot_add_scn_to_nid(unsigned long scn_addr)
+{
+        struct device_node *memory = NULL;
+        nodemask_t nodes;
+        int numa_domain = 0;
+
+        if (!numa_enabled || (min_common_depth < 0))
+                return numa_domain;
+
+        while ((memory = of_find_node_by_type(memory, "memory")) != NULL) {
+                unsigned long start, size;
+                int ranges;
+                unsigned int *memcell_buf;
+                unsigned int len;
+
+                memcell_buf = (unsigned int *)get_property(memory, "reg", &len);
+                if (!memcell_buf || len <= 0)
+                        continue;
+
+                /* ranges in cell */
+                ranges = (len >> 2) / (n_mem_addr_cells + n_mem_size_cells);
+ha_new_range:
+                start = read_n_cells(n_mem_addr_cells, &memcell_buf);
+                size = read_n_cells(n_mem_size_cells, &memcell_buf);
+                numa_domain = of_node_numa_domain(memory);
+
+                /* Domains not present at boot default to 0 */
+                if (!node_online(numa_domain))
+                        numa_domain = any_online_node(NODE_MASK_ALL);
+
+                if ((scn_addr >= start) && (scn_addr < (start + size))) {
+                        of_node_put(memory);
+                        goto got_numa_domain;
+                }
+
+                if (--ranges)           /* process all ranges in cell */
+                        goto ha_new_range;
+        }
+        BUG();  /* section address should be found above */
+
+        /* Temporary code to ensure that returned node is not empty */
+got_numa_domain:
+        nodes_setall(nodes);
+        while (NODE_DATA(numa_domain)->node_spanned_pages == 0) {
+                node_clear(numa_domain, nodes);
+                numa_domain = any_online_node(nodes);
+        }
+        return numa_domain;
+}
+#endif /* CONFIG_MEMORY_HOTPLUG */
diff --git a/arch/powerpc/mm/slb.c b/arch/powerpc/mm/slb.c
index 60e852f2f8e5..ffc8ed4de62d 100644
--- a/arch/powerpc/mm/slb.c
+++ b/arch/powerpc/mm/slb.c
@@ -75,7 +75,7 @@ static void slb_flush_and_rebolt(void)
         vflags = SLB_VSID_KERNEL | virtual_llp;
 
         ksp_esid_data = mk_esid_data(get_paca()->kstack, 2);
-        if ((ksp_esid_data & ESID_MASK) == KERNELBASE)
+        if ((ksp_esid_data & ESID_MASK) == PAGE_OFFSET)
                 ksp_esid_data &= ~SLB_ESID_V;
 
         /* We need to do this all in asm, so we're sure we don't touch
@@ -87,8 +87,8 @@ static void slb_flush_and_rebolt(void)
                      /* Slot 2 - kernel stack */
                      "slbmte    %2,%3\n"
                      "isync"
-                     :: "r"(mk_vsid_data(VMALLOCBASE, vflags)),
-                        "r"(mk_esid_data(VMALLOCBASE, 1)),
+                     :: "r"(mk_vsid_data(VMALLOC_START, vflags)),
+                        "r"(mk_esid_data(VMALLOC_START, 1)),
                         "r"(mk_vsid_data(ksp_esid_data, lflags)),
                         "r"(ksp_esid_data)
                      : "memory");
@@ -134,14 +134,14 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
         else
                 unmapped_base = TASK_UNMAPPED_BASE_USER64;
 
-        if (pc >= KERNELBASE)
+        if (is_kernel_addr(pc))
                 return;
         slb_allocate(pc);
 
         if (GET_ESID(pc) == GET_ESID(stack))
                 return;
 
-        if (stack >= KERNELBASE)
+        if (is_kernel_addr(stack))
                 return;
         slb_allocate(stack);
 
@@ -149,7 +149,7 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
             || (GET_ESID(stack) == GET_ESID(unmapped_base)))
                 return;
 
-        if (unmapped_base >= KERNELBASE)
+        if (is_kernel_addr(unmapped_base))
                 return;
         slb_allocate(unmapped_base);
 }
@@ -213,10 +213,10 @@ void slb_initialize(void)
         asm volatile("isync":::"memory");
         asm volatile("slbmte  %0,%0"::"r" (0) : "memory");
         asm volatile("isync; slbia; isync":::"memory");
-        create_slbe(KERNELBASE, lflags, 0);
+        create_slbe(PAGE_OFFSET, lflags, 0);
 
         /* VMALLOC space has 4K pages always for now */
-        create_slbe(VMALLOCBASE, vflags, 1);
+        create_slbe(VMALLOC_START, vflags, 1);
 
         /* We don't bolt the stack for the time being - we're in boot,
          * so the stack is in the bolted segment.  By the time it goes
diff --git a/arch/powerpc/mm/slb_low.S b/arch/powerpc/mm/slb_low.S
index 950ffc5848c7..d1acee38f163 100644
--- a/arch/powerpc/mm/slb_low.S
+++ b/arch/powerpc/mm/slb_low.S
@@ -37,9 +37,9 @@ _GLOBAL(slb_allocate_realmode)
 
         srdi    r9,r3,60                /* get region */
         srdi    r10,r3,28               /* get esid */
-        cmpldi  cr7,r9,0xc              /* cmp KERNELBASE for later use */
+        cmpldi  cr7,r9,0xc              /* cmp PAGE_OFFSET for later use */
 
-        /* r3 = address, r10 = esid, cr7 = <>KERNELBASE */
+        /* r3 = address, r10 = esid, cr7 = <> PAGE_OFFSET */
         blt     cr7,0f                  /* user or kernel? */
 
         /* kernel address: proto-VSID = ESID */
@@ -166,7 +166,7 @@ _GLOBAL(slb_allocate_user)
 /*
  * Finish loading of an SLB entry and return
  *
- * r3 = EA, r10 = proto-VSID, r11 = flags, clobbers r9, cr7 = <>KERNELBASE
+ * r3 = EA, r10 = proto-VSID, r11 = flags, clobbers r9, cr7 = <> PAGE_OFFSET
  */
 slb_finish_load:
         ASM_VSID_SCRAMBLE(r10,r9)
diff --git a/arch/powerpc/mm/stab.c b/arch/powerpc/mm/stab.c
index 51e7951414e5..82e4951826bc 100644
--- a/arch/powerpc/mm/stab.c
+++ b/arch/powerpc/mm/stab.c
@@ -40,7 +40,7 @@ static int make_ste(unsigned long stab, unsigned long esid, unsigned long vsid)
         unsigned long entry, group, old_esid, castout_entry, i;
         unsigned int global_entry;
         struct stab_entry *ste, *castout_ste;
-        unsigned long kernel_segment = (esid << SID_SHIFT) >= KERNELBASE;
+        unsigned long kernel_segment = (esid << SID_SHIFT) >= PAGE_OFFSET;
 
         vsid_data = vsid << STE_VSID_SHIFT;
         esid_data = esid << SID_SHIFT | STE_ESID_KP | STE_ESID_V;
@@ -83,7 +83,7 @@ static int make_ste(unsigned long stab, unsigned long esid, unsigned long vsid)
                 }
 
                 /* Dont cast out the first kernel segment */
-                if ((castout_ste->esid_data & ESID_MASK) != KERNELBASE)
+                if ((castout_ste->esid_data & ESID_MASK) != PAGE_OFFSET)
                         break;
 
                 castout_entry = (castout_entry + 1) & 0xf;
@@ -122,7 +122,7 @@ static int __ste_allocate(unsigned long ea, struct mm_struct *mm)
         unsigned long offset;
 
         /* Kernel or user address? */
-        if (ea >= KERNELBASE) {
+        if (is_kernel_addr(ea)) {
                 vsid = get_kernel_vsid(ea);
         } else {
                 if ((ea >= TASK_SIZE_USER64) || (! mm))
@@ -133,7 +133,7 @@ static int __ste_allocate(unsigned long ea, struct mm_struct *mm)
 
         stab_entry = make_ste(get_paca()->stab_addr, GET_ESID(ea), vsid);
 
-        if (ea < KERNELBASE) {
+        if (!is_kernel_addr(ea)) {
                 offset = __get_cpu_var(stab_cache_ptr);
                 if (offset < NR_STAB_CACHE_ENTRIES)
                         __get_cpu_var(stab_cache[offset++]) = stab_entry;
@@ -190,7 +190,7 @@ void switch_stab(struct task_struct *tsk, struct mm_struct *mm)
                      entry++, ste++) {
                         unsigned long ea;
                         ea = ste->esid_data & ESID_MASK;
-                        if (ea < KERNELBASE) {
+                        if (!is_kernel_addr(ea)) {
                                 ste->esid_data = 0;
                         }
                 }
@@ -251,7 +251,7 @@ void stabs_alloc(void)
                         panic("Unable to allocate segment table for CPU %d.\n",
                               cpu);
 
-                newstab += KERNELBASE;
+                newstab = (unsigned long)__va(newstab);
 
                 memset((void *)newstab, 0, HW_PAGE_SIZE);
 
@@ -270,11 +270,11 @@ void stabs_alloc(void)
  */
 void stab_initialize(unsigned long stab)
 {
-        unsigned long vsid = get_kernel_vsid(KERNELBASE);
+        unsigned long vsid = get_kernel_vsid(PAGE_OFFSET);
         unsigned long stabreal;
 
         asm volatile("isync; slbia; isync":::"memory");
-        make_ste(stab, GET_ESID(KERNELBASE), vsid);
+        make_ste(stab, GET_ESID(PAGE_OFFSET), vsid);
 
         /* Order update */
         asm volatile("sync":::"memory");
diff --git a/arch/powerpc/mm/tlb_64.c b/arch/powerpc/mm/tlb_64.c
index 859d29a0cac5..bb3afb6e6317 100644
--- a/arch/powerpc/mm/tlb_64.c
+++ b/arch/powerpc/mm/tlb_64.c
@@ -168,7 +168,7 @@ void hpte_update(struct mm_struct *mm, unsigned long addr,
                 batch->mm = mm;
                 batch->psize = psize;
         }
-        if (addr < KERNELBASE) {
+        if (!is_kernel_addr(addr)) {
                 vsid = get_vsid(mm->context.id, addr);
                 WARN_ON(vsid == 0);
         } else