Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-15 00:21:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

btrfs: ensure relocation never runs while we have send operations running

Browse Source 
Relocation and send do not play well together because while send is
running a block group can be relocated, a transaction committed and
the respective disk extents get re-allocated and written to or discarded
while send is about to do something with the extents.

This was explained in commit 9e967495e0 ("Btrfs: prevent send failures
and crashes due to concurrent relocation"), which prevented balance and
send from running in parallel but it did not address one remaining case
where chunk relocation can happen: shrinking a device (and device deletion
which shrinks a device's size to 0 before deleting the device).

We also have now one more case where relocation is triggered: on zoned
filesystems partially used block groups get relocated by a background
thread, introduced in commit 18bb8bbf13 ("btrfs: zoned: automatically
reclaim zones").

So make sure that instead of preventing balance from running when there
are ongoing send operations, we prevent relocation from happening.
This uses the infrastructure recently added by a patch that has the
subject: "btrfs: add cancellable chunk relocation support".

Also it adds a spinlock used exclusively for the exclusivity between
send and relocation, as before fs_info->balance_mutex was used, which
would make an attempt to run send to block waiting for balance to
finish, which can take a lot of time on large filesystems.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit is contained in:
Filipe Manana 2021-06-21 11:10:38 +01:00 committed by David Sterba
parent cbeaae4f6f
commit 1cea5cf0e6
6 changed files with 32 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
fs/btrfs/block-group.c
View File

@ -1491,7 +1491,7 @@ void btrfs_reclaim_bgs_work(struct work_struct *work)
container_of(work, struct btrfs_fs_info, reclaim_bgs_work);
struct btrfs_block_group *bg;
struct btrfs_space_info *space_info;
int ret;
LIST_HEAD(again_list);
if (!test_bit(BTRFS_FS_OPEN, &fs_info->flags))
return;
@ -1502,6 +1502,8 @@ void btrfs_reclaim_bgs_work(struct work_struct *work)
mutex_lock(&fs_info->reclaim_bgs_lock);
spin_lock(&fs_info->unused_bgs_lock);
while (!list_empty(&fs_info->reclaim_bgs)) {
int ret = 0;
bg = list_first_entry(&fs_info->reclaim_bgs,
struct btrfs_block_group,
bg_list);
@ -1547,9 +1549,13 @@ void btrfs_reclaim_bgs_work(struct work_struct *work)
bg->start);
next:
btrfs_put_block_group(bg);
spin_lock(&fs_info->unused_bgs_lock);
if (ret == -EAGAIN && list_empty(&bg->bg_list))
list_add_tail(&bg->bg_list, &again_list);
else
btrfs_put_block_group(bg);
}
list_splice_tail(&again_list, &fs_info->reclaim_bgs);
spin_unlock(&fs_info->unused_bgs_lock);
mutex_unlock(&fs_info->reclaim_bgs_lock);
btrfs_exclop_finish(fs_info);

5
fs/btrfs/ctree.h
View File

@ -561,13 +561,13 @@ enum {
/*
* Indicate that balance has been set up from the ioctl and is in the
* main phase. The fs_info::balance_ctl is initialized.
* Set and cleared while holding fs_info::balance_mutex.
*/
BTRFS_FS_BALANCE_RUNNING,
/*
* Indicate that relocation of a chunk has started, it's set per chunk
* and is toggled between chunks.
* Set, tested and cleared while holding fs_info::send_reloc_lock.
*/
BTRFS_FS_RELOC_RUNNING,
@ -995,9 +995,10 @@ struct btrfs_fs_info {
struct crypto_shash *csum_shash;
spinlock_t send_reloc_lock;
/*
* Number of send operations in progress.
* Updated while holding fs_info::balance_mutex.
* Updated while holding fs_info::send_reloc_lock.
*/
int send_in_progress;

1
fs/btrfs/disk-io.c
View File

@ -2999,6 +2999,7 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
spin_lock_init(&fs_info->swapfile_pins_lock);
fs_info->swapfile_pins = RB_ROOT;
spin_lock_init(&fs_info->send_reloc_lock);
fs_info->send_in_progress = 0;
fs_info->bg_reclaim_threshold = BTRFS_DEFAULT_RECLAIM_THRESH;

13
fs/btrfs/relocation.c
View File

@ -3789,14 +3789,25 @@ out:
* 0 success
* -EINPROGRESS operation is already in progress, that's probably a bug
* -ECANCELED cancellation request was set before the operation started
* -EAGAIN can not start because there are ongoing send operations
*/
static int reloc_chunk_start(struct btrfs_fs_info *fs_info)
{
spin_lock(&fs_info->send_reloc_lock);
if (fs_info->send_in_progress) {
btrfs_warn_rl(fs_info,
"cannot run relocation while send operations are in progress (%d in progress)",
fs_info->send_in_progress);
spin_unlock(&fs_info->send_reloc_lock);
return -EAGAIN;
}
if (test_and_set_bit(BTRFS_FS_RELOC_RUNNING, &fs_info->flags)) {
/* This should not happen */
spin_unlock(&fs_info->send_reloc_lock);
btrfs_err(fs_info, "reloc already running, cannot start");
return -EINPROGRESS;
}
spin_unlock(&fs_info->send_reloc_lock);
if (atomic_read(&fs_info->reloc_cancel_req) > 0) {
btrfs_info(fs_info, "chunk relocation canceled on start");
@ -3818,7 +3829,9 @@ static void reloc_chunk_end(struct btrfs_fs_info *fs_info)
/* Requested after start, clear bit first so any waiters can continue */
if (atomic_read(&fs_info->reloc_cancel_req) > 0)
btrfs_info(fs_info, "chunk relocation canceled during operation");
spin_lock(&fs_info->send_reloc_lock);
clear_and_wake_up_bit(BTRFS_FS_RELOC_RUNNING, &fs_info->flags);
spin_unlock(&fs_info->send_reloc_lock);
atomic_set(&fs_info->reloc_cancel_req, 0);
}

14
fs/btrfs/send.c
View File

@ -7416,23 +7416,23 @@ long btrfs_ioctl_send(struct file *mnt_file, struct btrfs_ioctl_send_args *arg)
if (ret)
goto out;
mutex_lock(&fs_info->balance_mutex);
if (test_bit(BTRFS_FS_BALANCE_RUNNING, &fs_info->flags)) {
mutex_unlock(&fs_info->balance_mutex);
spin_lock(&fs_info->send_reloc_lock);
if (test_bit(BTRFS_FS_RELOC_RUNNING, &fs_info->flags)) {
spin_unlock(&fs_info->send_reloc_lock);
btrfs_warn_rl(fs_info,
"cannot run send because a balance operation is in progress");
"cannot run send because a relocation operation is in progress");
ret = -EAGAIN;
goto out;
}
fs_info->send_in_progress++;
mutex_unlock(&fs_info->balance_mutex);
spin_unlock(&fs_info->send_reloc_lock);
current->journal_info = BTRFS_SEND_TRANS_STUB;
ret = send_subvol(sctx);
current->journal_info = NULL;
mutex_lock(&fs_info->balance_mutex);
spin_lock(&fs_info->send_reloc_lock);
fs_info->send_in_progress--;
mutex_unlock(&fs_info->balance_mutex);
spin_unlock(&fs_info->send_reloc_lock);
if (ret < 0)
goto out;

8
fs/btrfs/volumes.c
View File

@ -4217,14 +4217,6 @@ int btrfs_balance(struct btrfs_fs_info *fs_info,
btrfs_bg_type_to_raid_name(data_target));
}
if (fs_info->send_in_progress) {
btrfs_warn_rl(fs_info,
"cannot run balance while send operations are in progress (%d in progress)",
fs_info->send_in_progress);
ret = -EAGAIN;
goto out;
}
ret = insert_balance_item(fs_info, bctl);
if (ret && ret != -EEXIST)
goto out;