* [PATCH] ext4: don't use the orphan list when migrating an inode
@ 2022-01-06 5:05 Theodore Ts'o
2022-01-06 10:57 ` Jan Kara
2022-01-06 13:19 ` Lukas Czerner
0 siblings, 2 replies; 3+ messages in thread
From: Theodore Ts'o @ 2022-01-06 5:05 UTC (permalink / raw)
To: Ext4 Developers List; +Cc: lczerner, jack, Theodore Ts'o
We probably want to remove the indirect block to extents migration
feature after a deprecation window, but until then, let's fix a
potential data loss problem caused by the fact that we put the
tmp_inode on the orphan list. In the unlikely case where we crash and
do a journal recovery, the data blocks belonging to the inode inode
being migrated are also represented in the tmp_inode on the orphan
list --- and so its data blocks will get marked unallocated, and
available for reuse.
Instead, we stop putting the tmp_inode on the oprhan list. So in the
case where we crash while migrating the inode, we'll leak an inode,
which is not a disaster. It will be easily fixed the next time we run
fsck, and it's better than potentially having blocks getting claimed
by two different files, and losing data as a result.
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
fs/ext4/migrate.c | 19 ++++---------------
1 file changed, 4 insertions(+), 15 deletions(-)
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 36dfc88ce05b..ff8916e1d38e 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -437,12 +437,12 @@ int ext4_ext_migrate(struct inode *inode)
percpu_down_write(&sbi->s_writepages_rwsem);
/*
- * 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.
+ * Worst case we can touch the allocation bitmaps and a block
+ * group descriptor block. 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));
+ 3 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb));
if (IS_ERR(handle)) {
retval = PTR_ERR(handle);
@@ -463,10 +463,6 @@ int ext4_ext_migrate(struct inode *inode)
* Use the correct seed for checksum (i.e. the seed from 'inode'). This
* is so that the metadata blocks will have the correct checksum after
* the migration.
- *
- * Note however that, if a crash occurs during the migration process,
- * the recovery process is broken because the tmp_inode checksums will
- * be wrong and the orphans cleanup will fail.
*/
ei = EXT4_I(inode);
EXT4_I(tmp_inode)->i_csum_seed = ei->i_csum_seed;
@@ -478,7 +474,6 @@ int ext4_ext_migrate(struct inode *inode)
clear_nlink(tmp_inode);
ext4_ext_tree_init(handle, tmp_inode);
- ext4_orphan_add(handle, tmp_inode);
ext4_journal_stop(handle);
/*
@@ -503,12 +498,6 @@ int ext4_ext_migrate(struct inode *inode)
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_tmp_inode;
}
--
2.31.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] ext4: don't use the orphan list when migrating an inode
2022-01-06 5:05 [PATCH] ext4: don't use the orphan list when migrating an inode Theodore Ts'o
@ 2022-01-06 10:57 ` Jan Kara
2022-01-06 13:19 ` Lukas Czerner
1 sibling, 0 replies; 3+ messages in thread
From: Jan Kara @ 2022-01-06 10:57 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: Ext4 Developers List, lczerner, jack
On Thu 06-01-22 00:05:05, Theodore Ts'o wrote:
> We probably want to remove the indirect block to extents migration
> feature after a deprecation window, but until then, let's fix a
> potential data loss problem caused by the fact that we put the
> tmp_inode on the orphan list. In the unlikely case where we crash and
> do a journal recovery, the data blocks belonging to the inode inode
> being migrated are also represented in the tmp_inode on the orphan
> list --- and so its data blocks will get marked unallocated, and
> available for reuse.
>
> Instead, we stop putting the tmp_inode on the oprhan list. So in the
> case where we crash while migrating the inode, we'll leak an inode,
> which is not a disaster. It will be easily fixed the next time we run
> fsck, and it's better than potentially having blocks getting claimed
> by two different files, and losing data as a result.
>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/ext4/migrate.c | 19 ++++---------------
> 1 file changed, 4 insertions(+), 15 deletions(-)
>
> diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
> index 36dfc88ce05b..ff8916e1d38e 100644
> --- a/fs/ext4/migrate.c
> +++ b/fs/ext4/migrate.c
> @@ -437,12 +437,12 @@ int ext4_ext_migrate(struct inode *inode)
> percpu_down_write(&sbi->s_writepages_rwsem);
>
> /*
> - * 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.
> + * Worst case we can touch the allocation bitmaps and a block
> + * group descriptor block. 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));
> + 3 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb));
>
> if (IS_ERR(handle)) {
> retval = PTR_ERR(handle);
> @@ -463,10 +463,6 @@ int ext4_ext_migrate(struct inode *inode)
> * Use the correct seed for checksum (i.e. the seed from 'inode'). This
> * is so that the metadata blocks will have the correct checksum after
> * the migration.
> - *
> - * Note however that, if a crash occurs during the migration process,
> - * the recovery process is broken because the tmp_inode checksums will
> - * be wrong and the orphans cleanup will fail.
> */
> ei = EXT4_I(inode);
> EXT4_I(tmp_inode)->i_csum_seed = ei->i_csum_seed;
> @@ -478,7 +474,6 @@ int ext4_ext_migrate(struct inode *inode)
> clear_nlink(tmp_inode);
>
> ext4_ext_tree_init(handle, tmp_inode);
> - ext4_orphan_add(handle, tmp_inode);
> ext4_journal_stop(handle);
>
> /*
> @@ -503,12 +498,6 @@ int ext4_ext_migrate(struct inode *inode)
>
> 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_tmp_inode;
> }
> --
> 2.31.0
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] ext4: don't use the orphan list when migrating an inode
2022-01-06 5:05 [PATCH] ext4: don't use the orphan list when migrating an inode Theodore Ts'o
2022-01-06 10:57 ` Jan Kara
@ 2022-01-06 13:19 ` Lukas Czerner
1 sibling, 0 replies; 3+ messages in thread
From: Lukas Czerner @ 2022-01-06 13:19 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: Ext4 Developers List, jack
On Thu, Jan 06, 2022 at 12:05:05AM -0500, Theodore Ts'o wrote:
> We probably want to remove the indirect block to extents migration
> feature after a deprecation window, but until then, let's fix a
> potential data loss problem caused by the fact that we put the
> tmp_inode on the orphan list. In the unlikely case where we crash and
> do a journal recovery, the data blocks belonging to the inode inode
> being migrated are also represented in the tmp_inode on the orphan
> list --- and so its data blocks will get marked unallocated, and
> available for reuse.
>
> Instead, we stop putting the tmp_inode on the oprhan list. So in the
> case where we crash while migrating the inode, we'll leak an inode,
> which is not a disaster. It will be easily fixed the next time we run
> fsck, and it's better than potentially having blocks getting claimed
> by two different files, and losing data as a result.
Looks good to me. Thanks!
Reviewed-by: Lukas Czerner <lczerner@redhat.com>
>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>
> ---
> fs/ext4/migrate.c | 19 ++++---------------
> 1 file changed, 4 insertions(+), 15 deletions(-)
>
> diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
> index 36dfc88ce05b..ff8916e1d38e 100644
> --- a/fs/ext4/migrate.c
> +++ b/fs/ext4/migrate.c
> @@ -437,12 +437,12 @@ int ext4_ext_migrate(struct inode *inode)
> percpu_down_write(&sbi->s_writepages_rwsem);
>
> /*
> - * 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.
> + * Worst case we can touch the allocation bitmaps and a block
> + * group descriptor block. 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));
> + 3 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb));
>
> if (IS_ERR(handle)) {
> retval = PTR_ERR(handle);
> @@ -463,10 +463,6 @@ int ext4_ext_migrate(struct inode *inode)
> * Use the correct seed for checksum (i.e. the seed from 'inode'). This
> * is so that the metadata blocks will have the correct checksum after
> * the migration.
> - *
> - * Note however that, if a crash occurs during the migration process,
> - * the recovery process is broken because the tmp_inode checksums will
> - * be wrong and the orphans cleanup will fail.
> */
> ei = EXT4_I(inode);
> EXT4_I(tmp_inode)->i_csum_seed = ei->i_csum_seed;
> @@ -478,7 +474,6 @@ int ext4_ext_migrate(struct inode *inode)
> clear_nlink(tmp_inode);
>
> ext4_ext_tree_init(handle, tmp_inode);
> - ext4_orphan_add(handle, tmp_inode);
> ext4_journal_stop(handle);
>
> /*
> @@ -503,12 +498,6 @@ int ext4_ext_migrate(struct inode *inode)
>
> 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_tmp_inode;
> }
> --
> 2.31.0
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-01-06 15:45 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-01-06 5:05 [PATCH] ext4: don't use the orphan list when migrating an inode Theodore Ts'o
2022-01-06 10:57 ` Jan Kara
2022-01-06 13:19 ` Lukas Czerner
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.