--- fs/ext4/extents.c
+++ fs/ext4/extents.c
@@ -107,11 +107,8 @@ static int ext4_ext_truncate_extend_rest
if (err <= 0)
return err;
err = ext4_truncate_restart_trans(handle, inode, needed);
- /*
- * We have dropped i_data_sem so someone might have cached again
- * an extent we are going to truncate.
- */
- ext4_ext_invalidate_cache(inode);
+ if (err == 0)
+ err = -EAGAIN;
return err;
}
@@ -296,29 +293,44 @@ static inline int ext4_ext_space_root_id
* to allocate @blocks
* Worse case is one block per extent
*/
-int ext4_ext_calc_metadata_amount(struct inode *inode, int blocks)
+int ext4_ext_calc_metadata_amount(struct inode *inode, sector_t lblock)
{
- int lcap, icap, rcap, leafs, idxs, num;
- int newextents = blocks;
+ struct ext4_inode_info *ei = EXT4_I(inode);
+ int idxs, num = 0;
- rcap = ext4_ext_space_root_idx(inode, 0);
- lcap = ext4_ext_space_block(inode, 0);
- icap = ext4_ext_space_block_idx(inode, 0);
+ idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
+ / sizeof(struct ext4_extent_idx));
- /* number of new leaf blocks needed */
- num = leafs = (newextents + lcap - 1) / lcap;
+ /*
+ * If the new delayed allocation block is contiguous with the
+ * previous da block, it can share index blocks with the
+ * previous block, so we only need to allocate a new index
+ * block every idxs leaf blocks. At ldxs**2 blocks, we need
+ * an additional index block, and at ldxs**3 blocks, yet
+ * another index blocks.
+ */
+ if (ei->i_da_metadata_calc_len &&
+ ei->i_da_metadata_calc_last_lblock+1 == lblock) {
+ if ((ei->i_da_metadata_calc_len % idxs) == 0)
+ num++;
+ if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
+ num++;
+ if ((ei->i_da_metadata_calc_len % (idxs*idxs*idxs)) == 0) {
+ num++;
+ ei->i_da_metadata_calc_len = 0;
+ } else
+ ei->i_da_metadata_calc_len++;
+ ei->i_da_metadata_calc_last_lblock++;
+ return num;
+ }
/*
- * Worse case, we need separate index block(s)
- * to link all new leaf blocks
+ * In the worst case we need a new set of index blocks at
+ * every level of the inode's extent tree.
*/
- idxs = (leafs + icap - 1) / icap;
- do {
- num += idxs;
- idxs = (idxs + icap - 1) / icap;
- } while (idxs > rcap);
-
- return num;
+ ei->i_da_metadata_calc_len = 1;
+ ei->i_da_metadata_calc_last_lblock = lblock;
+ return ext_depth(inode) + 1;
}
static int
@@ -2248,7 +2260,7 @@ static int ext4_ext_remove_space(struct
int depth = ext_depth(inode);
struct ext4_ext_path *path;
handle_t *handle;
- int i = 0, err = 0;
+ int i, err;
ext_debug("truncate since %u\n", start);
@@ -2257,23 +2269,26 @@ static int ext4_ext_remove_space(struct
if (IS_ERR(handle))
return PTR_ERR(handle);
+again:
ext4_ext_invalidate_cache(inode);
/*
* We start scanning from right side, freeing all the blocks
* after i_size and walking into the tree depth-wise.
*/
+ depth = ext_depth(inode);
path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
if (path == NULL) {
ext4_journal_stop(handle);
return -ENOMEM;
}
+ path[0].p_depth = depth;
path[0].p_hdr = ext_inode_hdr(inode);
if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
err = -EIO;
goto out;
}
- path[0].p_depth = depth;
+ i = err = 0;
while (i >= 0 && err == 0) {
if (i == depth) {
@@ -2367,6 +2382,8 @@ static int ext4_ext_remove_space(struct
out:
ext4_ext_drop_refs(path);
kfree(path);
+ if (err == -EAGAIN)
+ goto again;
ext4_journal_stop(handle);
return err;
@@ -2431,7 +2448,7 @@ static void bi_complete(struct bio *bio,
/* FIXME!! we need to try to merge to left or right after zero-out */
static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
{
- int ret = -EIO;
+ int ret;
struct bio *bio;
int blkbits, blocksize;
sector_t ee_pblock;
@@ -2455,6 +2472,9 @@ static int ext4_ext_zeroout(struct inode
len = ee_len;
bio = bio_alloc(GFP_NOIO, len);
+ if (!bio)
+ return -ENOMEM;
+
bio->bi_sector = ee_pblock;
bio->bi_bdev = inode->i_sb->s_bdev;
@@ -2482,17 +2502,15 @@ static int ext4_ext_zeroout(struct inode
submit_bio(WRITE, bio);
wait_for_completion(&event);
- if (test_bit(BIO_UPTODATE, &bio->bi_flags))
- ret = 0;
- else {
- ret = -EIO;
- break;
+ if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
+ bio_put(bio);
+ return -EIO;
}
bio_put(bio);
ee_len -= done;
ee_pblock += done << (blkbits - 9);
}
- return ret;
+ return 0;
}
#define EXT4_EXT_ZERO_LEN 7
@@ -2517,11 +2535,21 @@ static int ext4_ext_convert_to_initializ
struct ext4_extent *ex2 = NULL;
struct ext4_extent *ex3 = NULL;
struct ext4_extent_header *eh;
- ext4_lblk_t ee_block;
+ ext4_lblk_t ee_block, eof_block;
unsigned int allocated, ee_len, depth;
ext4_fsblk_t newblock;
int err = 0;
int ret = 0;
+ int may_zeroout;
+
+ ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
+ "block %llu, max_blocks %u\n", inode->i_ino,
+ (unsigned long long)iblock, max_blocks);
+
+ eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
+ inode->i_sb->s_blocksize_bits;
+ if (eof_block < iblock + max_blocks)
+ eof_block = iblock + max_blocks;
depth = ext_depth(inode);
eh = path[depth].p_hdr;
@@ -2530,16 +2558,23 @@ static int ext4_ext_convert_to_initializ
ee_len = ext4_ext_get_actual_len(ex);
allocated = ee_len - (iblock - ee_block);
newblock = iblock - ee_block + ext_pblock(ex);
+
ex2 = ex;
orig_ex.ee_block = ex->ee_block;
orig_ex.ee_len = cpu_to_le16(ee_len);
ext4_ext_store_pblock(&orig_ex, ext_pblock(ex));
+ /*
+ * It is safe to convert extent to initialized via explicit
+ * zeroout only if extent is fully insde i_size or new_size.
+ */
+ may_zeroout = ee_block + ee_len <= eof_block;
+
err = ext4_ext_get_access(handle, inode, path + depth);
if (err)
goto out;
/* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
- if (ee_len <= 2*EXT4_EXT_ZERO_LEN) {
+ if (ee_len <= 2*EXT4_EXT_ZERO_LEN && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -2570,7 +2605,7 @@ static int ext4_ext_convert_to_initializ
if (allocated > max_blocks) {
unsigned int newdepth;
/* If extent has less than EXT4_EXT_ZERO_LEN zerout directly */
- if (allocated <= EXT4_EXT_ZERO_LEN) {
+ if (allocated <= EXT4_EXT_ZERO_LEN && may_zeroout) {
/*
* iblock == ee_block is handled by the zerouout
* at the beginning.
@@ -2646,7 +2681,7 @@ static int ext4_ext_convert_to_initializ
ex3->ee_len = cpu_to_le16(allocated - max_blocks);
ext4_ext_mark_uninitialized(ex3);
err = ext4_ext_insert_extent(handle, inode, path, ex3, 0);
- if (err == -ENOSPC) {
+ if (err == -ENOSPC && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -2670,8 +2705,10 @@ static int ext4_ext_convert_to_initializ
* update the extent length after successful insert of the
* split extent
*/
- orig_ex.ee_len = cpu_to_le16(ee_len -
- ext4_ext_get_actual_len(ex3));
+ ee_len -= ext4_ext_get_actual_len(ex3);
+ orig_ex.ee_len = cpu_to_le16(ee_len);
+ may_zeroout = ee_block + ee_len <= eof_block;
+
depth = newdepth;
ext4_ext_drop_refs(path);
path = ext4_ext_find_extent(inode, iblock, path);
@@ -2695,7 +2732,7 @@ static int ext4_ext_convert_to_initializ
* otherwise give the extent a chance to merge to left
*/
if (le16_to_cpu(orig_ex.ee_len) <= EXT4_EXT_ZERO_LEN &&
- iblock != ee_block) {
+ iblock != ee_block && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -2764,7 +2801,7 @@ static int ext4_ext_convert_to_initializ
goto out;
insert:
err = ext4_ext_insert_extent(handle, inode, path, &newex, 0);
- if (err == -ENOSPC) {
+ if (err == -ENOSPC && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -2824,14 +2861,21 @@ static int ext4_split_unwritten_extents(
struct ext4_extent *ex2 = NULL;
struct ext4_extent *ex3 = NULL;
struct ext4_extent_header *eh;
- ext4_lblk_t ee_block;
+ ext4_lblk_t ee_block, eof_block;
unsigned int allocated, ee_len, depth;
ext4_fsblk_t newblock;
int err = 0;
+ int may_zeroout;
+
+ ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
+ "block %llu, max_blocks %u\n", inode->i_ino,
+ (unsigned long long)iblock, max_blocks);
+
+ eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
+ inode->i_sb->s_blocksize_bits;
+ if (eof_block < iblock + max_blocks)
+ eof_block = iblock + max_blocks;
- ext_debug("ext4_split_unwritten_extents: inode %lu,"
- "iblock %llu, max_blocks %u\n", inode->i_ino,
- (unsigned long long)iblock, max_blocks);
depth = ext_depth(inode);
eh = path[depth].p_hdr;
ex = path[depth].p_ext;
@@ -2839,12 +2883,19 @@ static int ext4_split_unwritten_extents(
ee_len = ext4_ext_get_actual_len(ex);
allocated = ee_len - (iblock - ee_block);
newblock = iblock - ee_block + ext_pblock(ex);
+
ex2 = ex;
orig_ex.ee_block = ex->ee_block;
orig_ex.ee_len = cpu_to_le16(ee_len);
ext4_ext_store_pblock(&orig_ex, ext_pblock(ex));
/*
+ * It is safe to convert extent to initialized via explicit
+ * zeroout only if extent is fully insde i_size or new_size.
+ */
+ may_zeroout = ee_block + ee_len <= eof_block;
+
+ /*
* If the uninitialized extent begins at the same logical
* block where the write begins, and the write completely
* covers the extent, then we don't need to split it.
@@ -2878,7 +2929,7 @@ static int ext4_split_unwritten_extents(
ex3->ee_len = cpu_to_le16(allocated - max_blocks);
ext4_ext_mark_uninitialized(ex3);
err = ext4_ext_insert_extent(handle, inode, path, ex3, flags);
- if (err == -ENOSPC) {
+ if (err == -ENOSPC && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -2902,8 +2953,10 @@ static int ext4_split_unwritten_extents(
* update the extent length after successful insert of the
* split extent
*/
- orig_ex.ee_len = cpu_to_le16(ee_len -
- ext4_ext_get_actual_len(ex3));
+ ee_len -= ext4_ext_get_actual_len(ex3);
+ orig_ex.ee_len = cpu_to_le16(ee_len);
+ may_zeroout = ee_block + ee_len <= eof_block;
+
depth = newdepth;
ext4_ext_drop_refs(path);
path = ext4_ext_find_extent(inode, iblock, path);
@@ -2949,7 +3002,7 @@ static int ext4_split_unwritten_extents(
goto out;
insert:
err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
- if (err == -ENOSPC) {
+ if (err == -ENOSPC && may_zeroout) {
err = ext4_ext_zeroout(inode, &orig_ex);
if (err)
goto fix_extent_len;
@@ -3029,6 +3082,14 @@ out:
return err;
}
+static void unmap_underlying_metadata_blocks(struct block_device *bdev,
+ sector_t block, int count)
+{
+ int i;
+ for (i = 0; i < count; i++)
+ unmap_underlying_metadata(bdev, block + i);
+}
+
static int
ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode,
ext4_lblk_t iblock, unsigned int max_blocks,
@@ -3059,7 +3120,7 @@ ext4_ext_handle_uninitialized_extents(ha
if (io)
io->flag = DIO_AIO_UNWRITTEN;
else
- EXT4_I(inode)->i_state |= EXT4_STATE_DIO_UNWRITTEN;
+ ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN);
goto out;
}
/* async DIO end_io complete, convert the filled extent to written */
@@ -3104,6 +3165,30 @@ out:
} else
allocated = ret;
set_buffer_new(bh_result);
+ /*
+ * if we allocated more blocks than requested
+ * we need to make sure we unmap the extra block
+ * allocated. The actual needed block will get
+ * unmapped later when we find the buffer_head marked
+ * new.
+ */
+ if (allocated > max_blocks) {
+ unmap_underlying_metadata_blocks(inode->i_sb->s_bdev,
+ newblock + max_blocks,
+ allocated - max_blocks);
+ allocated = max_blocks;
+ }
+
+ /*
+ * If we have done fallocate with the offset that is already
+ * delayed allocated, we would have block reservation
+ * and quota reservation done in the delayed write path.
+ * But fallocate would have already updated quota and block
+ * count for this offset. So cancel these reservation
+ */
+ if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
+ ext4_da_update_reserve_space(inode, allocated, 0);
+
map_out:
set_buffer_mapped(bh_result);
out1:
@@ -3144,9 +3229,9 @@ int ext4_ext_get_blocks(handle_t *handle
{
struct ext4_ext_path *path = NULL;
struct ext4_extent_header *eh;
- struct ext4_extent newex, *ex;
+ struct ext4_extent newex, *ex, *last_ex;
ext4_fsblk_t newblock;
- int err = 0, depth, ret, cache_type;
+ int i, err = 0, depth, ret, cache_type;
unsigned int allocated = 0;
struct ext4_allocation_request ar;
ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
@@ -3196,7 +3281,13 @@ int ext4_ext_get_blocks(handle_t *handle
* this situation is possible, though, _during_ tree modification;
* this is why assert can't be put in ext4_ext_find_extent()
*/
- BUG_ON(path[depth].p_ext == NULL && depth != 0);
+ if (path[depth].p_ext == NULL && depth != 0) {
+ ext4_error(inode->i_sb, __func__, "bad extent address "
+ "inode: %lu, iblock: %lu, depth: %d",
+ inode->i_ino, (unsigned long) iblock, depth);
+ err = -EIO;
+ goto out2;
+ }
eh = path[depth].p_hdr;
ex = path[depth].p_ext;
@@ -3315,9 +3406,35 @@ int ext4_ext_get_blocks(handle_t *handle
if (io)
io->flag = DIO_AIO_UNWRITTEN;
else
- EXT4_I(inode)->i_state |=
- EXT4_STATE_DIO_UNWRITTEN;;
+ ext4_set_inode_state(inode,
+ EXT4_STATE_DIO_UNWRITTEN);
+ }
+ }
+
+ if (unlikely(ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))) {
+ if (unlikely(!eh->eh_entries)) {
+ ext4_error(inode->i_sb, __func__,
+ "inode#%lu, eh->eh_entries = 0 and "
+ "EOFBLOCKS_FL set", inode->i_ino);
+ err = -EIO;
+ goto out2;
}
+ last_ex = EXT_LAST_EXTENT(eh);
+ /*
+ * If the current leaf block was reached by looking at
+ * the last index block all the way down the tree, and
+ * we are extending the inode beyond the last extent
+ * in the current leaf block, then clear the
+ * EOFBLOCKS_FL flag.
+ */
+ for (i = depth-1; i >= 0; i--) {
+ if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
+ break;
+ }
+ if ((i < 0) &&
+ (iblock + ar.len > le32_to_cpu(last_ex->ee_block) +
+ ext4_ext_get_actual_len(last_ex)))
+ ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
}
err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
if (err) {
@@ -3333,9 +3450,18 @@ int ext4_ext_get_blocks(handle_t *handle
/* previous routine could use block we allocated */
newblock = ext_pblock(&newex);
allocated = ext4_ext_get_actual_len(&newex);
+ if (allocated > max_blocks)
+ allocated = max_blocks;
set_buffer_new(bh_result);
/*
+ * Update reserved blocks/metadata blocks after successful
+ * block allocation which had been deferred till now.
+ */
+ if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
+ ext4_da_update_reserve_space(inode, allocated, 1);
+
+ /*
* Cache the extent and update transaction to commit on fdatasync only
* when it is _not_ an uninitialized extent.
*/
@@ -3443,6 +3569,13 @@ static void ext4_falloc_update_inode(str
i_size_write(inode, new_size);
if (new_size > EXT4_I(inode)->i_disksize)
ext4_update_i_disksize(inode, new_size);
+ } else {
+ /*
+ * Mark that we allocate beyond EOF so the subsequent truncate
+ * can proceed even if the new size is the same as i_size.
+ */
+ if (new_size > i_size_read(inode))
+ ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
}
}
@@ -3470,7 +3603,7 @@ long ext4_fallocate(struct inode *inode,
* currently supporting (pre)allocate mode for extent-based
* files _only_
*/
- if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+ if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
return -EOPNOTSUPP;
/* preallocation to directories is currently not supported */
@@ -3489,6 +3622,11 @@ long ext4_fallocate(struct inode *inode,
*/
credits = ext4_chunk_trans_blocks(inode, max_blocks);
mutex_lock(&inode->i_mutex);
+ ret = inode_newsize_ok(inode, (len + offset));
+ if (ret) {
+ mutex_unlock(&inode->i_mutex);
+ return ret;
+ }
retry:
while (ret >= 0 && ret < max_blocks) {
block = block + ret;
@@ -3547,7 +3685,7 @@ retry:
* Returns 0 on success.
*/
int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset,
- loff_t len)
+ ssize_t len)
{
handle_t *handle;
ext4_lblk_t block;
@@ -3683,7 +3821,7 @@ static int ext4_xattr_fiemap(struct inod
int error = 0;
/* in-inode? */
- if (EXT4_I(inode)->i_state & EXT4_STATE_XATTR) {
+ if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
struct ext4_iloc iloc;
int offset; /* offset of xattr in inode */
@@ -3696,6 +3834,7 @@ static int ext4_xattr_fiemap(struct inod
physical += offset;
length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
flags |= FIEMAP_EXTENT_DATA_INLINE;
+ brelse(iloc.bh);
} else { /* external block */
physical = EXT4_I(inode)->i_file_acl << blockbits;
length = inode->i_sb->s_blocksize;
@@ -3711,11 +3850,10 @@ int ext4_fiemap(struct inode *inode, str
__u64 start, __u64 len)
{
ext4_lblk_t start_blk;
- ext4_lblk_t len_blks;
int error = 0;
/* fallback to generic here if not in extents fmt */
- if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+ if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
return generic_block_fiemap(inode, fieinfo, start, len,
ext4_get_block);
@@ -3725,8 +3863,14 @@ int ext4_fiemap(struct inode *inode, str
if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
error = ext4_xattr_fiemap(inode, fieinfo);
} else {
+ ext4_lblk_t len_blks;
+ __u64 last_blk;
+
start_blk = start >> inode->i_sb->s_blocksize_bits;
- len_blks = len >> inode->i_sb->s_blocksize_bits;
+ last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
+ if (last_blk >= EXT_MAX_BLOCK)
+ last_blk = EXT_MAX_BLOCK-1;
+ len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
/*
* Walk the extent tree gathering extent information.
projects / ziggy471-frankenstein-kernel.git / blobdiff