Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6

5 files changed, 166 insertions, 62 deletions

diff --git a/mm/filemap.c b/mm/filemap.c
index 96ac6b0eb6cb..698ea80f2102 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -1634,14 +1634,15 @@ EXPORT_SYMBOL(generic_file_readonly_mmap);
 static struct page *__read_cache_page(struct address_space *mapping,
                                 pgoff_t index,
                                 int (*filler)(void *,struct page*),
-                                void *data)
+                                void *data,
+                                gfp_t gfp)
 {
         struct page *page;
         int err;
 repeat:
         page = find_get_page(mapping, index);
         if (!page) {
-                page = page_cache_alloc_cold(mapping);
+                page = __page_cache_alloc(gfp | __GFP_COLD);
                 if (!page)
                         return ERR_PTR(-ENOMEM);
                 err = add_to_page_cache_lru(page, mapping, index, GFP_KERNEL);
@@ -1661,31 +1662,18 @@ repeat:
         return page;
 }
 
-/**
- * read_cache_page_async - read into page cache, fill it if needed
- * @mapping:    the page's address_space
- * @index:      the page index
- * @filler:     function to perform the read
- * @data:       destination for read data
- *
- * Same as read_cache_page, but don't wait for page to become unlocked
- * after submitting it to the filler.
- *
- * Read into the page cache. If a page already exists, and PageUptodate() is
- * not set, try to fill the page but don't wait for it to become unlocked.
- *
- * If the page does not get brought uptodate, return -EIO.
- */
-struct page *read_cache_page_async(struct address_space *mapping,
+static struct page *do_read_cache_page(struct address_space *mapping,
                                 pgoff_t index,
                                 int (*filler)(void *,struct page*),
-                                void *data)
+                                void *data,
+                                gfp_t gfp)
+
 {
         struct page *page;
         int err;
 
 retry:
-        page = __read_cache_page(mapping, index, filler, data);
+        page = __read_cache_page(mapping, index, filler, data, gfp);
         if (IS_ERR(page))
                 return page;
         if (PageUptodate(page))
@@ -1710,8 +1698,67 @@ out:
         mark_page_accessed(page);
         return page;
 }
+
+/**
+ * read_cache_page_async - read into page cache, fill it if needed
+ * @mapping:    the page's address_space
+ * @index:      the page index
+ * @filler:     function to perform the read
+ * @data:       destination for read data
+ *
+ * Same as read_cache_page, but don't wait for page to become unlocked
+ * after submitting it to the filler.
+ *
+ * Read into the page cache. If a page already exists, and PageUptodate() is
+ * not set, try to fill the page but don't wait for it to become unlocked.
+ *
+ * If the page does not get brought uptodate, return -EIO.
+ */
+struct page *read_cache_page_async(struct address_space *mapping,
+                                pgoff_t index,
+                                int (*filler)(void *,struct page*),
+                                void *data)
+{
+        return do_read_cache_page(mapping, index, filler, data, mapping_gfp_mask(mapping));
+}
 EXPORT_SYMBOL(read_cache_page_async);
 
+static struct page *wait_on_page_read(struct page *page)
+{
+        if (!IS_ERR(page)) {
+                wait_on_page_locked(page);
+                if (!PageUptodate(page)) {
+                        page_cache_release(page);
+                        page = ERR_PTR(-EIO);
+                }
+        }
+        return page;
+}
+
+/**
+ * read_cache_page_gfp - read into page cache, using specified page allocation flags.
+ * @mapping:    the page's address_space
+ * @index:      the page index
+ * @gfp:        the page allocator flags to use if allocating
+ *
+ * This is the same as "read_mapping_page(mapping, index, NULL)", but with
+ * any new page allocations done using the specified allocation flags. Note
+ * that the Radix tree operations will still use GFP_KERNEL, so you can't
+ * expect to do this atomically or anything like that - but you can pass in
+ * other page requirements.
+ *
+ * If the page does not get brought uptodate, return -EIO.
+ */
+struct page *read_cache_page_gfp(struct address_space *mapping,
+                                pgoff_t index,
+                                gfp_t gfp)
+{
+        filler_t *filler = (filler_t *)mapping->a_ops->readpage;
+
+        return wait_on_page_read(do_read_cache_page(mapping, index, filler, NULL, gfp));
+}
+EXPORT_SYMBOL(read_cache_page_gfp);
+
 /**
  * read_cache_page - read into page cache, fill it if needed
  * @mapping:    the page's address_space
@@ -1729,18 +1776,7 @@ struct page *read_cache_page(struct address_space *mapping,
                                 int (*filler)(void *,struct page*),
                                 void *data)
 {
-        struct page *page;
-
-        page = read_cache_page_async(mapping, index, filler, data);
-        if (IS_ERR(page))
-                goto out;
-        wait_on_page_locked(page);
-        if (!PageUptodate(page)) {
-                page_cache_release(page);
-                page = ERR_PTR(-EIO);
-        }
- out:
-        return page;
+        return wait_on_page_read(read_cache_page_async(mapping, index, filler, data));
 }
 EXPORT_SYMBOL(read_cache_page);
 
@@ -2196,6 +2232,9 @@ again:
                 if (unlikely(status))
                         break;
 
+                if (mapping_writably_mapped(mapping))
+                        flush_dcache_page(page);
+
                 pagefault_disable();
                 copied = iov_iter_copy_from_user_atomic(page, i, offset, bytes);
                 pagefault_enable();
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index e91b81b63670..2d16fa6b8c2d 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1515,10 +1515,9 @@ static struct attribute_group hstate_attr_group = {
         .attrs = hstate_attrs,
 };
 
-static int __init hugetlb_sysfs_add_hstate(struct hstate *h,
-                                struct kobject *parent,
-                                struct kobject **hstate_kobjs,
-                                struct attribute_group *hstate_attr_group)
+static int hugetlb_sysfs_add_hstate(struct hstate *h, struct kobject *parent,
+                                    struct kobject **hstate_kobjs,
+                                    struct attribute_group *hstate_attr_group)
 {
         int retval;
         int hi = h - hstates;
diff --git a/mm/migrate.c b/mm/migrate.c
index efddbf0926b2..9a0db5bbabe4 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -912,6 +912,9 @@ static int do_pages_move(struct mm_struct *mm, struct task_struct *task,
                                 goto out_pm;
 
                         err = -ENODEV;
+                        if (node < 0 || node >= MAX_NUMNODES)
+                                goto out_pm;
+
                         if (!node_state(node, N_HIGH_MEMORY))
                                 goto out_pm;
 
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index d2a8889b4c58..8deb9d0fd5b1 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -556,8 +556,9 @@ static void free_pcppages_bulk(struct zone *zone, int count,
                         page = list_entry(list->prev, struct page, lru);
                         /* must delete as __free_one_page list manipulates */
                         list_del(&page->lru);
-                        __free_one_page(page, zone, 0, migratetype);
-                        trace_mm_page_pcpu_drain(page, 0, migratetype);
+                        /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
+                        __free_one_page(page, zone, 0, page_private(page));
+                        trace_mm_page_pcpu_drain(page, 0, page_private(page));
                 } while (--count && --batch_free && !list_empty(list));
         }
         spin_unlock(&zone->lock);
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index d55d905463eb..ae007462b7f6 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -509,6 +509,9 @@ static unsigned long lazy_max_pages(void)
 
 static atomic_t vmap_lazy_nr = ATOMIC_INIT(0);
 
+/* for per-CPU blocks */
+static void purge_fragmented_blocks_allcpus(void);
+
 /*
  * Purges all lazily-freed vmap areas.
  *
@@ -539,6 +542,9 @@ static void __purge_vmap_area_lazy(unsigned long *start, unsigned long *end,
         } else
                 spin_lock(&purge_lock);
 
+        if (sync)
+                purge_fragmented_blocks_allcpus();
+
         rcu_read_lock();
         list_for_each_entry_rcu(va, &vmap_area_list, list) {
                 if (va->flags & VM_LAZY_FREE) {
@@ -667,8 +673,6 @@ static bool vmap_initialized __read_mostly = false;
 struct vmap_block_queue {
         spinlock_t lock;
         struct list_head free;
-        struct list_head dirty;
-        unsigned int nr_dirty;
 };
 
 struct vmap_block {
@@ -678,10 +682,9 @@ struct vmap_block {
         unsigned long free, dirty;
         DECLARE_BITMAP(alloc_map, VMAP_BBMAP_BITS);
         DECLARE_BITMAP(dirty_map, VMAP_BBMAP_BITS);
-        union {
-                struct list_head free_list;
-                struct rcu_head rcu_head;
-        };
+        struct list_head free_list;
+        struct rcu_head rcu_head;
+        struct list_head purge;
 };
 
 /* Queue of free and dirty vmap blocks, for allocation and flushing purposes */
@@ -757,7 +760,7 @@ static struct vmap_block *new_vmap_block(gfp_t gfp_mask)
         vbq = &get_cpu_var(vmap_block_queue);
         vb->vbq = vbq;
         spin_lock(&vbq->lock);
-        list_add(&vb->free_list, &vbq->free);
+        list_add_rcu(&vb->free_list, &vbq->free);
         spin_unlock(&vbq->lock);
         put_cpu_var(vmap_block_queue);
 
@@ -776,8 +779,6 @@ static void free_vmap_block(struct vmap_block *vb)
         struct vmap_block *tmp;
         unsigned long vb_idx;
 
-        BUG_ON(!list_empty(&vb->free_list));
-
         vb_idx = addr_to_vb_idx(vb->va->va_start);
         spin_lock(&vmap_block_tree_lock);
         tmp = radix_tree_delete(&vmap_block_tree, vb_idx);
@@ -788,12 +789,61 @@ static void free_vmap_block(struct vmap_block *vb)
         call_rcu(&vb->rcu_head, rcu_free_vb);
 }
 
+static void purge_fragmented_blocks(int cpu)
+{
+        LIST_HEAD(purge);
+        struct vmap_block *vb;
+        struct vmap_block *n_vb;
+        struct vmap_block_queue *vbq = &per_cpu(vmap_block_queue, cpu);
+
+        rcu_read_lock();
+        list_for_each_entry_rcu(vb, &vbq->free, free_list) {
+
+                if (!(vb->free + vb->dirty == VMAP_BBMAP_BITS && vb->dirty != VMAP_BBMAP_BITS))
+                        continue;
+
+                spin_lock(&vb->lock);
+                if (vb->free + vb->dirty == VMAP_BBMAP_BITS && vb->dirty != VMAP_BBMAP_BITS) {
+                        vb->free = 0; /* prevent further allocs after releasing lock */
+                        vb->dirty = VMAP_BBMAP_BITS; /* prevent purging it again */
+                        bitmap_fill(vb->alloc_map, VMAP_BBMAP_BITS);
+                        bitmap_fill(vb->dirty_map, VMAP_BBMAP_BITS);
+                        spin_lock(&vbq->lock);
+                        list_del_rcu(&vb->free_list);
+                        spin_unlock(&vbq->lock);
+                        spin_unlock(&vb->lock);
+                        list_add_tail(&vb->purge, &purge);
+                } else
+                        spin_unlock(&vb->lock);
+        }
+        rcu_read_unlock();
+
+        list_for_each_entry_safe(vb, n_vb, &purge, purge) {
+                list_del(&vb->purge);
+                free_vmap_block(vb);
+        }
+}
+
+static void purge_fragmented_blocks_thiscpu(void)
+{
+        purge_fragmented_blocks(smp_processor_id());
+}
+
+static void purge_fragmented_blocks_allcpus(void)
+{
+        int cpu;
+
+        for_each_possible_cpu(cpu)
+                purge_fragmented_blocks(cpu);
+}
+
 static void *vb_alloc(unsigned long size, gfp_t gfp_mask)
 {
         struct vmap_block_queue *vbq;
         struct vmap_block *vb;
         unsigned long addr = 0;
         unsigned int order;
+        int purge = 0;
 
         BUG_ON(size & ~PAGE_MASK);
         BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC);
@@ -806,24 +856,38 @@ again:
                 int i;
 
                 spin_lock(&vb->lock);
+                if (vb->free < 1UL << order)
+                        goto next;
+
                 i = bitmap_find_free_region(vb->alloc_map,
                                                 VMAP_BBMAP_BITS, order);
 
-                if (i >= 0) {
-                        addr = vb->va->va_start + (i << PAGE_SHIFT);
-                        BUG_ON(addr_to_vb_idx(addr) !=
-                                        addr_to_vb_idx(vb->va->va_start));
-                        vb->free -= 1UL << order;
-                        if (vb->free == 0) {
-                                spin_lock(&vbq->lock);
-                                list_del_init(&vb->free_list);
-                                spin_unlock(&vbq->lock);
+                if (i < 0) {
+                        if (vb->free + vb->dirty == VMAP_BBMAP_BITS) {
+                                /* fragmented and no outstanding allocations */
+                                BUG_ON(vb->dirty != VMAP_BBMAP_BITS);
+                                purge = 1;
                         }
-                        spin_unlock(&vb->lock);
-                        break;
+                        goto next;
                 }
+                addr = vb->va->va_start + (i << PAGE_SHIFT);
+                BUG_ON(addr_to_vb_idx(addr) !=
+                                addr_to_vb_idx(vb->va->va_start));
+                vb->free -= 1UL << order;
+                if (vb->free == 0) {
+                        spin_lock(&vbq->lock);
+                        list_del_rcu(&vb->free_list);
+                        spin_unlock(&vbq->lock);
+                }
+                spin_unlock(&vb->lock);
+                break;
+next:
                 spin_unlock(&vb->lock);
         }
+
+        if (purge)
+                purge_fragmented_blocks_thiscpu();
+
         put_cpu_var(vmap_block_queue);
         rcu_read_unlock();
 
@@ -860,11 +924,11 @@ static void vb_free(const void *addr, unsigned long size)
         BUG_ON(!vb);
 
         spin_lock(&vb->lock);
-        bitmap_allocate_region(vb->dirty_map, offset >> PAGE_SHIFT, order);
+        BUG_ON(bitmap_allocate_region(vb->dirty_map, offset >> PAGE_SHIFT, order));
 
         vb->dirty += 1UL << order;
         if (vb->dirty == VMAP_BBMAP_BITS) {
-                BUG_ON(vb->free || !list_empty(&vb->free_list));
+                BUG_ON(vb->free);
                 spin_unlock(&vb->lock);
                 free_vmap_block(vb);
         } else
@@ -1033,8 +1097,6 @@ void __init vmalloc_init(void)
                 vbq = &per_cpu(vmap_block_queue, i);
                 spin_lock_init(&vbq->lock);
                 INIT_LIST_HEAD(&vbq->free);
-                INIT_LIST_HEAD(&vbq->dirty);
-                vbq->nr_dirty = 0;
         }
 
         /* Import existing vmlist entries. */