1 files changed, 17 insertions, 40 deletions

diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c
index 09e838801e39..7ac6b62fb7c1 100644
--- a/arch/sparc/mm/init_64.c
+++ b/arch/sparc/mm/init_64.c
@@ -324,18 +324,6 @@ static void __update_mmu_tsb_insert(struct mm_struct *mm, unsigned long tsb_inde
 	tsb_insert(tsb, tag, tte);
 }
 
-#if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE)
-static inline bool is_hugetlb_pte(pte_t pte)
-{
-	if ((tlb_type == hypervisor &&
-	     (pte_val(pte) & _PAGE_SZALL_4V) == _PAGE_SZHUGE_4V) ||
-	    (tlb_type != hypervisor &&
-	     (pte_val(pte) & _PAGE_SZALL_4U) == _PAGE_SZHUGE_4U))
-		return true;
-	return false;
-}
-#endif
-
 void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t *ptep)
 {
 	struct mm_struct *mm;
@@ -358,10 +346,13 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address, pte_t *
 	spin_lock_irqsave(&mm->context.lock, flags);
 
 #if defined(CONFIG_HUGETLB_PAGE) || defined(CONFIG_TRANSPARENT_HUGEPAGE)
-	if (mm->context.huge_pte_count && is_hugetlb_pte(pte))
+	if ((mm->context.hugetlb_pte_count || mm->context.thp_pte_count) &&
+	    is_hugetlb_pte(pte)) {
+		/* We are fabricating 8MB pages using 4MB real hw pages.  */
+		pte_val(pte) |= (address & (1UL << REAL_HPAGE_SHIFT));
 		__update_mmu_tsb_insert(mm, MM_TSB_HUGE, REAL_HPAGE_SHIFT,
 					address, pte_val(pte));
-	else
+	} else
 #endif
 		__update_mmu_tsb_insert(mm, MM_TSB_BASE, PAGE_SHIFT,
 					address, pte_val(pte));
@@ -1169,7 +1160,7 @@ int __node_distance(int from, int to)
 	return numa_latency[from][to];
 }
 
-static int find_best_numa_node_for_mlgroup(struct mdesc_mlgroup *grp)
+static int __init find_best_numa_node_for_mlgroup(struct mdesc_mlgroup *grp)
 {
 	int i;
 
@@ -1182,8 +1173,8 @@ static int find_best_numa_node_for_mlgroup(struct mdesc_mlgroup *grp)
 	return i;
 }
 
-static void find_numa_latencies_for_group(struct mdesc_handle *md, u64 grp,
-					  int index)
+static void __init find_numa_latencies_for_group(struct mdesc_handle *md,
+						 u64 grp, int index)
 {
 	u64 arc;
 
@@ -2090,7 +2081,6 @@ void __init paging_init(void)
 {
 	unsigned long end_pfn, shift, phys_base;
 	unsigned long real_end, i;
-	int node;
 
 	setup_page_offset();
 
@@ -2259,21 +2249,6 @@ void __init paging_init(void)
 	/* Setup bootmem... */
 	last_valid_pfn = end_pfn = bootmem_init(phys_base);
 
-	/* Once the OF device tree and MDESC have been setup, we know
-	 * the list of possible cpus.  Therefore we can allocate the
-	 * IRQ stacks.
-	 */
-	for_each_possible_cpu(i) {
-		node = cpu_to_node(i);
-
-		softirq_stack[i] = __alloc_bootmem_node(NODE_DATA(node),
-							THREAD_SIZE,
-							THREAD_SIZE, 0);
-		hardirq_stack[i] = __alloc_bootmem_node(NODE_DATA(node),
-							THREAD_SIZE,
-							THREAD_SIZE, 0);
-	}
-
 	kernel_physical_mapping_init();
 
 	{
@@ -2716,8 +2691,7 @@ void __flush_tlb_all(void)
 pte_t *pte_alloc_one_kernel(struct mm_struct *mm,
 			    unsigned long address)
 {
-	struct page *page = alloc_page(GFP_KERNEL | __GFP_NOTRACK |
-				       __GFP_REPEAT | __GFP_ZERO);
+	struct page *page = alloc_page(GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO);
 	pte_t *pte = NULL;
 
 	if (page)
@@ -2729,8 +2703,7 @@ pte_t *pte_alloc_one_kernel(struct mm_struct *mm,
 pgtable_t pte_alloc_one(struct mm_struct *mm,
 			unsigned long address)
 {
-	struct page *page = alloc_page(GFP_KERNEL | __GFP_NOTRACK |
-				       __GFP_REPEAT | __GFP_ZERO);
+	struct page *page = alloc_page(GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO);
 	if (!page)
 		return NULL;
 	if (!pgtable_page_ctor(page)) {
@@ -2836,9 +2809,10 @@ void hugetlb_setup(struct pt_regs *regs)
 	 * the Data-TLB for huge pages.
 	 */
 	if (tlb_type == cheetah_plus) {
+		bool need_context_reload = false;
 		unsigned long ctx;
 
-		spin_lock(&ctx_alloc_lock);
+		spin_lock_irq(&ctx_alloc_lock);
 		ctx = mm->context.sparc64_ctx_val;
 		ctx &= ~CTX_PGSZ_MASK;
 		ctx |= CTX_PGSZ_BASE << CTX_PGSZ0_SHIFT;
@@ -2857,9 +2831,12 @@ void hugetlb_setup(struct pt_regs *regs)
 			 * also executing in this address space.
 			 */
 			mm->context.sparc64_ctx_val = ctx;
-			on_each_cpu(context_reload, mm, 0);
+			need_context_reload = true;
 		}
-		spin_unlock(&ctx_alloc_lock);
+		spin_unlock_irq(&ctx_alloc_lock);
+
+		if (need_context_reload)
+			on_each_cpu(context_reload, mm, 0);
 	}
 }
 #endif