aboutsummaryrefslogtreecommitdiff
path: root/fs/btrfs/file.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/btrfs/file.c')
-rw-r--r--fs/btrfs/file.c281
1 files changed, 211 insertions, 70 deletions
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index aeb84469d2c..a005fe2c072 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -24,17 +24,18 @@
#include <linux/string.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
+#include <linux/aio.h>
#include <linux/falloc.h>
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
#include <linux/slab.h>
+#include <linux/btrfs.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "btrfs_inode.h"
-#include "ioctl.h"
#include "print-tree.h"
#include "tree-log.h"
#include "locking.h"
@@ -192,8 +193,8 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
* the same inode in the tree, we will merge them together (by
* __btrfs_add_inode_defrag()) and free the one that we want to requeue.
*/
-void btrfs_requeue_inode_defrag(struct inode *inode,
- struct inode_defrag *defrag)
+static void btrfs_requeue_inode_defrag(struct inode *inode,
+ struct inode_defrag *defrag)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret;
@@ -308,10 +309,6 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info,
ret = PTR_ERR(inode_root);
goto cleanup;
}
- if (btrfs_root_refs(&inode_root->root_item) == 0) {
- ret = -ENOENT;
- goto cleanup;
- }
key.objectid = defrag->ino;
btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
@@ -374,6 +371,11 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
atomic_inc(&fs_info->defrag_running);
while(1) {
+ /* Pause the auto defragger. */
+ if (test_bit(BTRFS_FS_STATE_REMOUNTING,
+ &fs_info->fs_state))
+ break;
+
if (!__need_auto_defrag(fs_info->tree_root))
break;
@@ -468,7 +470,7 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages,
/*
* unlocks pages after btrfs_file_write is done with them
*/
-void btrfs_drop_pages(struct page **pages, size_t num_pages)
+static void btrfs_drop_pages(struct page **pages, size_t num_pages)
{
size_t i;
for (i = 0; i < num_pages; i++) {
@@ -492,9 +494,9 @@ void btrfs_drop_pages(struct page **pages, size_t num_pages)
* doing real data extents, marking pages dirty and delalloc as required.
*/
int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode,
- struct page **pages, size_t num_pages,
- loff_t pos, size_t write_bytes,
- struct extent_state **cached)
+ struct page **pages, size_t num_pages,
+ loff_t pos, size_t write_bytes,
+ struct extent_state **cached)
{
int err = 0;
int i;
@@ -505,8 +507,7 @@ int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode,
loff_t isize = i_size_read(inode);
start_pos = pos & ~((u64)root->sectorsize - 1);
- num_bytes = (write_bytes + pos - start_pos +
- root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
+ num_bytes = ALIGN(write_bytes + pos - start_pos, root->sectorsize);
end_of_last_block = start_pos + num_bytes - 1;
err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block,
@@ -548,6 +549,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
int testend = 1;
unsigned long flags;
int compressed = 0;
+ bool modified;
WARN_ON(end < start);
if (end == (u64)-1) {
@@ -557,6 +559,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
while (1) {
int no_splits = 0;
+ modified = false;
if (!split)
split = alloc_extent_map();
if (!split2)
@@ -587,6 +590,8 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
}
compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
clear_bit(EXTENT_FLAG_PINNED, &em->flags);
+ clear_bit(EXTENT_FLAG_LOGGING, &flags);
+ modified = !list_empty(&em->list);
remove_extent_mapping(em_tree, em);
if (no_splits)
goto next;
@@ -602,15 +607,15 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
split->block_len = em->block_len;
else
split->block_len = split->len;
+ split->ram_bytes = em->ram_bytes;
split->orig_block_len = max(split->block_len,
em->orig_block_len);
split->generation = gen;
split->bdev = em->bdev;
split->flags = flags;
split->compress_type = em->compress_type;
- ret = add_extent_mapping(em_tree, split);
+ ret = add_extent_mapping(em_tree, split, modified);
BUG_ON(ret); /* Logic error */
- list_move(&split->list, &em_tree->modified_extents);
free_extent_map(split);
split = split2;
split2 = NULL;
@@ -627,6 +632,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
split->generation = gen;
split->orig_block_len = max(em->block_len,
em->orig_block_len);
+ split->ram_bytes = em->ram_bytes;
if (compressed) {
split->block_len = em->block_len;
@@ -638,9 +644,8 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
split->orig_start = em->orig_start;
}
- ret = add_extent_mapping(em_tree, split);
+ ret = add_extent_mapping(em_tree, split, modified);
BUG_ON(ret); /* Logic error */
- list_move(&split->list, &em_tree->modified_extents);
free_extent_map(split);
split = NULL;
}
@@ -816,7 +821,7 @@ next_slot:
memcpy(&new_key, &key, sizeof(new_key));
new_key.offset = end;
- btrfs_set_item_key_safe(trans, root, path, &new_key);
+ btrfs_set_item_key_safe(root, path, &new_key);
extent_offset += end - key.offset;
btrfs_set_file_extent_offset(leaf, fi, extent_offset);
@@ -1032,7 +1037,7 @@ again:
ino, bytenr, orig_offset,
&other_start, &other_end)) {
new_key.offset = end;
- btrfs_set_item_key_safe(trans, root, path, &new_key);
+ btrfs_set_item_key_safe(root, path, &new_key);
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
btrfs_set_file_extent_generation(leaf, fi,
@@ -1066,7 +1071,7 @@ again:
trans->transid);
path->slots[0]++;
new_key.offset = start;
- btrfs_set_item_key_safe(trans, root, path, &new_key);
+ btrfs_set_item_key_safe(root, path, &new_key);
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
@@ -1225,7 +1230,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file,
struct extent_state *cached_state = NULL;
int i;
unsigned long index = pos >> PAGE_CACHE_SHIFT;
- struct inode *inode = fdentry(file)->d_inode;
+ struct inode *inode = file_inode(file);
gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
int err = 0;
int faili = 0;
@@ -1308,17 +1313,69 @@ fail:
}
+static noinline int check_can_nocow(struct inode *inode, loff_t pos,
+ size_t *write_bytes)
+{
+ struct btrfs_trans_handle *trans;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ struct btrfs_ordered_extent *ordered;
+ u64 lockstart, lockend;
+ u64 num_bytes;
+ int ret;
+
+ lockstart = round_down(pos, root->sectorsize);
+ lockend = lockstart + round_up(*write_bytes, root->sectorsize) - 1;
+
+ while (1) {
+ lock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ ordered = btrfs_lookup_ordered_range(inode, lockstart,
+ lockend - lockstart + 1);
+ if (!ordered) {
+ break;
+ }
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ btrfs_start_ordered_extent(inode, ordered, 1);
+ btrfs_put_ordered_extent(ordered);
+ }
+
+ trans = btrfs_join_transaction(root);
+ if (IS_ERR(trans)) {
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+ return PTR_ERR(trans);
+ }
+
+ num_bytes = lockend - lockstart + 1;
+ ret = can_nocow_extent(trans, inode, lockstart, &num_bytes, NULL, NULL,
+ NULL);
+ btrfs_end_transaction(trans, root);
+ if (ret <= 0) {
+ ret = 0;
+ } else {
+ clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend,
+ EXTENT_DIRTY | EXTENT_DELALLOC |
+ EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
+ NULL, GFP_NOFS);
+ *write_bytes = min_t(size_t, *write_bytes, num_bytes);
+ }
+
+ unlock_extent(&BTRFS_I(inode)->io_tree, lockstart, lockend);
+
+ return ret;
+}
+
static noinline ssize_t __btrfs_buffered_write(struct file *file,
struct iov_iter *i,
loff_t pos)
{
- struct inode *inode = fdentry(file)->d_inode;
+ struct inode *inode = file_inode(file);
struct btrfs_root *root = BTRFS_I(inode)->root;
struct page **pages = NULL;
+ u64 release_bytes = 0;
unsigned long first_index;
size_t num_written = 0;
int nrptrs;
int ret = 0;
+ bool only_release_metadata = false;
bool force_page_uptodate = false;
nrptrs = min((iov_iter_count(i) + PAGE_CACHE_SIZE - 1) /
@@ -1339,6 +1396,7 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
offset);
size_t num_pages = (write_bytes + offset +
PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+ size_t reserve_bytes;
size_t dirty_pages;
size_t copied;
@@ -1353,11 +1411,41 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
break;
}
- ret = btrfs_delalloc_reserve_space(inode,
- num_pages << PAGE_CACHE_SHIFT);
+ reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
+ ret = btrfs_check_data_free_space(inode, reserve_bytes);
+ if (ret == -ENOSPC &&
+ (BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
+ BTRFS_INODE_PREALLOC))) {
+ ret = check_can_nocow(inode, pos, &write_bytes);
+ if (ret > 0) {
+ only_release_metadata = true;
+ /*
+ * our prealloc extent may be smaller than
+ * write_bytes, so scale down.
+ */
+ num_pages = (write_bytes + offset +
+ PAGE_CACHE_SIZE - 1) >>
+ PAGE_CACHE_SHIFT;
+ reserve_bytes = num_pages << PAGE_CACHE_SHIFT;
+ ret = 0;
+ } else {
+ ret = -ENOSPC;
+ }
+ }
+
if (ret)
break;
+ ret = btrfs_delalloc_reserve_metadata(inode, reserve_bytes);
+ if (ret) {
+ if (!only_release_metadata)
+ btrfs_free_reserved_data_space(inode,
+ reserve_bytes);
+ break;
+ }
+
+ release_bytes = reserve_bytes;
+
/*
* This is going to setup the pages array with the number of
* pages we want, so we don't really need to worry about the
@@ -1366,11 +1454,8 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
ret = prepare_pages(root, file, pages, num_pages,
pos, first_index, write_bytes,
force_page_uptodate);
- if (ret) {
- btrfs_delalloc_release_space(inode,
- num_pages << PAGE_CACHE_SHIFT);
+ if (ret)
break;
- }
copied = btrfs_copy_from_user(pos, num_pages,
write_bytes, pages, i);
@@ -1400,30 +1485,46 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
* managed to copy.
*/
if (num_pages > dirty_pages) {
+ release_bytes = (num_pages - dirty_pages) <<
+ PAGE_CACHE_SHIFT;
if (copied > 0) {
spin_lock(&BTRFS_I(inode)->lock);
BTRFS_I(inode)->outstanding_extents++;
spin_unlock(&BTRFS_I(inode)->lock);
}
- btrfs_delalloc_release_space(inode,
- (num_pages - dirty_pages) <<
- PAGE_CACHE_SHIFT);
+ if (only_release_metadata)
+ btrfs_delalloc_release_metadata(inode,
+ release_bytes);
+ else
+ btrfs_delalloc_release_space(inode,
+ release_bytes);
}
+ release_bytes = dirty_pages << PAGE_CACHE_SHIFT;
if (copied > 0) {
ret = btrfs_dirty_pages(root, inode, pages,
dirty_pages, pos, copied,
NULL);
if (ret) {
- btrfs_delalloc_release_space(inode,
- dirty_pages << PAGE_CACHE_SHIFT);
btrfs_drop_pages(pages, num_pages);
break;
}
}
+ release_bytes = 0;
btrfs_drop_pages(pages, num_pages);
+ if (only_release_metadata && copied > 0) {
+ u64 lockstart = round_down(pos, root->sectorsize);
+ u64 lockend = lockstart +
+ (dirty_pages << PAGE_CACHE_SHIFT) - 1;
+
+ set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
+ lockend, EXTENT_NORESERVE, NULL,
+ NULL, GFP_NOFS);
+ only_release_metadata = false;
+ }
+
cond_resched();
balance_dirty_pages_ratelimited(inode->i_mapping);
@@ -1436,6 +1537,13 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
kfree(pages);
+ if (release_bytes) {
+ if (only_release_metadata)
+ btrfs_delalloc_release_metadata(inode, release_bytes);
+ else
+ btrfs_delalloc_release_space(inode, release_bytes);
+ }
+
return num_written ? num_written : ret;
}
@@ -1500,7 +1608,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
unsigned long nr_segs, loff_t pos)
{
struct file *file = iocb->ki_filp;
- struct inode *inode = fdentry(file)->d_inode;
+ struct inode *inode = file_inode(file);
struct btrfs_root *root = BTRFS_I(inode)->root;
loff_t *ppos = &iocb->ki_pos;
u64 start_pos;
@@ -1509,8 +1617,6 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
size_t count, ocount;
bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);
- sb_start_write(inode->i_sb);
-
mutex_lock(&inode->i_mutex);
err = generic_segment_checks(iov, &nr_segs, &ocount, VERIFY_READ);
@@ -1544,7 +1650,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
* although we have opened a file as writable, we have
* to stop this write operation to ensure FS consistency.
*/
- if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
+ if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) {
mutex_unlock(&inode->i_mutex);
err = -EROFS;
goto out;
@@ -1612,7 +1718,6 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
if (sync)
atomic_dec(&BTRFS_I(inode)->sync_writers);
out:
- sb_end_write(inode->i_sb);
current->backing_dev_info = NULL;
return num_written ? num_written : err;
}
@@ -1627,7 +1732,20 @@ int btrfs_release_file(struct inode *inode, struct file *filp)
*/
if (test_and_clear_bit(BTRFS_INODE_ORDERED_DATA_CLOSE,
&BTRFS_I(inode)->runtime_flags)) {
- btrfs_add_ordered_operation(NULL, BTRFS_I(inode)->root, inode);
+ struct btrfs_trans_handle *trans;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+
+ /*
+ * We need to block on a committing transaction to keep us from
+ * throwing a ordered operation on to the list and causing
+ * something like sync to deadlock trying to flush out this
+ * inode.
+ */
+ trans = btrfs_start_transaction(root, 0);
+ if (IS_ERR(trans))
+ return PTR_ERR(trans);
+ btrfs_add_ordered_operation(trans, BTRFS_I(inode)->root, inode);
+ btrfs_end_transaction(trans, root);
if (inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT)
filemap_flush(inode->i_mapping);
}
@@ -1654,16 +1772,21 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret = 0;
struct btrfs_trans_handle *trans;
+ bool full_sync = 0;
trace_btrfs_sync_file(file, datasync);
/*
* We write the dirty pages in the range and wait until they complete
* out of the ->i_mutex. If so, we can flush the dirty pages by
- * multi-task, and make the performance up.
+ * multi-task, and make the performance up. See
+ * btrfs_wait_ordered_range for an explanation of the ASYNC check.
*/
atomic_inc(&BTRFS_I(inode)->sync_writers);
- ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+ ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
+ if (!ret && test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
+ &BTRFS_I(inode)->runtime_flags))
+ ret = filemap_fdatawrite_range(inode->i_mapping, start, end);
atomic_dec(&BTRFS_I(inode)->sync_writers);
if (ret)
return ret;
@@ -1675,7 +1798,10 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
* range being left.
*/
atomic_inc(&root->log_batch);
- btrfs_wait_ordered_range(inode, start, end - start + 1);
+ full_sync = test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
+ &BTRFS_I(inode)->runtime_flags);
+ if (full_sync)
+ btrfs_wait_ordered_range(inode, start, end - start + 1);
atomic_inc(&root->log_batch);
/*
@@ -1742,13 +1868,25 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
if (ret != BTRFS_NO_LOG_SYNC) {
if (ret > 0) {
+ /*
+ * If we didn't already wait for ordered extents we need
+ * to do that now.
+ */
+ if (!full_sync)
+ btrfs_wait_ordered_range(inode, start,
+ end - start + 1);
ret = btrfs_commit_transaction(trans, root);
} else {
ret = btrfs_sync_log(trans, root);
- if (ret == 0)
+ if (ret == 0) {
ret = btrfs_end_transaction(trans, root);
- else
+ } else {
+ if (!full_sync)
+ btrfs_wait_ordered_range(inode, start,
+ end -
+ start + 1);
ret = btrfs_commit_transaction(trans, root);
+ }
}
} else {
ret = btrfs_end_transaction(trans, root);
@@ -1847,7 +1985,7 @@ static int fill_holes(struct btrfs_trans_handle *trans, struct inode *inode,
path->slots[0]++;
key.offset = offset;
- btrfs_set_item_key_safe(trans, root, path, &key);
+ btrfs_set_item_key_safe(root, path, &key);
fi = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_file_extent_item);
num_bytes = btrfs_file_extent_num_bytes(leaf, fi) + end -
@@ -1877,6 +2015,7 @@ out:
} else {
hole_em->start = offset;
hole_em->len = end - offset;
+ hole_em->ram_bytes = hole_em->len;
hole_em->orig_start = offset;
hole_em->block_start = EXTENT_MAP_HOLE;
@@ -1889,10 +2028,7 @@ out:
do {
btrfs_drop_extent_cache(inode, offset, end - 1, 0);
write_lock(&em_tree->lock);
- ret = add_extent_mapping(em_tree, hole_em);
- if (!ret)
- list_move(&hole_em->list,
- &em_tree->modified_extents);
+ ret = add_extent_mapping(em_tree, hole_em, 1);
write_unlock(&em_tree->lock);
} while (ret == -EEXIST);
free_extent_map(hole_em);
@@ -2102,8 +2238,9 @@ out:
static long btrfs_fallocate(struct file *file, int mode,
loff_t offset, loff_t len)
{
- struct inode *inode = file->f_path.dentry->d_inode;
+ struct inode *inode = file_inode(file);
struct extent_state *cached_state = NULL;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
u64 cur_offset;
u64 last_byte;
u64 alloc_start;
@@ -2131,12 +2268,11 @@ static long btrfs_fallocate(struct file *file, int mode,
ret = btrfs_check_data_free_space(inode, alloc_end - alloc_start);
if (ret)
return ret;
-
- /*
- * wait for ordered IO before we have any locks. We'll loop again
- * below with the locks held.
- */
- btrfs_wait_ordered_range(inode, alloc_start, alloc_end - alloc_start);
+ if (root->fs_info->quota_enabled) {
+ ret = btrfs_qgroup_reserve(root, alloc_end - alloc_start);
+ if (ret)
+ goto out_reserve_fail;
+ }
mutex_lock(&inode->i_mutex);
ret = inode_newsize_ok(inode, alloc_end);
@@ -2148,8 +2284,23 @@ static long btrfs_fallocate(struct file *file, int mode,
alloc_start);
if (ret)
goto out;
+ } else {
+ /*
+ * If we are fallocating from the end of the file onward we
+ * need to zero out the end of the page if i_size lands in the
+ * middle of a page.
+ */
+ ret = btrfs_truncate_page(inode, inode->i_size, 0, 0);
+ if (ret)
+ goto out;
}
+ /*
+ * wait for ordered IO before we have any locks. We'll loop again
+ * below with the locks held.
+ */
+ btrfs_wait_ordered_range(inode, alloc_start, alloc_end - alloc_start);
+
locked_end = alloc_end - 1;
while (1) {
struct btrfs_ordered_extent *ordered;
@@ -2234,6 +2385,9 @@ static long btrfs_fallocate(struct file *file, int mode,
&cached_state, GFP_NOFS);
out:
mutex_unlock(&inode->i_mutex);
+ if (root->fs_info->quota_enabled)
+ btrfs_qgroup_free(root, alloc_end - alloc_start);
+out_reserve_fail:
/* Let go of our reservation. */
btrfs_free_reserved_data_space(inode, alloc_end - alloc_start);
return ret;
@@ -2379,20 +2533,7 @@ static loff_t btrfs_file_llseek(struct file *file, loff_t offset, int whence)
}
}
- if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) {
- offset = -EINVAL;
- goto out;
- }
- if (offset > inode->i_sb->s_maxbytes) {
- offset = -EINVAL;
- goto out;
- }
-
- /* Special lock needed here? */
- if (offset != file->f_pos) {
- file->f_pos = offset;
- file->f_version = 0;
- }
+ offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes);
out:
mutex_unlock(&inode->i_mutex);
return offset;
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-10-22 01:13:36 +0000