#include <linux/gfp.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
+#include <linux/writeback.h>
+#include <linux/pagevec.h>
#include "ctree.h"
#include "transaction.h"
#include "btrfs_inode.h"
#include "extent_io.h"
-
static u64 entry_end(struct btrfs_ordered_extent *entry)
{
if (entry->file_offset + entry->len < entry->file_offset)
return entry->file_offset + entry->len;
}
+/* returns NULL if the insertion worked, or it returns the node it did find
+ * in the tree
+ */
static struct rb_node *tree_insert(struct rb_root *root, u64 file_offset,
struct rb_node *node)
{
- struct rb_node ** p = &root->rb_node;
- struct rb_node * parent = NULL;
+ struct rb_node **p = &root->rb_node;
+ struct rb_node *parent = NULL;
struct btrfs_ordered_extent *entry;
- while(*p) {
+ while (*p) {
parent = *p;
entry = rb_entry(parent, struct btrfs_ordered_extent, rb_node);
return NULL;
}
+/*
+ * look for a given offset in the tree, and if it can't be found return the
+ * first lesser offset
+ */
static struct rb_node *__tree_search(struct rb_root *root, u64 file_offset,
struct rb_node **prev_ret)
{
- struct rb_node * n = root->rb_node;
+ struct rb_node *n = root->rb_node;
struct rb_node *prev = NULL;
struct rb_node *test;
struct btrfs_ordered_extent *entry;
struct btrfs_ordered_extent *prev_entry = NULL;
- while(n) {
+ while (n) {
entry = rb_entry(n, struct btrfs_ordered_extent, rb_node);
prev = n;
prev_entry = entry;
if (!prev_ret)
return NULL;
- while(prev && file_offset >= entry_end(prev_entry)) {
+ while (prev && file_offset >= entry_end(prev_entry)) {
test = rb_next(prev);
if (!test)
break;
if (prev)
prev_entry = rb_entry(prev, struct btrfs_ordered_extent,
rb_node);
- while(prev && file_offset < entry_end(prev_entry)) {
+ while (prev && file_offset < entry_end(prev_entry)) {
test = rb_prev(prev);
if (!test)
break;
return NULL;
}
+/*
+ * helper to check if a given offset is inside a given entry
+ */
static int offset_in_entry(struct btrfs_ordered_extent *entry, u64 file_offset)
{
if (file_offset < entry->file_offset ||
return 1;
}
+/*
+ * look find the first ordered struct that has this offset, otherwise
+ * the first one less than this offset
+ */
static inline struct rb_node *tree_search(struct btrfs_ordered_inode_tree *tree,
u64 file_offset)
{
* inserted.
*/
int btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
- u64 start, u64 len)
+ u64 start, u64 len, u64 disk_len, int type)
{
struct btrfs_ordered_inode_tree *tree;
struct rb_node *node;
entry->file_offset = file_offset;
entry->start = start;
entry->len = len;
+ entry->disk_len = disk_len;
+ entry->inode = inode;
+ if (type != BTRFS_ORDERED_IO_DONE && type != BTRFS_ORDERED_COMPLETE)
+ set_bit(type, &entry->flags);
+
/* one ref for the tree */
atomic_set(&entry->refs, 1);
init_waitqueue_head(&entry->wait);
INIT_LIST_HEAD(&entry->list);
+ INIT_LIST_HEAD(&entry->root_extent_list);
node = tree_insert(&tree->tree, file_offset,
&entry->rb_node);
- if (node) {
- entry = rb_entry(node, struct btrfs_ordered_extent, rb_node);
- atomic_inc(&entry->refs);
- }
+ BUG_ON(node);
+
set_extent_ordered(&BTRFS_I(inode)->io_tree, file_offset,
entry_end(entry) - 1, GFP_NOFS);
+ spin_lock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
+ list_add_tail(&entry->root_extent_list,
+ &BTRFS_I(inode)->root->fs_info->ordered_extents);
+ spin_unlock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
+
mutex_unlock(&tree->mutex);
BUG_ON(node);
return 0;
struct btrfs_ordered_sum *sum;
if (atomic_dec_and_test(&entry->refs)) {
- while(!list_empty(&entry->list)) {
+ while (!list_empty(&entry->list)) {
cur = entry->list.next;
sum = list_entry(cur, struct btrfs_ordered_sum, list);
list_del(&sum->list);
rb_erase(node, &tree->tree);
tree->last = NULL;
set_bit(BTRFS_ORDERED_COMPLETE, &entry->flags);
+
+ spin_lock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
+ list_del_init(&entry->root_extent_list);
+ spin_unlock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
+
mutex_unlock(&tree->mutex);
wake_up(&entry->wait);
return 0;
}
+/*
+ * wait for all the ordered extents in a root. This is done when balancing
+ * space between drives.
+ */
+int btrfs_wait_ordered_extents(struct btrfs_root *root, int nocow_only)
+{
+ struct list_head splice;
+ struct list_head *cur;
+ struct btrfs_ordered_extent *ordered;
+ struct inode *inode;
+
+ INIT_LIST_HEAD(&splice);
+
+ spin_lock(&root->fs_info->ordered_extent_lock);
+ list_splice_init(&root->fs_info->ordered_extents, &splice);
+ while (!list_empty(&splice)) {
+ cur = splice.next;
+ ordered = list_entry(cur, struct btrfs_ordered_extent,
+ root_extent_list);
+ if (nocow_only &&
+ !test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags) &&
+ !test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags)) {
+ list_move(&ordered->root_extent_list,
+ &root->fs_info->ordered_extents);
+ cond_resched_lock(&root->fs_info->ordered_extent_lock);
+ continue;
+ }
+
+ list_del_init(&ordered->root_extent_list);
+ atomic_inc(&ordered->refs);
+
+ /*
+ * the inode may be getting freed (in sys_unlink path).
+ */
+ inode = igrab(ordered->inode);
+
+ spin_unlock(&root->fs_info->ordered_extent_lock);
+
+ if (inode) {
+ btrfs_start_ordered_extent(inode, ordered, 1);
+ btrfs_put_ordered_extent(ordered);
+ iput(inode);
+ } else {
+ btrfs_put_ordered_extent(ordered);
+ }
+
+ spin_lock(&root->fs_info->ordered_extent_lock);
+ }
+ spin_unlock(&root->fs_info->ordered_extent_lock);
+ return 0;
+}
+
/*
* Used to start IO or wait for a given ordered extent to finish.
*
* start IO on any dirty ones so the wait doesn't stall waiting
* for pdflush to find them
*/
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
- do_sync_file_range(file, start, end, SYNC_FILE_RANGE_WRITE);
-#else
- do_sync_mapping_range(inode->i_mapping, start, end,
- SYNC_FILE_RANGE_WRITE);
-#endif
- if (wait)
+ btrfs_fdatawrite_range(inode->i_mapping, start, end, WB_SYNC_ALL);
+ if (wait) {
wait_event(entry->wait, test_bit(BTRFS_ORDERED_COMPLETE,
&entry->flags));
+ }
}
/*
* Used to wait on ordered extents across a large range of bytes.
*/
-void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
+int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
{
u64 end;
u64 orig_end;
u64 wait_end;
struct btrfs_ordered_extent *ordered;
- u64 mask = BTRFS_I(inode)->root->sectorsize - 1;
if (start + len < start) {
- wait_end = (inode->i_size + mask) & ~mask;
- orig_end = (u64)-1;
+ orig_end = INT_LIMIT(loff_t);
} else {
orig_end = start + len - 1;
- wait_end = orig_end;
+ if (orig_end > INT_LIMIT(loff_t))
+ orig_end = INT_LIMIT(loff_t);
}
-
+ wait_end = orig_end;
+again:
/* start IO across the range first to instantiate any delalloc
* extents
*/
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
- do_sync_file_range(file, start, wait_end, SYNC_FILE_RANGE_WRITE);
-#else
- do_sync_mapping_range(inode->i_mapping, start, wait_end,
- SYNC_FILE_RANGE_WRITE);
-#endif
- end = orig_end;
- wait_on_extent_writeback(&BTRFS_I(inode)->io_tree, start, orig_end);
+ btrfs_fdatawrite_range(inode->i_mapping, start, orig_end, WB_SYNC_NONE);
+
+ /* The compression code will leave pages locked but return from
+ * writepage without setting the page writeback. Starting again
+ * with WB_SYNC_ALL will end up waiting for the IO to actually start.
+ */
+ btrfs_fdatawrite_range(inode->i_mapping, start, orig_end, WB_SYNC_ALL);
- while(1) {
+ btrfs_wait_on_page_writeback_range(inode->i_mapping,
+ start >> PAGE_CACHE_SHIFT,
+ orig_end >> PAGE_CACHE_SHIFT);
+
+ end = orig_end;
+ while (1) {
ordered = btrfs_lookup_first_ordered_extent(inode, end);
- if (!ordered) {
+ if (!ordered)
break;
- }
if (ordered->file_offset > orig_end) {
btrfs_put_ordered_extent(ordered);
break;
break;
end--;
}
+ if (test_range_bit(&BTRFS_I(inode)->io_tree, start, orig_end,
+ EXTENT_ORDERED | EXTENT_DELALLOC, 0)) {
+ schedule_timeout(1);
+ goto again;
+ }
+ return 0;
}
/*
* if none is found
*/
struct btrfs_ordered_extent *
-btrfs_lookup_first_ordered_extent(struct inode * inode, u64 file_offset)
+btrfs_lookup_first_ordered_extent(struct inode *inode, u64 file_offset)
{
struct btrfs_ordered_inode_tree *tree;
struct rb_node *node;
* yet
*/
node = &ordered->rb_node;
- while(1) {
+ while (1) {
node = rb_prev(node);
if (!node)
break;
* between our ordered extent and the next one.
*/
test = rb_entry(node, struct btrfs_ordered_extent, rb_node);
- if (test->file_offset > entry_end(ordered)) {
- i_size_test = test->file_offset - 1;
- }
+ if (test->file_offset > entry_end(ordered))
+ i_size_test = test->file_offset;
} else {
i_size_test = i_size_read(inode);
}
* disk_i_size to the end of the region.
*/
if (i_size_test > entry_end(ordered) &&
- !test_range_bit(io_tree, entry_end(ordered), i_size_test,
+ !test_range_bit(io_tree, entry_end(ordered), i_size_test - 1,
EXTENT_DELALLOC, 0)) {
new_i_size = min_t(u64, i_size_test, i_size_read(inode));
}
* try to find a checksum. This is used because we allow pages to
* be reclaimed before their checksum is actually put into the btree
*/
-int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u32 *sum)
+int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
+ u32 *sum)
{
struct btrfs_ordered_sum *ordered_sum;
struct btrfs_sector_sum *sector_sums;
struct btrfs_ordered_extent *ordered;
struct btrfs_ordered_inode_tree *tree = &BTRFS_I(inode)->ordered_tree;
- struct list_head *cur;
unsigned long num_sectors;
unsigned long i;
u32 sectorsize = BTRFS_I(inode)->root->sectorsize;
return 1;
mutex_lock(&tree->mutex);
- list_for_each_prev(cur, &ordered->list) {
- ordered_sum = list_entry(cur, struct btrfs_ordered_sum, list);
- if (offset >= ordered_sum->file_offset) {
+ list_for_each_entry_reverse(ordered_sum, &ordered->list, list) {
+ if (disk_bytenr >= ordered_sum->bytenr) {
num_sectors = ordered_sum->len / sectorsize;
- sector_sums = &ordered_sum->sums;
+ sector_sums = ordered_sum->sums;
for (i = 0; i < num_sectors; i++) {
- if (sector_sums[i].offset == offset) {
-printk("find ordered sum inode %lu offset %Lu\n", inode->i_ino, offset);
+ if (sector_sums[i].bytenr == disk_bytenr) {
*sum = sector_sums[i].sum;
ret = 0;
goto out;
}
out:
mutex_unlock(&tree->mutex);
+ btrfs_put_ordered_extent(ordered);
return ret;
}
+
+/**
+ * taken from mm/filemap.c because it isn't exported
+ *
+ * __filemap_fdatawrite_range - start writeback on mapping dirty pages in range
+ * @mapping: address space structure to write
+ * @start: offset in bytes where the range starts
+ * @end: offset in bytes where the range ends (inclusive)
+ * @sync_mode: enable synchronous operation
+ *
+ * Start writeback against all of a mapping's dirty pages that lie
+ * within the byte offsets <start, end> inclusive.
+ *
+ * If sync_mode is WB_SYNC_ALL then this is a "data integrity" operation, as
+ * opposed to a regular memory cleansing writeback. The difference between
+ * these two operations is that if a dirty page/buffer is encountered, it must
+ * be waited upon, and not just skipped over.
+ */
+int btrfs_fdatawrite_range(struct address_space *mapping, loff_t start,
+ loff_t end, int sync_mode)
+{
+ struct writeback_control wbc = {
+ .sync_mode = sync_mode,
+ .nr_to_write = mapping->nrpages * 2,
+ .range_start = start,
+ .range_end = end,
+ .for_writepages = 1,
+ };
+ return btrfs_writepages(mapping, &wbc);
+}
+
+/**
+ * taken from mm/filemap.c because it isn't exported
+ *
+ * wait_on_page_writeback_range - wait for writeback to complete
+ * @mapping: target address_space
+ * @start: beginning page index
+ * @end: ending page index
+ *
+ * Wait for writeback to complete against pages indexed by start->end
+ * inclusive
+ */
+int btrfs_wait_on_page_writeback_range(struct address_space *mapping,
+ pgoff_t start, pgoff_t end)
+{
+ struct pagevec pvec;
+ int nr_pages;
+ int ret = 0;
+ pgoff_t index;
+
+ if (end < start)
+ return 0;
+
+ pagevec_init(&pvec, 0);
+ index = start;
+ while ((index <= end) &&
+ (nr_pages = pagevec_lookup_tag(&pvec, mapping, &index,
+ PAGECACHE_TAG_WRITEBACK,
+ min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1)) != 0) {
+ unsigned i;
+
+ for (i = 0; i < nr_pages; i++) {
+ struct page *page = pvec.pages[i];
+
+ /* until radix tree lookup accepts end_index */
+ if (page->index > end)
+ continue;
+
+ wait_on_page_writeback(page);
+ if (PageError(page))
+ ret = -EIO;
+ }
+ pagevec_release(&pvec);
+ cond_resched();
+ }
+
+ /* Check for outstanding write errors */
+ if (test_and_clear_bit(AS_ENOSPC, &mapping->flags))
+ ret = -ENOSPC;
+ if (test_and_clear_bit(AS_EIO, &mapping->flags))
+ ret = -EIO;
+
+ return ret;
+}