Btrfs: make filesystem read-only when submitting barrier fails
So far the return code of barrier_all_devices() is ignored, which
means that errors are ignored. The result can be a corrupt
filesystem which is not consistent.
This commit adds code to evaluate the return code of
barrier_all_devices(). The normal btrfs_error() mechanism is used to
switch the filesystem into read-only mode when errors are detected.
In order to decide whether barrier_all_devices() should return
error or success, the number of disks that are allowed to fail the
barrier submission is calculated. This calculation accounts for the
worst RAID level of metadata, system and data. If single, dup or
RAID0 is in use, a single disk error is already considered to be
fatal. Otherwise a single disk error is tolerated.
The calculation of the number of disks that are tolerated to fail
the barrier operation is performed when the filesystem gets mounted,
when a balance operation is started and finished, and when devices
are added or removed.
Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 50dcd0f..1630be8 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1468,6 +1468,8 @@
/* next backup root to be overwritten */
int backup_root_index;
+
+ int num_tolerated_disk_barrier_failures;
};
/*
@@ -3361,6 +3363,9 @@
int btrfs_defrag_file(struct inode *inode, struct file *file,
struct btrfs_ioctl_defrag_range_args *range,
u64 newer_than, unsigned long max_pages);
+void btrfs_get_block_group_info(struct list_head *groups_list,
+ struct btrfs_ioctl_space_info *space);
+
/* file.c */
int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
struct inode *inode);
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index c699955..8355236 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2505,6 +2505,8 @@
printk(KERN_ERR "Failed to read block groups: %d\n", ret);
goto fail_block_groups;
}
+ fs_info->num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root,
"btrfs-cleaner");
@@ -2888,12 +2890,10 @@
printk_in_rcu("btrfs: disabling barriers on dev %s\n",
rcu_str_deref(device->name));
device->nobarriers = 1;
- }
- if (!bio_flagged(bio, BIO_UPTODATE)) {
+ } else if (!bio_flagged(bio, BIO_UPTODATE)) {
ret = -EIO;
- if (!bio_flagged(bio, BIO_EOPNOTSUPP))
- btrfs_dev_stat_inc_and_print(device,
- BTRFS_DEV_STAT_FLUSH_ERRS);
+ btrfs_dev_stat_inc_and_print(device,
+ BTRFS_DEV_STAT_FLUSH_ERRS);
}
/* drop the reference from the wait == 0 run */
@@ -2932,14 +2932,15 @@
{
struct list_head *head;
struct btrfs_device *dev;
- int errors = 0;
+ int errors_send = 0;
+ int errors_wait = 0;
int ret;
/* send down all the barriers */
head = &info->fs_devices->devices;
list_for_each_entry_rcu(dev, head, dev_list) {
if (!dev->bdev) {
- errors++;
+ errors_send++;
continue;
}
if (!dev->in_fs_metadata || !dev->writeable)
@@ -2947,13 +2948,13 @@
ret = write_dev_flush(dev, 0);
if (ret)
- errors++;
+ errors_send++;
}
/* wait for all the barriers */
list_for_each_entry_rcu(dev, head, dev_list) {
if (!dev->bdev) {
- errors++;
+ errors_wait++;
continue;
}
if (!dev->in_fs_metadata || !dev->writeable)
@@ -2961,13 +2962,87 @@
ret = write_dev_flush(dev, 1);
if (ret)
- errors++;
+ errors_wait++;
}
- if (errors)
+ if (errors_send > info->num_tolerated_disk_barrier_failures ||
+ errors_wait > info->num_tolerated_disk_barrier_failures)
return -EIO;
return 0;
}
+int btrfs_calc_num_tolerated_disk_barrier_failures(
+ struct btrfs_fs_info *fs_info)
+{
+ struct btrfs_ioctl_space_info space;
+ struct btrfs_space_info *sinfo;
+ u64 types[] = {BTRFS_BLOCK_GROUP_DATA,
+ BTRFS_BLOCK_GROUP_SYSTEM,
+ BTRFS_BLOCK_GROUP_METADATA,
+ BTRFS_BLOCK_GROUP_DATA | BTRFS_BLOCK_GROUP_METADATA};
+ int num_types = 4;
+ int i;
+ int c;
+ int num_tolerated_disk_barrier_failures =
+ (int)fs_info->fs_devices->num_devices;
+
+ for (i = 0; i < num_types; i++) {
+ struct btrfs_space_info *tmp;
+
+ sinfo = NULL;
+ rcu_read_lock();
+ list_for_each_entry_rcu(tmp, &fs_info->space_info, list) {
+ if (tmp->flags == types[i]) {
+ sinfo = tmp;
+ break;
+ }
+ }
+ rcu_read_unlock();
+
+ if (!sinfo)
+ continue;
+
+ down_read(&sinfo->groups_sem);
+ for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
+ if (!list_empty(&sinfo->block_groups[c])) {
+ u64 flags;
+
+ btrfs_get_block_group_info(
+ &sinfo->block_groups[c], &space);
+ if (space.total_bytes == 0 ||
+ space.used_bytes == 0)
+ continue;
+ flags = space.flags;
+ /*
+ * return
+ * 0: if dup, single or RAID0 is configured for
+ * any of metadata, system or data, else
+ * 1: if RAID5 is configured, or if RAID1 or
+ * RAID10 is configured and only two mirrors
+ * are used, else
+ * 2: if RAID6 is configured, else
+ * num_mirrors - 1: if RAID1 or RAID10 is
+ * configured and more than
+ * 2 mirrors are used.
+ */
+ if (num_tolerated_disk_barrier_failures > 0 &&
+ ((flags & (BTRFS_BLOCK_GROUP_DUP |
+ BTRFS_BLOCK_GROUP_RAID0)) ||
+ ((flags & BTRFS_BLOCK_GROUP_PROFILE_MASK)
+ == 0)))
+ num_tolerated_disk_barrier_failures = 0;
+ else if (num_tolerated_disk_barrier_failures > 1
+ &&
+ (flags & (BTRFS_BLOCK_GROUP_RAID1 |
+ BTRFS_BLOCK_GROUP_RAID10)))
+ num_tolerated_disk_barrier_failures = 1;
+ }
+ }
+ up_read(&sinfo->groups_sem);
+ }
+
+ return num_tolerated_disk_barrier_failures;
+}
+
int write_all_supers(struct btrfs_root *root, int max_mirrors)
{
struct list_head *head;
@@ -2990,8 +3065,16 @@
mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
head = &root->fs_info->fs_devices->devices;
- if (do_barriers)
- barrier_all_devices(root->fs_info);
+ if (do_barriers) {
+ ret = barrier_all_devices(root->fs_info);
+ if (ret) {
+ mutex_unlock(
+ &root->fs_info->fs_devices->device_list_mutex);
+ btrfs_error(root->fs_info, ret,
+ "errors while submitting device barriers.");
+ return ret;
+ }
+ }
list_for_each_entry_rcu(dev, head, dev_list) {
if (!dev->bdev) {
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index c5b00a7..2025a91 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -95,6 +95,8 @@
u64 objectid);
int btree_lock_page_hook(struct page *page, void *data,
void (*flush_fn)(void *));
+int btrfs_calc_num_tolerated_disk_barrier_failures(
+ struct btrfs_fs_info *fs_info);
#ifdef CONFIG_DEBUG_LOCK_ALLOC
void btrfs_init_lockdep(void);
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index d6836af6..f5a2e6c 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -2875,8 +2875,8 @@
return 0;
}
-static void get_block_group_info(struct list_head *groups_list,
- struct btrfs_ioctl_space_info *space)
+void btrfs_get_block_group_info(struct list_head *groups_list,
+ struct btrfs_ioctl_space_info *space)
{
struct btrfs_block_group_cache *block_group;
@@ -2984,8 +2984,8 @@
down_read(&info->groups_sem);
for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
if (!list_empty(&info->block_groups[c])) {
- get_block_group_info(&info->block_groups[c],
- &space);
+ btrfs_get_block_group_info(
+ &info->block_groups[c], &space);
memcpy(dest, &space, sizeof(space));
dest++;
space_args.total_spaces++;
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 47911fd..67eab2d 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2425,9 +2425,12 @@
* in and cause problems either.
*/
btrfs_scrub_pause_super(root);
- write_ctree_super(trans, root->fs_info->tree_root, 1);
+ ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
btrfs_scrub_continue_super(root);
- ret = 0;
+ if (ret) {
+ btrfs_abort_transaction(trans, root, ret);
+ goto out_wake_log_root;
+ }
mutex_lock(&root->log_mutex);
if (root->last_log_commit < log_transid)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index dfe5e3a..029b903 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1475,6 +1475,9 @@
free_fs_devices(cur_devices);
}
+ root->fs_info->num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(root->fs_info);
+
/*
* at this point, the device is zero sized. We want to
* remove it from the devices list and zero out the old super
@@ -1799,6 +1802,8 @@
btrfs_clear_space_info_full(root->fs_info);
unlock_chunks(root);
+ root->fs_info->num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(root->fs_info);
ret = btrfs_commit_transaction(trans, root);
if (seeding_dev) {
@@ -2809,6 +2814,26 @@
}
}
+ if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
+ int num_tolerated_disk_barrier_failures;
+ u64 target = bctl->sys.target;
+
+ num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
+ if (num_tolerated_disk_barrier_failures > 0 &&
+ (target &
+ (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID0 |
+ BTRFS_AVAIL_ALLOC_BIT_SINGLE)))
+ num_tolerated_disk_barrier_failures = 0;
+ else if (num_tolerated_disk_barrier_failures > 1 &&
+ (target &
+ (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)))
+ num_tolerated_disk_barrier_failures = 1;
+
+ fs_info->num_tolerated_disk_barrier_failures =
+ num_tolerated_disk_barrier_failures;
+ }
+
ret = insert_balance_item(fs_info->tree_root, bctl);
if (ret && ret != -EEXIST)
goto out;
@@ -2841,6 +2866,11 @@
__cancel_balance(fs_info);
}
+ if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
+ fs_info->num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
+ }
+
wake_up(&fs_info->balance_wait_q);
return ret;