linux/fs/ext4/migrate.c
Tahsin Erdogan 1b917ed8ae ext4: do not set posix acls on xattr inodes
We don't need acls on xattr inodes because they are not directly
accessible from user mode.

Besides lockdep complains about recursive locking of xattr_sem as seen
below.

  =============================================
  [ INFO: possible recursive locking detected ]
  4.11.0-rc8+ #402 Not tainted
  ---------------------------------------------
  python/1894 is trying to acquire lock:
   (&ei->xattr_sem){++++..}, at: [<ffffffff804878a6>] ext4_xattr_get+0x66/0x270

  but task is already holding lock:
   (&ei->xattr_sem){++++..}, at: [<ffffffff80489500>] ext4_xattr_set_handle+0xa0/0x5d0

  other info that might help us debug this:
   Possible unsafe locking scenario:

         CPU0
         ----
    lock(&ei->xattr_sem);
    lock(&ei->xattr_sem);

   *** DEADLOCK ***

   May be due to missing lock nesting notation

  3 locks held by python/1894:
   #0:  (sb_writers#10){.+.+.+}, at: [<ffffffff803d829f>] mnt_want_write+0x1f/0x50
   #1:  (&sb->s_type->i_mutex_key#15){+.+...}, at: [<ffffffff803dda27>] vfs_setxattr+0x57/0xb0
   #2:  (&ei->xattr_sem){++++..}, at: [<ffffffff80489500>] ext4_xattr_set_handle+0xa0/0x5d0

  stack backtrace:
  CPU: 0 PID: 1894 Comm: python Not tainted 4.11.0-rc8+ #402
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
  Call Trace:
   dump_stack+0x67/0x99
   __lock_acquire+0x5f3/0x1830
   lock_acquire+0xb5/0x1d0
   down_read+0x2f/0x60
   ext4_xattr_get+0x66/0x270
   ext4_get_acl+0x43/0x1e0
   get_acl+0x72/0xf0
   posix_acl_create+0x5e/0x170
   ext4_init_acl+0x21/0xc0
   __ext4_new_inode+0xffd/0x16b0
   ext4_xattr_set_entry+0x5ea/0xb70
   ext4_xattr_block_set+0x1b5/0x970
   ext4_xattr_set_handle+0x351/0x5d0
   ext4_xattr_set+0x124/0x180
   ext4_xattr_user_set+0x34/0x40
   __vfs_setxattr+0x66/0x80
   __vfs_setxattr_noperm+0x69/0x1c0
   vfs_setxattr+0xa2/0xb0
   setxattr+0x129/0x160
   path_setxattr+0x87/0xb0
   SyS_setxattr+0xf/0x20
   entry_SYSCALL_64_fastpath+0x18/0xad

Signed-off-by: Tahsin Erdogan <tahsin@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
2017-06-21 21:21:39 -04:00

681 lines
17 KiB
C

/*
* Copyright IBM Corporation, 2007
* Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of version 2.1 of the GNU Lesser General Public License
* as published by the Free Software Foundation.
*
* This program is distributed in the hope that it would be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
*/
#include <linux/slab.h>
#include "ext4_jbd2.h"
#include "ext4_extents.h"
/*
* The contiguous blocks details which can be
* represented by a single extent
*/
struct migrate_struct {
ext4_lblk_t first_block, last_block, curr_block;
ext4_fsblk_t first_pblock, last_pblock;
};
static int finish_range(handle_t *handle, struct inode *inode,
struct migrate_struct *lb)
{
int retval = 0, needed;
struct ext4_extent newext;
struct ext4_ext_path *path;
if (lb->first_pblock == 0)
return 0;
/* Add the extent to temp inode*/
newext.ee_block = cpu_to_le32(lb->first_block);
newext.ee_len = cpu_to_le16(lb->last_block - lb->first_block + 1);
ext4_ext_store_pblock(&newext, lb->first_pblock);
/* Locking only for convinience since we are operating on temp inode */
down_write(&EXT4_I(inode)->i_data_sem);
path = ext4_find_extent(inode, lb->first_block, NULL, 0);
if (IS_ERR(path)) {
retval = PTR_ERR(path);
path = NULL;
goto err_out;
}
/*
* Calculate the credit needed to inserting this extent
* Since we are doing this in loop we may accumalate extra
* credit. But below we try to not accumalate too much
* of them by restarting the journal.
*/
needed = ext4_ext_calc_credits_for_single_extent(inode,
lb->last_block - lb->first_block + 1, path);
/*
* Make sure the credit we accumalated is not really high
*/
if (needed && ext4_handle_has_enough_credits(handle,
EXT4_RESERVE_TRANS_BLOCKS)) {
up_write((&EXT4_I(inode)->i_data_sem));
retval = ext4_journal_restart(handle, needed);
down_write((&EXT4_I(inode)->i_data_sem));
if (retval)
goto err_out;
} else if (needed) {
retval = ext4_journal_extend(handle, needed);
if (retval) {
/*
* IF not able to extend the journal restart the journal
*/
up_write((&EXT4_I(inode)->i_data_sem));
retval = ext4_journal_restart(handle, needed);
down_write((&EXT4_I(inode)->i_data_sem));
if (retval)
goto err_out;
}
}
retval = ext4_ext_insert_extent(handle, inode, &path, &newext, 0);
err_out:
up_write((&EXT4_I(inode)->i_data_sem));
ext4_ext_drop_refs(path);
kfree(path);
lb->first_pblock = 0;
return retval;
}
static int update_extent_range(handle_t *handle, struct inode *inode,
ext4_fsblk_t pblock, struct migrate_struct *lb)
{
int retval;
/*
* See if we can add on to the existing range (if it exists)
*/
if (lb->first_pblock &&
(lb->last_pblock+1 == pblock) &&
(lb->last_block+1 == lb->curr_block)) {
lb->last_pblock = pblock;
lb->last_block = lb->curr_block;
lb->curr_block++;
return 0;
}
/*
* Start a new range.
*/
retval = finish_range(handle, inode, lb);
lb->first_pblock = lb->last_pblock = pblock;
lb->first_block = lb->last_block = lb->curr_block;
lb->curr_block++;
return retval;
}
static int update_ind_extent_range(handle_t *handle, struct inode *inode,
ext4_fsblk_t pblock,
struct migrate_struct *lb)
{
struct buffer_head *bh;
__le32 *i_data;
int i, retval = 0;
unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
bh = sb_bread(inode->i_sb, pblock);
if (!bh)
return -EIO;
i_data = (__le32 *)bh->b_data;
for (i = 0; i < max_entries; i++) {
if (i_data[i]) {
retval = update_extent_range(handle, inode,
le32_to_cpu(i_data[i]), lb);
if (retval)
break;
} else {
lb->curr_block++;
}
}
put_bh(bh);
return retval;
}
static int update_dind_extent_range(handle_t *handle, struct inode *inode,
ext4_fsblk_t pblock,
struct migrate_struct *lb)
{
struct buffer_head *bh;
__le32 *i_data;
int i, retval = 0;
unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
bh = sb_bread(inode->i_sb, pblock);
if (!bh)
return -EIO;
i_data = (__le32 *)bh->b_data;
for (i = 0; i < max_entries; i++) {
if (i_data[i]) {
retval = update_ind_extent_range(handle, inode,
le32_to_cpu(i_data[i]), lb);
if (retval)
break;
} else {
/* Only update the file block number */
lb->curr_block += max_entries;
}
}
put_bh(bh);
return retval;
}
static int update_tind_extent_range(handle_t *handle, struct inode *inode,
ext4_fsblk_t pblock,
struct migrate_struct *lb)
{
struct buffer_head *bh;
__le32 *i_data;
int i, retval = 0;
unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
bh = sb_bread(inode->i_sb, pblock);
if (!bh)
return -EIO;
i_data = (__le32 *)bh->b_data;
for (i = 0; i < max_entries; i++) {
if (i_data[i]) {
retval = update_dind_extent_range(handle, inode,
le32_to_cpu(i_data[i]), lb);
if (retval)
break;
} else {
/* Only update the file block number */
lb->curr_block += max_entries * max_entries;
}
}
put_bh(bh);
return retval;
}
static int extend_credit_for_blkdel(handle_t *handle, struct inode *inode)
{
int retval = 0, needed;
if (ext4_handle_has_enough_credits(handle, EXT4_RESERVE_TRANS_BLOCKS+1))
return 0;
/*
* We are freeing a blocks. During this we touch
* superblock, group descriptor and block bitmap.
* So allocate a credit of 3. We may update
* quota (user and group).
*/
needed = 3 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
if (ext4_journal_extend(handle, needed) != 0)
retval = ext4_journal_restart(handle, needed);
return retval;
}
static int free_dind_blocks(handle_t *handle,
struct inode *inode, __le32 i_data)
{
int i;
__le32 *tmp_idata;
struct buffer_head *bh;
unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
bh = sb_bread(inode->i_sb, le32_to_cpu(i_data));
if (!bh)
return -EIO;
tmp_idata = (__le32 *)bh->b_data;
for (i = 0; i < max_entries; i++) {
if (tmp_idata[i]) {
extend_credit_for_blkdel(handle, inode);
ext4_free_blocks(handle, inode, NULL,
le32_to_cpu(tmp_idata[i]), 1,
EXT4_FREE_BLOCKS_METADATA |
EXT4_FREE_BLOCKS_FORGET);
}
}
put_bh(bh);
extend_credit_for_blkdel(handle, inode);
ext4_free_blocks(handle, inode, NULL, le32_to_cpu(i_data), 1,
EXT4_FREE_BLOCKS_METADATA |
EXT4_FREE_BLOCKS_FORGET);
return 0;
}
static int free_tind_blocks(handle_t *handle,
struct inode *inode, __le32 i_data)
{
int i, retval = 0;
__le32 *tmp_idata;
struct buffer_head *bh;
unsigned long max_entries = inode->i_sb->s_blocksize >> 2;
bh = sb_bread(inode->i_sb, le32_to_cpu(i_data));
if (!bh)
return -EIO;
tmp_idata = (__le32 *)bh->b_data;
for (i = 0; i < max_entries; i++) {
if (tmp_idata[i]) {
retval = free_dind_blocks(handle,
inode, tmp_idata[i]);
if (retval) {
put_bh(bh);
return retval;
}
}
}
put_bh(bh);
extend_credit_for_blkdel(handle, inode);
ext4_free_blocks(handle, inode, NULL, le32_to_cpu(i_data), 1,
EXT4_FREE_BLOCKS_METADATA |
EXT4_FREE_BLOCKS_FORGET);
return 0;
}
static int free_ind_block(handle_t *handle, struct inode *inode, __le32 *i_data)
{
int retval;
/* ei->i_data[EXT4_IND_BLOCK] */
if (i_data[0]) {
extend_credit_for_blkdel(handle, inode);
ext4_free_blocks(handle, inode, NULL,
le32_to_cpu(i_data[0]), 1,
EXT4_FREE_BLOCKS_METADATA |
EXT4_FREE_BLOCKS_FORGET);
}
/* ei->i_data[EXT4_DIND_BLOCK] */
if (i_data[1]) {
retval = free_dind_blocks(handle, inode, i_data[1]);
if (retval)
return retval;
}
/* ei->i_data[EXT4_TIND_BLOCK] */
if (i_data[2]) {
retval = free_tind_blocks(handle, inode, i_data[2]);
if (retval)
return retval;
}
return 0;
}
static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode,
struct inode *tmp_inode)
{
int retval;
__le32 i_data[3];
struct ext4_inode_info *ei = EXT4_I(inode);
struct ext4_inode_info *tmp_ei = EXT4_I(tmp_inode);
/*
* One credit accounted for writing the
* i_data field of the original inode
*/
retval = ext4_journal_extend(handle, 1);
if (retval) {
retval = ext4_journal_restart(handle, 1);
if (retval)
goto err_out;
}
i_data[0] = ei->i_data[EXT4_IND_BLOCK];
i_data[1] = ei->i_data[EXT4_DIND_BLOCK];
i_data[2] = ei->i_data[EXT4_TIND_BLOCK];
down_write(&EXT4_I(inode)->i_data_sem);
/*
* if EXT4_STATE_EXT_MIGRATE is cleared a block allocation
* happened after we started the migrate. We need to
* fail the migrate
*/
if (!ext4_test_inode_state(inode, EXT4_STATE_EXT_MIGRATE)) {
retval = -EAGAIN;
up_write(&EXT4_I(inode)->i_data_sem);
goto err_out;
} else
ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
/*
* We have the extent map build with the tmp inode.
* Now copy the i_data across
*/
ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS);
memcpy(ei->i_data, tmp_ei->i_data, sizeof(ei->i_data));
/*
* Update i_blocks with the new blocks that got
* allocated while adding extents for extent index
* blocks.
*
* While converting to extents we need not
* update the original inode i_blocks for extent blocks
* via quota APIs. The quota update happened via tmp_inode already.
*/
spin_lock(&inode->i_lock);
inode->i_blocks += tmp_inode->i_blocks;
spin_unlock(&inode->i_lock);
up_write(&EXT4_I(inode)->i_data_sem);
/*
* We mark the inode dirty after, because we decrement the
* i_blocks when freeing the indirect meta-data blocks
*/
retval = free_ind_block(handle, inode, i_data);
ext4_mark_inode_dirty(handle, inode);
err_out:
return retval;
}
static int free_ext_idx(handle_t *handle, struct inode *inode,
struct ext4_extent_idx *ix)
{
int i, retval = 0;
ext4_fsblk_t block;
struct buffer_head *bh;
struct ext4_extent_header *eh;
block = ext4_idx_pblock(ix);
bh = sb_bread(inode->i_sb, block);
if (!bh)
return -EIO;
eh = (struct ext4_extent_header *)bh->b_data;
if (eh->eh_depth != 0) {
ix = EXT_FIRST_INDEX(eh);
for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ix++) {
retval = free_ext_idx(handle, inode, ix);
if (retval)
break;
}
}
put_bh(bh);
extend_credit_for_blkdel(handle, inode);
ext4_free_blocks(handle, inode, NULL, block, 1,
EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
return retval;
}
/*
* Free the extent meta data blocks only
*/
static int free_ext_block(handle_t *handle, struct inode *inode)
{
int i, retval = 0;
struct ext4_inode_info *ei = EXT4_I(inode);
struct ext4_extent_header *eh = (struct ext4_extent_header *)ei->i_data;
struct ext4_extent_idx *ix;
if (eh->eh_depth == 0)
/*
* No extra blocks allocated for extent meta data
*/
return 0;
ix = EXT_FIRST_INDEX(eh);
for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ix++) {
retval = free_ext_idx(handle, inode, ix);
if (retval)
return retval;
}
return retval;
}
int ext4_ext_migrate(struct inode *inode)
{
handle_t *handle;
int retval = 0, i;
__le32 *i_data;
struct ext4_inode_info *ei;
struct inode *tmp_inode = NULL;
struct migrate_struct lb;
unsigned long max_entries;
__u32 goal;
uid_t owner[2];
/*
* If the filesystem does not support extents, or the inode
* already is extent-based, error out.
*/
if (!ext4_has_feature_extents(inode->i_sb) ||
(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
return -EINVAL;
if (S_ISLNK(inode->i_mode) && inode->i_blocks == 0)
/*
* don't migrate fast symlink
*/
return retval;
/*
* Worst case we can touch the allocation bitmaps, a bgd
* block, and a block to link in the orphan list. We do need
* need to worry about credits for modifying the quota inode.
*/
handle = ext4_journal_start(inode, EXT4_HT_MIGRATE,
4 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb));
if (IS_ERR(handle)) {
retval = PTR_ERR(handle);
return retval;
}
goal = (((inode->i_ino - 1) / EXT4_INODES_PER_GROUP(inode->i_sb)) *
EXT4_INODES_PER_GROUP(inode->i_sb)) + 1;
owner[0] = i_uid_read(inode);
owner[1] = i_gid_read(inode);
tmp_inode = ext4_new_inode(handle, d_inode(inode->i_sb->s_root),
S_IFREG, NULL, goal, owner, 0);
if (IS_ERR(tmp_inode)) {
retval = PTR_ERR(tmp_inode);
ext4_journal_stop(handle);
return retval;
}
i_size_write(tmp_inode, i_size_read(inode));
/*
* Set the i_nlink to zero so it will be deleted later
* when we drop inode reference.
*/
clear_nlink(tmp_inode);
ext4_ext_tree_init(handle, tmp_inode);
ext4_orphan_add(handle, tmp_inode);
ext4_journal_stop(handle);
/*
* start with one credit accounted for
* superblock modification.
*
* For the tmp_inode we already have committed the
* transaction that created the inode. Later as and
* when we add extents we extent the journal
*/
/*
* Even though we take i_mutex we can still cause block
* allocation via mmap write to holes. If we have allocated
* new blocks we fail migrate. New block allocation will
* clear EXT4_STATE_EXT_MIGRATE flag. The flag is updated
* with i_data_sem held to prevent racing with block
* allocation.
*/
down_read(&EXT4_I(inode)->i_data_sem);
ext4_set_inode_state(inode, EXT4_STATE_EXT_MIGRATE);
up_read((&EXT4_I(inode)->i_data_sem));
handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1);
if (IS_ERR(handle)) {
/*
* It is impossible to update on-disk structures without
* a handle, so just rollback in-core changes and live other
* work to orphan_list_cleanup()
*/
ext4_orphan_del(NULL, tmp_inode);
retval = PTR_ERR(handle);
goto out;
}
ei = EXT4_I(inode);
i_data = ei->i_data;
memset(&lb, 0, sizeof(lb));
/* 32 bit block address 4 bytes */
max_entries = inode->i_sb->s_blocksize >> 2;
for (i = 0; i < EXT4_NDIR_BLOCKS; i++) {
if (i_data[i]) {
retval = update_extent_range(handle, tmp_inode,
le32_to_cpu(i_data[i]), &lb);
if (retval)
goto err_out;
} else
lb.curr_block++;
}
if (i_data[EXT4_IND_BLOCK]) {
retval = update_ind_extent_range(handle, tmp_inode,
le32_to_cpu(i_data[EXT4_IND_BLOCK]), &lb);
if (retval)
goto err_out;
} else
lb.curr_block += max_entries;
if (i_data[EXT4_DIND_BLOCK]) {
retval = update_dind_extent_range(handle, tmp_inode,
le32_to_cpu(i_data[EXT4_DIND_BLOCK]), &lb);
if (retval)
goto err_out;
} else
lb.curr_block += max_entries * max_entries;
if (i_data[EXT4_TIND_BLOCK]) {
retval = update_tind_extent_range(handle, tmp_inode,
le32_to_cpu(i_data[EXT4_TIND_BLOCK]), &lb);
if (retval)
goto err_out;
}
/*
* Build the last extent
*/
retval = finish_range(handle, tmp_inode, &lb);
err_out:
if (retval)
/*
* Failure case delete the extent information with the
* tmp_inode
*/
free_ext_block(handle, tmp_inode);
else {
retval = ext4_ext_swap_inode_data(handle, inode, tmp_inode);
if (retval)
/*
* if we fail to swap inode data free the extent
* details of the tmp inode
*/
free_ext_block(handle, tmp_inode);
}
/* We mark the tmp_inode dirty via ext4_ext_tree_init. */
if (ext4_journal_extend(handle, 1) != 0)
ext4_journal_restart(handle, 1);
/*
* Mark the tmp_inode as of size zero
*/
i_size_write(tmp_inode, 0);
/*
* set the i_blocks count to zero
* so that the ext4_evict_inode() does the
* right job
*
* We don't need to take the i_lock because
* the inode is not visible to user space.
*/
tmp_inode->i_blocks = 0;
/* Reset the extent details */
ext4_ext_tree_init(handle, tmp_inode);
ext4_journal_stop(handle);
out:
unlock_new_inode(tmp_inode);
iput(tmp_inode);
return retval;
}
/*
* Migrate a simple extent-based inode to use the i_blocks[] array
*/
int ext4_ind_migrate(struct inode *inode)
{
struct ext4_extent_header *eh;
struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
struct ext4_inode_info *ei = EXT4_I(inode);
struct ext4_extent *ex;
unsigned int i, len;
ext4_lblk_t start, end;
ext4_fsblk_t blk;
handle_t *handle;
int ret;
if (!ext4_has_feature_extents(inode->i_sb) ||
(!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
return -EINVAL;
if (ext4_has_feature_bigalloc(inode->i_sb))
return -EOPNOTSUPP;
/*
* In order to get correct extent info, force all delayed allocation
* blocks to be allocated, otherwise delayed allocation blocks may not
* be reflected and bypass the checks on extent header.
*/
if (test_opt(inode->i_sb, DELALLOC))
ext4_alloc_da_blocks(inode);
handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1);
if (IS_ERR(handle))
return PTR_ERR(handle);
down_write(&EXT4_I(inode)->i_data_sem);
ret = ext4_ext_check_inode(inode);
if (ret)
goto errout;
eh = ext_inode_hdr(inode);
ex = EXT_FIRST_EXTENT(eh);
if (ext4_blocks_count(es) > EXT4_MAX_BLOCK_FILE_PHYS ||
eh->eh_depth != 0 || le16_to_cpu(eh->eh_entries) > 1) {
ret = -EOPNOTSUPP;
goto errout;
}
if (eh->eh_entries == 0)
blk = len = start = end = 0;
else {
len = le16_to_cpu(ex->ee_len);
blk = ext4_ext_pblock(ex);
start = le32_to_cpu(ex->ee_block);
end = start + len - 1;
if (end >= EXT4_NDIR_BLOCKS) {
ret = -EOPNOTSUPP;
goto errout;
}
}
ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS);
memset(ei->i_data, 0, sizeof(ei->i_data));
for (i = start; i <= end; i++)
ei->i_data[i] = cpu_to_le32(blk++);
ext4_mark_inode_dirty(handle, inode);
errout:
ext4_journal_stop(handle);
up_write(&EXT4_I(inode)->i_data_sem);
return ret;
}