unsigned int height; /* Height from the bottom */
unsigned int count;
struct rcu_head rcu_head;
- void *slots[RADIX_TREE_MAP_SIZE];
+ void __rcu *slots[RADIX_TREE_MAP_SIZE];
unsigned long tags[RADIX_TREE_MAX_TAGS][RADIX_TREE_TAG_LONGS];
};
};
static DEFINE_PER_CPU(struct radix_tree_preload, radix_tree_preloads) = { 0, };
+static inline void *ptr_to_indirect(void *ptr)
+{
+ return (void *)((unsigned long)ptr | RADIX_TREE_INDIRECT_PTR);
+}
+
+static inline void *indirect_to_ptr(void *ptr)
+{
+ return (void *)((unsigned long)ptr & ~RADIX_TREE_INDIRECT_PTR);
+}
+
static inline gfp_t root_gfp_mask(struct radix_tree_root *root)
{
return root->gfp_mask & __GFP_BITS_MASK;
{
struct radix_tree_node *node =
container_of(head, struct radix_tree_node, rcu_head);
+ int i;
/*
* must only free zeroed nodes into the slab. radix_tree_shrink
* can leave us with a non-NULL entry in the first slot, so clear
* that here to make sure.
*/
- tag_clear(node, 0, 0);
- tag_clear(node, 1, 0);
+ for (i = 0; i < RADIX_TREE_MAX_TAGS; i++)
+ tag_clear(node, i, 0);
+
node->slots[0] = NULL;
node->count = 0;
return -ENOMEM;
/* Increase the height. */
- node->slots[0] = radix_tree_indirect_to_ptr(root->rnode);
+ node->slots[0] = indirect_to_ptr(root->rnode);
/* Propagate the aggregated tag info into the new root */
for (tag = 0; tag < RADIX_TREE_MAX_TAGS; tag++) {
newheight = root->height+1;
node->height = newheight;
node->count = 1;
- node = radix_tree_ptr_to_indirect(node);
+ node = ptr_to_indirect(node);
rcu_assign_pointer(root->rnode, node);
root->height = newheight;
} while (height > root->height);
return error;
}
- slot = radix_tree_indirect_to_ptr(root->rnode);
+ slot = indirect_to_ptr(root->rnode);
height = root->height;
shift = (height-1) * RADIX_TREE_MAP_SHIFT;
rcu_assign_pointer(node->slots[offset], slot);
node->count++;
} else
- rcu_assign_pointer(root->rnode,
- radix_tree_ptr_to_indirect(slot));
+ rcu_assign_pointer(root->rnode, ptr_to_indirect(slot));
}
/* Go a level down */
return NULL;
return is_slot ? (void *)&root->rnode : node;
}
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
height = node->height;
if (index > radix_tree_maxindex(height))
height--;
} while (height > 0);
- return is_slot ? (void *)slot:node;
+ return is_slot ? (void *)slot : indirect_to_ptr(node);
}
/**
height = root->height;
BUG_ON(index > radix_tree_maxindex(height));
- slot = radix_tree_indirect_to_ptr(root->rnode);
+ slot = indirect_to_ptr(root->rnode);
shift = (height - 1) * RADIX_TREE_MAP_SHIFT;
while (height > 0) {
shift = (height - 1) * RADIX_TREE_MAP_SHIFT;
pathp->node = NULL;
- slot = radix_tree_indirect_to_ptr(root->rnode);
+ slot = indirect_to_ptr(root->rnode);
while (height > 0) {
int offset;
{
unsigned int height, shift;
struct radix_tree_node *node;
- int saw_unset_tag = 0;
/* check the root's tag bit */
if (!root_tag_get(root, tag))
if (!radix_tree_is_indirect_ptr(node))
return (index == 0);
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
height = node->height;
if (index > radix_tree_maxindex(height))
return 0;
offset = (index >> shift) & RADIX_TREE_MAP_MASK;
-
- /*
- * This is just a debug check. Later, we can bale as soon as
- * we see an unset tag.
- */
if (!tag_get(node, tag, offset))
- saw_unset_tag = 1;
+ return 0;
if (height == 1)
- return !!tag_get(node, tag, offset);
+ return 1;
node = rcu_dereference_raw(node->slots[offset]);
shift -= RADIX_TREE_MAP_SHIFT;
height--;
* also settag. The function stops either after tagging nr_to_tag items or
* after reaching last_index.
*
+ * The tags must be set from the leaf level only and propagated back up the
+ * path to the root. We must do this so that we resolve the full path before
+ * setting any tags on intermediate nodes. If we set tags as we descend, then
+ * we can get to the leaf node and find that the index that has the iftag
+ * set is outside the range we are scanning. This reults in dangling tags and
+ * can lead to problems with later tag operations (e.g. livelocks on lookups).
+ *
* The function returns number of leaves where the tag was set and sets
* *first_indexp to the first unscanned index.
* WARNING! *first_indexp can wrap if last_index is ULONG_MAX. Caller must
unsigned long nr_to_tag,
unsigned int iftag, unsigned int settag)
{
- unsigned int height = root->height, shift;
- unsigned long tagged = 0, index = *first_indexp;
- struct radix_tree_node *open_slots[height], *slot;
+ unsigned int height = root->height;
+ struct radix_tree_path path[height];
+ struct radix_tree_path *pathp = path;
+ struct radix_tree_node *slot;
+ unsigned int shift;
+ unsigned long tagged = 0;
+ unsigned long index = *first_indexp;
last_index = min(last_index, radix_tree_maxindex(height));
if (index > last_index)
}
shift = (height - 1) * RADIX_TREE_MAP_SHIFT;
- slot = radix_tree_indirect_to_ptr(root->rnode);
+ slot = indirect_to_ptr(root->rnode);
+
+ /*
+ * we fill the path from (root->height - 2) to 0, leaving the index at
+ * (root->height - 1) as a terminator. Zero the node in the terminator
+ * so that we can use this to end walk loops back up the path.
+ */
+ path[height - 1].node = NULL;
for (;;) {
int offset;
goto next;
if (!tag_get(slot, iftag, offset))
goto next;
+ if (height > 1) {
+ /* Go down one level */
+ height--;
+ shift -= RADIX_TREE_MAP_SHIFT;
+ path[height - 1].node = slot;
+ path[height - 1].offset = offset;
+ slot = slot->slots[offset];
+ continue;
+ }
+
+ /* tag the leaf */
+ tagged++;
tag_set(slot, settag, offset);
- if (height == 1) {
- tagged++;
- goto next;
+
+ /* walk back up the path tagging interior nodes */
+ pathp = &path[0];
+ while (pathp->node) {
+ /* stop if we find a node with the tag already set */
+ if (tag_get(pathp->node, settag, pathp->offset))
+ break;
+ tag_set(pathp->node, settag, pathp->offset);
+ pathp++;
}
- /* Go down one level */
- height--;
- shift -= RADIX_TREE_MAP_SHIFT;
- open_slots[height] = slot;
- slot = slot->slots[offset];
- continue;
+
next:
/* Go to next item at level determined by 'shift' */
index = ((index >> shift) + 1) << shift;
* last_index is guaranteed to be in the tree, what
* we do below cannot wander astray.
*/
- slot = open_slots[height];
+ slot = path[height - 1].node;
height++;
shift += RADIX_TREE_MAP_SHIFT;
}
}
/*
- * The iftag must have been set somewhere because otherwise
- * we would return immediated at the beginning of the function
+ * We need not to tag the root tag if there is no tag which is set with
+ * settag within the range from *first_indexp to last_index.
*/
- root_tag_set(root, settag);
+ if (tagged > 0)
+ root_tag_set(root, settag);
*first_indexp = index;
return tagged;
EXPORT_SYMBOL(radix_tree_prev_hole);
static unsigned int
-__lookup(struct radix_tree_node *slot, void ***results, unsigned long index,
- unsigned int max_items, unsigned long *next_index)
+__lookup(struct radix_tree_node *slot, void ***results, unsigned long *indices,
+ unsigned long index, unsigned int max_items, unsigned long *next_index)
{
unsigned int nr_found = 0;
unsigned int shift, height;
/* Bottom level: grab some items */
for (i = index & RADIX_TREE_MAP_MASK; i < RADIX_TREE_MAP_SIZE; i++) {
- index++;
if (slot->slots[i]) {
- results[nr_found++] = &(slot->slots[i]);
- if (nr_found == max_items)
+ results[nr_found] = &(slot->slots[i]);
+ if (indices)
+ indices[nr_found] = index;
+ if (++nr_found == max_items) {
+ index++;
goto out;
+ }
}
+ index++;
}
out:
*next_index = index;
results[0] = node;
return 1;
}
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
max_index = radix_tree_maxindex(node->height);
if (cur_index > max_index)
break;
- slots_found = __lookup(node, (void ***)results + ret, cur_index,
- max_items - ret, &next_index);
+ slots_found = __lookup(node, (void ***)results + ret, NULL,
+ cur_index, max_items - ret, &next_index);
nr_found = 0;
for (i = 0; i < slots_found; i++) {
struct radix_tree_node *slot;
slot = *(((void ***)results)[ret + i]);
if (!slot)
continue;
- results[ret + nr_found] = rcu_dereference_raw(slot);
+ results[ret + nr_found] =
+ indirect_to_ptr(rcu_dereference_raw(slot));
nr_found++;
}
ret += nr_found;
* radix_tree_gang_lookup_slot - perform multiple slot lookup on radix tree
* @root: radix tree root
* @results: where the results of the lookup are placed
+ * @indices: where their indices should be placed (but usually NULL)
* @first_index: start the lookup from this key
* @max_items: place up to this many items at *results
*
* protection, radix_tree_deref_slot may fail requiring a retry.
*/
unsigned int
-radix_tree_gang_lookup_slot(struct radix_tree_root *root, void ***results,
+radix_tree_gang_lookup_slot(struct radix_tree_root *root,
+ void ***results, unsigned long *indices,
unsigned long first_index, unsigned int max_items)
{
unsigned long max_index;
if (first_index > 0)
return 0;
results[0] = (void **)&root->rnode;
+ if (indices)
+ indices[0] = 0;
return 1;
}
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
max_index = radix_tree_maxindex(node->height);
if (cur_index > max_index)
break;
- slots_found = __lookup(node, results + ret, cur_index,
- max_items - ret, &next_index);
+ slots_found = __lookup(node, results + ret,
+ indices ? indices + ret : NULL,
+ cur_index, max_items - ret, &next_index);
ret += slots_found;
if (next_index == 0)
break;
results[0] = node;
return 1;
}
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
max_index = radix_tree_maxindex(node->height);
slot = *(((void ***)results)[ret + i]);
if (!slot)
continue;
- results[ret + nr_found] = rcu_dereference_raw(slot);
+ results[ret + nr_found] =
+ indirect_to_ptr(rcu_dereference_raw(slot));
nr_found++;
}
ret += nr_found;
results[0] = (void **)&root->rnode;
return 1;
}
- node = radix_tree_indirect_to_ptr(node);
+ node = indirect_to_ptr(node);
max_index = radix_tree_maxindex(node->height);
}
EXPORT_SYMBOL(radix_tree_gang_lookup_tag_slot);
+#if defined(CONFIG_SHMEM) && defined(CONFIG_SWAP)
+#include <linux/sched.h> /* for cond_resched() */
+
+/*
+ * This linear search is at present only useful to shmem_unuse_inode().
+ */
+static unsigned long __locate(struct radix_tree_node *slot, void *item,
+ unsigned long index, unsigned long *found_index)
+{
+ unsigned int shift, height;
+ unsigned long i;
+
+ height = slot->height;
+ shift = (height-1) * RADIX_TREE_MAP_SHIFT;
+
+ for ( ; height > 1; height--) {
+ i = (index >> shift) & RADIX_TREE_MAP_MASK;
+ for (;;) {
+ if (slot->slots[i] != NULL)
+ break;
+ index &= ~((1UL << shift) - 1);
+ index += 1UL << shift;
+ if (index == 0)
+ goto out; /* 32-bit wraparound */
+ i++;
+ if (i == RADIX_TREE_MAP_SIZE)
+ goto out;
+ }
+
+ shift -= RADIX_TREE_MAP_SHIFT;
+ slot = rcu_dereference_raw(slot->slots[i]);
+ if (slot == NULL)
+ goto out;
+ }
+
+ /* Bottom level: check items */
+ for (i = 0; i < RADIX_TREE_MAP_SIZE; i++) {
+ if (slot->slots[i] == item) {
+ *found_index = index + i;
+ index = 0;
+ goto out;
+ }
+ }
+ index += RADIX_TREE_MAP_SIZE;
+out:
+ return index;
+}
+
+/**
+ * radix_tree_locate_item - search through radix tree for item
+ * @root: radix tree root
+ * @item: item to be found
+ *
+ * Returns index where item was found, or -1 if not found.
+ * Caller must hold no lock (since this time-consuming function needs
+ * to be preemptible), and must check afterwards if item is still there.
+ */
+unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item)
+{
+ struct radix_tree_node *node;
+ unsigned long max_index;
+ unsigned long cur_index = 0;
+ unsigned long found_index = -1;
+
+ do {
+ rcu_read_lock();
+ node = rcu_dereference_raw(root->rnode);
+ if (!radix_tree_is_indirect_ptr(node)) {
+ rcu_read_unlock();
+ if (node == item)
+ found_index = 0;
+ break;
+ }
+
+ node = indirect_to_ptr(node);
+ max_index = radix_tree_maxindex(node->height);
+ if (cur_index > max_index)
+ break;
+
+ cur_index = __locate(node, item, cur_index, &found_index);
+ rcu_read_unlock();
+ cond_resched();
+ } while (cur_index != 0 && cur_index <= max_index);
+
+ return found_index;
+}
+#else
+unsigned long radix_tree_locate_item(struct radix_tree_root *root, void *item)
+{
+ return -1;
+}
+#endif /* CONFIG_SHMEM && CONFIG_SWAP */
/**
* radix_tree_shrink - shrink height of a radix tree to minimal
void *newptr;
BUG_ON(!radix_tree_is_indirect_ptr(to_free));
- to_free = radix_tree_indirect_to_ptr(to_free);
+ to_free = indirect_to_ptr(to_free);
/*
* The candidate node has more than one child, or its child
/*
* We don't need rcu_assign_pointer(), since we are simply
- * moving the node from one part of the tree to another. If
- * it was safe to dereference the old pointer to it
+ * moving the node from one part of the tree to another: if it
+ * was safe to dereference the old pointer to it
* (to_free->slots[0]), it will be safe to dereference the new
- * one (root->rnode).
+ * one (root->rnode) as far as dependent read barriers go.
*/
newptr = to_free->slots[0];
if (root->height > 1)
- newptr = radix_tree_ptr_to_indirect(newptr);
+ newptr = ptr_to_indirect(newptr);
root->rnode = newptr;
root->height--;
+
+ /*
+ * We have a dilemma here. The node's slot[0] must not be
+ * NULLed in case there are concurrent lookups expecting to
+ * find the item. However if this was a bottom-level node,
+ * then it may be subject to the slot pointer being visible
+ * to callers dereferencing it. If item corresponding to
+ * slot[0] is subsequently deleted, these callers would expect
+ * their slot to become empty sooner or later.
+ *
+ * For example, lockless pagecache will look up a slot, deref
+ * the page pointer, and if the page is 0 refcount it means it
+ * was concurrently deleted from pagecache so try the deref
+ * again. Fortunately there is already a requirement for logic
+ * to retry the entire slot lookup -- the indirect pointer
+ * problem (replacing direct root node with an indirect pointer
+ * also results in a stale slot). So tag the slot as indirect
+ * to force callers to retry.
+ */
+ if (root->height == 0)
+ *((unsigned long *)&to_free->slots[0]) |=
+ RADIX_TREE_INDIRECT_PTR;
+
radix_tree_node_free(to_free);
}
}
root->rnode = NULL;
goto out;
}
- slot = radix_tree_indirect_to_ptr(slot);
+ slot = indirect_to_ptr(slot);
shift = (height - 1) * RADIX_TREE_MAP_SHIFT;
pathp->node = NULL;
radix_tree_node_free(to_free);
if (pathp->node->count) {
- if (pathp->node ==
- radix_tree_indirect_to_ptr(root->rnode))
+ if (pathp->node == indirect_to_ptr(root->rnode))
radix_tree_shrink(root);
goto out;
}