#include <linux/time.h>
#include <linux/init.h>
#include <linux/string.h>
-#include <linux/smp_lock.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
-#include <linux/version.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
/* simple helper to fault in pages and copy. This should go away
* and be replaced with calls into generic code.
*/
-static int noinline btrfs_copy_from_user(loff_t pos, int num_pages,
+static noinline int btrfs_copy_from_user(loff_t pos, int num_pages,
int write_bytes,
struct page **prepared_pages,
- const char __user * buf)
+ const char __user *buf)
{
long page_fault = 0;
int i;
/*
* unlocks pages after btrfs_file_write is done with them
*/
-static void noinline btrfs_drop_pages(struct page **pages, size_t num_pages)
+static noinline void btrfs_drop_pages(struct page **pages, size_t num_pages)
{
size_t i;
for (i = 0; i < num_pages; i++) {
* this also makes the decision about creating an inline extent vs
* doing real data extents, marking pages dirty and delalloc as required.
*/
-static int noinline dirty_and_release_pages(struct btrfs_trans_handle *trans,
+static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct file *file,
struct page **pages,
int err = 0;
int i;
struct inode *inode = fdentry(file)->d_inode;
- struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
- u64 hint_byte;
u64 num_bytes;
u64 start_pos;
u64 end_of_last_block;
root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
end_of_last_block = start_pos + num_bytes - 1;
+ err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
+ if (err)
+ return err;
- lock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
- trans = btrfs_join_transaction(root, 1);
- if (!trans) {
- err = -ENOMEM;
- goto out_unlock;
- }
- btrfs_set_trans_block_group(trans, inode);
- hint_byte = 0;
-
- if ((end_of_last_block & 4095) == 0) {
- printk("strange end of last %Lu %zu %Lu\n", start_pos, write_bytes, end_of_last_block);
- }
- set_extent_uptodate(io_tree, start_pos, end_of_last_block, GFP_NOFS);
-
- /* check for reserved extents on each page, we don't want
- * to reset the delalloc bit on things that already have
- * extents reserved.
- */
- btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
for (i = 0; i < num_pages; i++) {
struct page *p = pages[i];
SetPageUptodate(p);
}
if (end_pos > isize) {
i_size_write(inode, end_pos);
- btrfs_update_inode(trans, root, inode);
+ /* we've only changed i_size in ram, and we haven't updated
+ * the disk i_size. There is no need to log the inode
+ * at this time.
+ */
}
- err = btrfs_end_transaction(trans, root);
-out_unlock:
- unlock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
return err;
}
len = (u64)-1;
testend = 0;
}
- while(1) {
+ while (1) {
if (!split)
split = alloc_extent_map(GFP_NOFS);
if (!split2)
split2 = alloc_extent_map(GFP_NOFS);
- spin_lock(&em_tree->lock);
+ write_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree, start, len);
if (!em) {
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
break;
}
flags = em->flags;
if (skip_pinned && test_bit(EXTENT_FLAG_PINNED, &em->flags)) {
- spin_unlock(&em_tree->lock);
if (em->start <= start &&
(!testend || em->start + em->len >= start + len)) {
free_extent_map(em);
+ write_unlock(&em_tree->lock);
break;
}
if (start < em->start) {
start = em->start + em->len;
}
free_extent_map(em);
+ write_unlock(&em_tree->lock);
continue;
}
compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
free_extent_map(split);
split = NULL;
}
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
/* once for us */
free_extent_map(em);
return 0;
}
-int btrfs_check_file(struct btrfs_root *root, struct inode *inode)
-{
- return 0;
-#if 0
- struct btrfs_path *path;
- struct btrfs_key found_key;
- struct extent_buffer *leaf;
- struct btrfs_file_extent_item *extent;
- u64 last_offset = 0;
- int nritems;
- int slot;
- int found_type;
- int ret;
- int err = 0;
- u64 extent_end = 0;
-
- path = btrfs_alloc_path();
- ret = btrfs_lookup_file_extent(NULL, root, path, inode->i_ino,
- last_offset, 0);
- while(1) {
- nritems = btrfs_header_nritems(path->nodes[0]);
- if (path->slots[0] >= nritems) {
- ret = btrfs_next_leaf(root, path);
- if (ret)
- goto out;
- nritems = btrfs_header_nritems(path->nodes[0]);
- }
- slot = path->slots[0];
- leaf = path->nodes[0];
- btrfs_item_key_to_cpu(leaf, &found_key, slot);
- if (found_key.objectid != inode->i_ino)
- break;
- if (found_key.type != BTRFS_EXTENT_DATA_KEY)
- goto out;
-
- if (found_key.offset < last_offset) {
- WARN_ON(1);
- btrfs_print_leaf(root, leaf);
- printk("inode %lu found offset %Lu expected %Lu\n",
- inode->i_ino, found_key.offset, last_offset);
- err = 1;
- goto out;
- }
- extent = btrfs_item_ptr(leaf, slot,
- struct btrfs_file_extent_item);
- found_type = btrfs_file_extent_type(leaf, extent);
- if (found_type == BTRFS_FILE_EXTENT_REG) {
- extent_end = found_key.offset +
- btrfs_file_extent_num_bytes(leaf, extent);
- } else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
- struct btrfs_item *item;
- item = btrfs_item_nr(leaf, slot);
- extent_end = found_key.offset +
- btrfs_file_extent_inline_len(leaf, extent);
- extent_end = (extent_end + root->sectorsize - 1) &
- ~((u64)root->sectorsize -1 );
- }
- last_offset = extent_end;
- path->slots[0]++;
- }
- if (0 && last_offset < inode->i_size) {
- WARN_ON(1);
- btrfs_print_leaf(root, leaf);
- printk("inode %lu found offset %Lu size %Lu\n", inode->i_ino,
- last_offset, inode->i_size);
- err = 1;
-
- }
-out:
- btrfs_free_path(path);
- return err;
-#endif
-}
-
/*
* this is very complex, but the basic idea is to drop all extents
* in the range start - end. hint_block is filled in with a block number
* inline_limit is used to tell this code which offsets in the file to keep
* if they contain inline extents.
*/
-int noinline btrfs_drop_extents(struct btrfs_trans_handle *trans,
+noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans,
struct btrfs_root *root, struct inode *inode,
- u64 start, u64 end, u64 inline_limit, u64 *hint_byte)
+ u64 start, u64 end, u64 locked_end,
+ u64 inline_limit, u64 *hint_byte, int drop_cache)
{
u64 extent_end = 0;
- u64 locked_end = end;
u64 search_start = start;
- u64 leaf_start;
u64 ram_bytes = 0;
- u64 orig_parent = 0;
u64 disk_bytenr = 0;
+ u64 orig_locked_end = locked_end;
u8 compression;
u8 encryption;
u16 other_encoding = 0;
- u64 root_gen;
- u64 root_owner;
struct extent_buffer *leaf;
struct btrfs_file_extent_item *extent;
struct btrfs_path *path;
int ret;
inline_limit = 0;
- btrfs_drop_extent_cache(inode, start, end - 1, 0);
+ if (drop_cache)
+ btrfs_drop_extent_cache(inode, start, end - 1, 0);
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
- while(1) {
+ while (1) {
recow = 0;
btrfs_release_path(root, path);
ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino,
bookend = 0;
found_extent = 0;
found_inline = 0;
- leaf_start = 0;
- root_gen = 0;
- root_owner = 0;
compression = 0;
encryption = 0;
extent = NULL;
if (found_extent) {
read_extent_buffer(leaf, &old, (unsigned long)extent,
sizeof(old));
- root_gen = btrfs_header_generation(leaf);
- root_owner = btrfs_header_owner(leaf);
- leaf_start = leaf->start;
}
if (end < extent_end && end >= key.offset) {
}
locked_end = extent_end;
}
- orig_parent = path->nodes[0]->start;
disk_bytenr = le64_to_cpu(old.disk_bytenr);
if (disk_bytenr != 0) {
ret = btrfs_inc_extent_ref(trans, root,
disk_bytenr,
- le64_to_cpu(old.disk_num_bytes),
- orig_parent, root->root_key.objectid,
- trans->transid, inode->i_ino);
+ le64_to_cpu(old.disk_num_bytes), 0,
+ root->root_key.objectid,
+ key.objectid, key.offset -
+ le64_to_cpu(old.offset));
BUG_ON(ret);
}
}
inode_sub_bytes(inode, old_num -
new_num);
}
- if (!compression && !encryption) {
- btrfs_set_file_extent_ram_bytes(leaf,
- extent, new_num);
- }
btrfs_set_file_extent_num_bytes(leaf,
extent, new_num);
btrfs_mark_buffer_dirty(leaf);
btrfs_set_key_type(&ins, BTRFS_EXTENT_DATA_KEY);
btrfs_release_path(root, path);
+ path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, root, path, &ins,
sizeof(*extent));
BUG_ON(ret);
ram_bytes);
btrfs_set_file_extent_type(leaf, extent, found_type);
+ btrfs_unlock_up_safe(path, 1);
btrfs_mark_buffer_dirty(path->nodes[0]);
+ btrfs_set_lock_blocking(path->nodes[0]);
- if (disk_bytenr != 0) {
- ret = btrfs_update_extent_ref(trans, root,
- disk_bytenr, orig_parent,
- leaf->start,
- root->root_key.objectid,
- trans->transid, ins.objectid);
-
- BUG_ON(ret);
- }
+ path->leave_spinning = 0;
btrfs_release_path(root, path);
- if (disk_bytenr != 0) {
+ if (disk_bytenr != 0)
inode_add_bytes(inode, extent_end - end);
- }
}
if (found_extent && !keep) {
ret = btrfs_free_extent(trans, root,
old_disk_bytenr,
le64_to_cpu(old.disk_num_bytes),
- leaf_start, root_owner,
- root_gen, key.objectid, 0);
+ 0, root->root_key.objectid,
+ key.objectid, key.offset -
+ le64_to_cpu(old.offset));
BUG_ON(ret);
*hint_byte = old_disk_bytenr;
}
}
out:
btrfs_free_path(path);
- if (locked_end > end) {
- unlock_extent(&BTRFS_I(inode)->io_tree, end, locked_end - 1,
- GFP_NOFS);
+ if (locked_end > orig_locked_end) {
+ unlock_extent(&BTRFS_I(inode)->io_tree, orig_locked_end,
+ locked_end - 1, GFP_NOFS);
}
- btrfs_check_file(root, inode);
return ret;
}
u64 bytenr;
u64 num_bytes;
u64 extent_end;
- u64 extent_offset;
+ u64 orig_offset;
u64 other_start;
u64 other_end;
u64 split = start;
u64 locked_end = end;
- u64 orig_parent;
int extent_type;
int split_end = 1;
int ret;
bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi);
- extent_offset = btrfs_file_extent_offset(leaf, fi);
+ orig_offset = key.offset - btrfs_file_extent_offset(leaf, fi);
if (key.offset == start)
split = end;
if (key.offset == start && extent_end == end) {
int del_nr = 0;
int del_slot = 0;
- u64 leaf_owner = btrfs_header_owner(leaf);
- u64 leaf_gen = btrfs_header_generation(leaf);
other_start = end;
other_end = 0;
if (extent_mergeable(leaf, path->slots[0] + 1, inode->i_ino,
del_slot = path->slots[0] + 1;
del_nr++;
ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
- leaf->start, leaf_owner,
- leaf_gen, inode->i_ino, 0);
+ 0, root->root_key.objectid,
+ inode->i_ino, orig_offset);
BUG_ON(ret);
}
other_start = 0;
del_slot = path->slots[0];
del_nr++;
ret = btrfs_free_extent(trans, root, bytenr, num_bytes,
- leaf->start, leaf_owner,
- leaf_gen, inode->i_ino, 0);
+ 0, root->root_key.objectid,
+ inode->i_ino, orig_offset);
BUG_ON(ret);
}
split_end = 0;
ret = btrfs_del_items(trans, root, path, del_slot, del_nr);
BUG_ON(ret);
- goto done;
+ goto release;
} else if (split == start) {
if (locked_end < extent_end) {
ret = try_lock_extent(&BTRFS_I(inode)->io_tree,
locked_end = extent_end;
}
btrfs_set_file_extent_num_bytes(leaf, fi, split - key.offset);
- extent_offset += split - key.offset;
} else {
BUG_ON(key.offset != start);
- btrfs_set_file_extent_offset(leaf, fi, extent_offset +
- split - key.offset);
- btrfs_set_file_extent_num_bytes(leaf, fi, extent_end - split);
key.offset = split;
+ btrfs_set_file_extent_offset(leaf, fi, key.offset -
+ orig_offset);
+ btrfs_set_file_extent_num_bytes(leaf, fi, extent_end - split);
btrfs_set_item_key_safe(trans, root, path, &key);
extent_end = split;
}
struct btrfs_file_extent_item);
key.offset = split;
btrfs_set_item_key_safe(trans, root, path, &key);
- btrfs_set_file_extent_offset(leaf, fi, extent_offset);
+ btrfs_set_file_extent_offset(leaf, fi, key.offset -
+ orig_offset);
btrfs_set_file_extent_num_bytes(leaf, fi,
other_end - split);
goto done;
btrfs_mark_buffer_dirty(leaf);
- orig_parent = leaf->start;
- ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes,
- orig_parent, root->root_key.objectid,
- trans->transid, inode->i_ino);
+ ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes, 0,
+ root->root_key.objectid,
+ inode->i_ino, orig_offset);
BUG_ON(ret);
btrfs_release_path(root, path);
btrfs_set_file_extent_type(leaf, fi, extent_type);
btrfs_set_file_extent_disk_bytenr(leaf, fi, bytenr);
btrfs_set_file_extent_disk_num_bytes(leaf, fi, num_bytes);
- btrfs_set_file_extent_offset(leaf, fi, extent_offset);
+ btrfs_set_file_extent_offset(leaf, fi, key.offset - orig_offset);
btrfs_set_file_extent_num_bytes(leaf, fi, extent_end - key.offset);
btrfs_set_file_extent_ram_bytes(leaf, fi, num_bytes);
btrfs_set_file_extent_compression(leaf, fi, 0);
btrfs_set_file_extent_encryption(leaf, fi, 0);
btrfs_set_file_extent_other_encoding(leaf, fi, 0);
-
- if (orig_parent != leaf->start) {
- ret = btrfs_update_extent_ref(trans, root, bytenr,
- orig_parent, leaf->start,
- root->root_key.objectid,
- trans->transid, inode->i_ino);
- BUG_ON(ret);
- }
done:
btrfs_mark_buffer_dirty(leaf);
+
+release:
btrfs_release_path(root, path);
if (split_end && split == start) {
split = end;
* waits for data=ordered extents to finish before allowing the pages to be
* modified.
*/
-static int noinline prepare_pages(struct btrfs_root *root, struct file *file,
+static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
struct page **pages, size_t num_pages,
loff_t pos, unsigned long first_index,
unsigned long last_index, size_t write_bytes)
struct btrfs_ordered_extent *ordered;
lock_extent(&BTRFS_I(inode)->io_tree,
start_pos, last_pos - 1, GFP_NOFS);
- ordered = btrfs_lookup_first_ordered_extent(inode, last_pos -1);
+ ordered = btrfs_lookup_first_ordered_extent(inode,
+ last_pos - 1);
if (ordered &&
ordered->file_offset + ordered->len > start_pos &&
ordered->file_offset < last_pos) {
start_pos = pos;
vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+
+ /* do the reserve before the mutex lock in case we have to do some
+ * flushing. We wouldn't deadlock, but this is more polite.
+ */
+ err = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+ if (err)
+ goto out_nolock;
+
+ mutex_lock(&inode->i_mutex);
+
current->backing_dev_info = inode->i_mapping->backing_dev_info;
err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode));
if (err)
- goto out_nolock;
+ goto out;
+
if (count == 0)
- goto out_nolock;
+ goto out;
err = file_remove_suid(file);
if (err)
- goto out_nolock;
+ goto out;
+
file_update_time(file);
pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
- mutex_lock(&inode->i_mutex);
+ /* generic_write_checks can change our pos */
+ start_pos = pos;
+
BTRFS_I(inode)->sequence++;
first_index = pos >> PAGE_CACHE_SHIFT;
last_index = (pos + count) >> PAGE_CACHE_SHIFT;
/*
- * if this is a nodatasum mount, force summing off for the inode
- * all the time. That way a later mount with summing on won't
- * get confused
- */
- if (btrfs_test_opt(root, NODATASUM))
- btrfs_set_flag(inode, NODATASUM);
-
- /*
* there are lots of better ways to do this, but this code
* makes sure the first and last page in the file range are
* up to date and ready for cow
}
}
- while(count > 0) {
+ while (count > 0) {
size_t offset = pos & (PAGE_CACHE_SIZE - 1);
size_t write_bytes = min(count, nrptrs *
(size_t)PAGE_CACHE_SIZE -
PAGE_CACHE_SHIFT;
WARN_ON(num_pages > nrptrs);
- memset(pages, 0, sizeof(pages));
+ memset(pages, 0, sizeof(struct page *) * nrptrs);
- ret = btrfs_check_free_space(root, write_bytes, 0);
+ ret = btrfs_check_data_free_space(root, inode, write_bytes);
if (ret)
goto out;
ret = prepare_pages(root, file, pages, num_pages,
pos, first_index, last_index,
write_bytes);
- if (ret)
+ if (ret) {
+ btrfs_free_reserved_data_space(root, inode,
+ write_bytes);
goto out;
+ }
ret = btrfs_copy_from_user(pos, num_pages,
write_bytes, pages, buf);
if (ret) {
+ btrfs_free_reserved_data_space(root, inode,
+ write_bytes);
btrfs_drop_pages(pages, num_pages);
goto out;
}
ret = dirty_and_release_pages(NULL, root, file, pages,
num_pages, pos, write_bytes);
btrfs_drop_pages(pages, num_pages);
- if (ret)
+ if (ret) {
+ btrfs_free_reserved_data_space(root, inode,
+ write_bytes);
goto out;
+ }
if (will_write) {
- btrfs_fdatawrite_range(inode->i_mapping, pos,
- pos + write_bytes - 1,
- WB_SYNC_NONE);
+ filemap_fdatawrite_range(inode->i_mapping, pos,
+ pos + write_bytes - 1);
} else {
balance_dirty_pages_ratelimited_nr(inode->i_mapping,
num_pages);
}
out:
mutex_unlock(&inode->i_mutex);
+ if (ret)
+ err = ret;
+ btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
out_nolock:
kfree(pages);
page_cache_release(pinned[1]);
*ppos = pos;
+ /*
+ * we want to make sure fsync finds this change
+ * but we haven't joined a transaction running right now.
+ *
+ * Later on, someone is sure to update the inode and get the
+ * real transid recorded.
+ *
+ * We set last_trans now to the fs_info generation + 1,
+ * this will either be one more than the running transaction
+ * or the generation used for the next transaction if there isn't
+ * one running right now.
+ */
+ BTRFS_I(inode)->last_trans = root->fs_info->generation + 1;
+
if (num_written > 0 && will_write) {
struct btrfs_trans_handle *trans;
ret = btrfs_log_dentry_safe(trans, root,
file->f_dentry);
if (ret == 0) {
- btrfs_sync_log(trans, root);
- btrfs_end_transaction(trans, root);
+ ret = btrfs_sync_log(trans, root);
+ if (ret == 0)
+ btrfs_end_transaction(trans, root);
+ else
+ btrfs_commit_transaction(trans, root);
} else {
btrfs_commit_transaction(trans, root);
}
return num_written ? num_written : err;
}
-int btrfs_release_file(struct inode * inode, struct file * filp)
+int btrfs_release_file(struct inode *inode, struct file *filp)
{
+ /*
+ * ordered_data_close is set by settattr when we are about to truncate
+ * a file from a non-zero size to a zero size. This tries to
+ * flush down new bytes that may have been written if the
+ * application were using truncate to replace a file in place.
+ */
+ if (BTRFS_I(inode)->ordered_data_close) {
+ BTRFS_I(inode)->ordered_data_close = 0;
+ btrfs_add_ordered_operation(NULL, BTRFS_I(inode)->root, inode);
+ if (inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT)
+ filemap_flush(inode->i_mapping);
+ }
if (filp->private_data)
btrfs_ioctl_trans_end(filp);
return 0;
}
mutex_unlock(&root->fs_info->trans_mutex);
- root->fs_info->tree_log_batch++;
+ root->log_batch++;
filemap_fdatawrite(inode->i_mapping);
btrfs_wait_ordered_range(inode, 0, (u64)-1);
- root->fs_info->tree_log_batch++;
+ root->log_batch++;
+ if (datasync && !(inode->i_state & I_DIRTY_PAGES))
+ goto out;
/*
* ok we haven't committed the transaction yet, lets do a commit
*/
- if (file->private_data)
+ if (file && file->private_data)
btrfs_ioctl_trans_end(file);
trans = btrfs_start_transaction(root, 1);
goto out;
}
- ret = btrfs_log_dentry_safe(trans, root, file->f_dentry);
- if (ret < 0) {
+ ret = btrfs_log_dentry_safe(trans, root, dentry);
+ if (ret < 0)
goto out;
- }
/* we've logged all the items and now have a consistent
* version of the file in the log. It is possible that
* file again, but that will end up using the synchronization
* inside btrfs_sync_log to keep things safe.
*/
- mutex_unlock(&file->f_dentry->d_inode->i_mutex);
+ mutex_unlock(&dentry->d_inode->i_mutex);
if (ret > 0) {
ret = btrfs_commit_transaction(trans, root);
} else {
- btrfs_sync_log(trans, root);
- ret = btrfs_end_transaction(trans, root);
+ ret = btrfs_sync_log(trans, root);
+ if (ret == 0)
+ ret = btrfs_end_transaction(trans, root);
+ else
+ ret = btrfs_commit_transaction(trans, root);
}
- mutex_lock(&file->f_dentry->d_inode->i_mutex);
+ mutex_lock(&dentry->d_inode->i_mutex);
out:
return ret > 0 ? EIO : ret;
}
-static struct vm_operations_struct btrfs_file_vm_ops = {
+static const struct vm_operations_struct btrfs_file_vm_ops = {
.fault = filemap_fault,
.page_mkwrite = btrfs_page_mkwrite,
};
return 0;
}
-struct file_operations btrfs_file_operations = {
+const struct file_operations btrfs_file_operations = {
.llseek = generic_file_llseek,
.read = do_sync_read,
.aio_read = generic_file_aio_read,
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff