#include <linux/buffer_head.h> // for block_sync_page
#include <linux/workqueue.h>
#include <linux/kthread.h>
-# include <linux/freezer.h>
+#include <linux/freezer.h>
+#include "compat.h"
#include "crc32c.h"
#include "ctree.h"
#include "disk-io.h"
* extents on the btree inode are pretty simple, there's one extent
* that covers the entire device
*/
-struct extent_map *btree_get_extent(struct inode *inode, struct page *page,
- size_t page_offset, u64 start, u64 len,
- int create)
+static struct extent_map *btree_get_extent(struct inode *inode,
+ struct page *page, size_t page_offset, u64 start, u64 len,
+ int create)
{
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
struct extent_map *em;
static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
int verify)
{
- char result[BTRFS_CRC32_SIZE];
+ u16 csum_size =
+ btrfs_super_csum_size(&root->fs_info->super_copy);
+ char *result = NULL;
unsigned long len;
unsigned long cur_len;
unsigned long offset = BTRFS_CSUM_SIZE;
unsigned long map_len;
int err;
u32 crc = ~(u32)0;
+ unsigned long inline_result;
len = buf->len - offset;
while(len > 0) {
offset += cur_len;
unmap_extent_buffer(buf, map_token, KM_USER0);
}
+ if (csum_size > sizeof(inline_result)) {
+ result = kzalloc(csum_size * sizeof(char), GFP_NOFS);
+ if (!result)
+ return 1;
+ } else {
+ result = (char *)&inline_result;
+ }
+
btrfs_csum_final(crc, result);
if (verify) {
/* FIXME, this is not good */
- if (memcmp_extent_buffer(buf, result, 0, BTRFS_CRC32_SIZE)) {
+ if (memcmp_extent_buffer(buf, result, 0, csum_size)) {
u32 val;
u32 found = 0;
- memcpy(&found, result, BTRFS_CRC32_SIZE);
+ memcpy(&found, result, csum_size);
- read_extent_buffer(buf, &val, 0, BTRFS_CRC32_SIZE);
+ read_extent_buffer(buf, &val, 0, csum_size);
printk("btrfs: %s checksum verify failed on %llu "
"wanted %X found %X level %d\n",
root->fs_info->sb->s_id,
buf->start, val, found, btrfs_header_level(buf));
+ if (result != (char *)&inline_result)
+ kfree(result);
return 1;
}
} else {
- write_extent_buffer(buf, result, 0, BTRFS_CRC32_SIZE);
+ write_extent_buffer(buf, result, 0, csum_size);
}
+ if (result != (char *)&inline_result)
+ kfree(result);
return 0;
}
* checksum a dirty tree block before IO. This has extra checks to make
* sure we only fill in the checksum field in the first page of a multi-page block
*/
-int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
+static int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
{
struct extent_io_tree *tree;
u64 start = (u64)page->index << PAGE_CACHE_SHIFT;
return 0;
}
-int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
+static int check_tree_block_fsid(struct btrfs_root *root,
+ struct extent_buffer *eb)
+{
+ struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
+ u8 fsid[BTRFS_UUID_SIZE];
+ int ret = 1;
+
+ read_extent_buffer(eb, fsid, (unsigned long)btrfs_header_fsid(eb),
+ BTRFS_FSID_SIZE);
+ while (fs_devices) {
+ if (!memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE)) {
+ ret = 0;
+ break;
+ }
+ fs_devices = fs_devices->seed;
+ }
+ return ret;
+}
+
+static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
struct extent_state *state)
{
struct extent_io_tree *tree;
ret = -EIO;
goto err;
}
- if (memcmp_extent_buffer(eb, root->fs_info->fsid,
- (unsigned long)btrfs_header_fsid(eb),
- BTRFS_FSID_SIZE)) {
+ if (check_tree_block_fsid(root, eb)) {
printk("bad fsid on block %Lu\n", eb->start);
ret = -EIO;
goto err;
end_io_wq->error = err;
end_io_wq->work.func = end_workqueue_fn;
end_io_wq->work.flags = 0;
- if (bio->bi_rw & (1 << BIO_RW))
+
+ if (bio->bi_rw & (1 << BIO_RW)) {
btrfs_queue_worker(&fs_info->endio_write_workers,
&end_io_wq->work);
- else
- btrfs_queue_worker(&fs_info->endio_workers, &end_io_wq->work);
+ } else {
+ if (end_io_wq->metadata)
+ btrfs_queue_worker(&fs_info->endio_meta_workers,
+ &end_io_wq->work);
+ else
+ btrfs_queue_worker(&fs_info->endio_workers,
+ &end_io_wq->work);
+ }
}
int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
extent_submit_bio_hook_t *submit_bio_done)
{
struct async_submit_bio *async;
- int limit = btrfs_async_submit_limit(fs_info);
async = kmalloc(sizeof(*async), GFP_NOFS);
if (!async)
async->work.flags = 0;
async->bio_flags = bio_flags;
- while(atomic_read(&fs_info->async_submit_draining) &&
- atomic_read(&fs_info->nr_async_submits)) {
- wait_event(fs_info->async_submit_wait,
- (atomic_read(&fs_info->nr_async_submits) == 0));
- }
-
atomic_inc(&fs_info->nr_async_submits);
btrfs_queue_worker(&fs_info->workers, &async->work);
-
+#if 0
+ int limit = btrfs_async_submit_limit(fs_info);
if (atomic_read(&fs_info->nr_async_submits) > limit) {
wait_event_timeout(fs_info->async_submit_wait,
(atomic_read(&fs_info->nr_async_submits) < limit),
(atomic_read(&fs_info->nr_async_bios) < limit),
HZ/10);
}
-
+#endif
while(atomic_read(&fs_info->async_submit_draining) &&
atomic_read(&fs_info->nr_async_submits)) {
wait_event(fs_info->async_submit_wait,
BUG_ON(ret);
return btrfs_map_bio(BTRFS_I(inode)->root, rw, bio,
- mirror_num, 1);
+ mirror_num, 0);
}
return btrfs_wq_submit_bio(BTRFS_I(inode)->root->fs_info,
inode, rw, bio, mirror_num, 0,
return extent_writepages(tree, mapping, btree_get_extent, wbc);
}
-int btree_readpage(struct file *file, struct page *page)
+static int btree_readpage(struct file *file, struct page *page)
{
struct extent_io_tree *tree;
tree = &BTRFS_I(page->mapping->host)->io_tree;
u64 objectid)
{
root->node = NULL;
- root->inode = NULL;
root->commit_root = NULL;
root->ref_tree = NULL;
root->sectorsize = sectorsize;
root->defrag_running = 0;
root->defrag_level = 0;
root->root_key.objectid = objectid;
+ root->anon_super.s_root = NULL;
+ root->anon_super.s_dev = 0;
+ INIT_LIST_HEAD(&root->anon_super.s_list);
+ INIT_LIST_HEAD(&root->anon_super.s_instances);
+ init_rwsem(&root->anon_super.s_umount);
+
return 0;
}
root = btrfs_read_fs_root_no_radix(fs_info->tree_root, location);
if (IS_ERR(root))
return root;
+
+ set_anon_super(&root->anon_super, NULL);
+
ret = radix_tree_insert(&fs_info->fs_roots_radix,
(unsigned long)root->root_key.objectid,
root);
kfree(root);
return ERR_PTR(ret);
}
- ret = btrfs_find_dead_roots(fs_info->tree_root,
- root->root_key.objectid, root);
- BUG_ON(ret);
-
+ if (!(fs_info->sb->s_flags & MS_RDONLY)) {
+ ret = btrfs_find_dead_roots(fs_info->tree_root,
+ root->root_key.objectid, root);
+ BUG_ON(ret);
+ btrfs_orphan_cleanup(root);
+ }
return root;
}
kfree(root);
return ERR_PTR(ret);
}
-
+#if 0
ret = btrfs_sysfs_add_root(root);
if (ret) {
free_extent_buffer(root->node);
kfree(root);
return ERR_PTR(ret);
}
+#endif
root->in_sysfs = 1;
return root;
}
struct list_head *cur;
struct btrfs_device *device;
struct backing_dev_info *bdi;
-
+#if 0
if ((bdi_bits & (1 << BDI_write_congested)) &&
btrfs_congested_async(info, 0))
return 1;
-
+#endif
list_for_each(cur, &info->fs_devices->devices) {
device = list_entry(cur, struct btrfs_device, dev_list);
if (!device->bdev)
info = (struct btrfs_fs_info *)bdi->unplug_io_data;
list_for_each(cur, &info->fs_devices->devices) {
device = list_entry(cur, struct btrfs_device, dev_list);
+ if (!device->bdev)
+ continue;
+
bdi = blk_get_backing_dev_info(device->bdev);
if (bdi->unplug_io_fn) {
bdi->unplug_io_fn(bdi, page);
}
}
-void btrfs_unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
+static void btrfs_unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
{
struct inode *inode;
struct extent_map_tree *em_tree;
u64 offset;
/* the generic O_DIRECT read code does this */
- if (!page) {
+ if (1 || !page) {
__unplug_io_fn(bdi, page);
return;
}
return;
inode = mapping->host;
+
+ /*
+ * don't do the expensive searching for a small number of
+ * devices
+ */
+ if (BTRFS_I(inode)->root->fs_info->fs_devices->open_devices <= 2) {
+ __unplug_io_fn(bdi, page);
+ return;
+ }
+
offset = page_offset(page);
em_tree = &BTRFS_I(inode)->extent_tree;
* blocksize <= pagesize, it is basically a noop
*/
if (end_io_wq->metadata && !bio_ready_for_csum(bio)) {
- btrfs_queue_worker(&fs_info->endio_workers,
+ btrfs_queue_worker(&fs_info->endio_meta_workers,
&end_io_wq->work);
return;
}
u32 blocksize;
u32 stripesize;
u64 generation;
+ u64 features;
+ struct btrfs_key location;
struct buffer_head *bh;
struct btrfs_root *extent_root = kzalloc(sizeof(struct btrfs_root),
GFP_NOFS);
+ struct btrfs_root *csum_root = kzalloc(sizeof(struct btrfs_root),
+ GFP_NOFS);
struct btrfs_root *tree_root = kzalloc(sizeof(struct btrfs_root),
GFP_NOFS);
struct btrfs_fs_info *fs_info = kzalloc(sizeof(*fs_info),
struct btrfs_super_block *disk_super;
if (!extent_root || !tree_root || !fs_info ||
- !chunk_root || !dev_root) {
+ !chunk_root || !dev_root || !csum_root) {
err = -ENOMEM;
goto fail;
}
init_completion(&fs_info->kobj_unregister);
fs_info->tree_root = tree_root;
fs_info->extent_root = extent_root;
+ fs_info->csum_root = csum_root;
fs_info->chunk_root = chunk_root;
fs_info->dev_root = dev_root;
fs_info->fs_devices = fs_devices;
fs_info, BTRFS_ROOT_TREE_OBJECTID);
- bh = __bread(fs_devices->latest_bdev,
- BTRFS_SUPER_INFO_OFFSET / 4096, 4096);
+ bh = btrfs_read_dev_super(fs_devices->latest_bdev);
if (!bh)
goto fail_iput;
disk_super = &fs_info->super_copy;
if (!btrfs_super_root(disk_super))
- goto fail_sb_buffer;
+ goto fail_iput;
- err = btrfs_parse_options(tree_root, options);
- if (err)
- goto fail_sb_buffer;
+ ret = btrfs_parse_options(tree_root, options);
+ if (ret) {
+ err = ret;
+ goto fail_iput;
+ }
+
+ features = btrfs_super_incompat_flags(disk_super) &
+ ~BTRFS_FEATURE_INCOMPAT_SUPP;
+ if (features) {
+ printk(KERN_ERR "BTRFS: couldn't mount because of "
+ "unsupported optional features (%Lx).\n",
+ features);
+ err = -EINVAL;
+ goto fail_iput;
+ }
+
+ features = btrfs_super_compat_ro_flags(disk_super) &
+ ~BTRFS_FEATURE_COMPAT_RO_SUPP;
+ if (!(sb->s_flags & MS_RDONLY) && features) {
+ printk(KERN_ERR "BTRFS: couldn't mount RDWR because of "
+ "unsupported option features (%Lx).\n",
+ features);
+ err = -EINVAL;
+ goto fail_iput;
+ }
/*
* we need to start all the end_io workers up front because the
btrfs_init_workers(&fs_info->fixup_workers, "fixup", 1);
btrfs_init_workers(&fs_info->endio_workers, "endio",
fs_info->thread_pool_size);
+ btrfs_init_workers(&fs_info->endio_meta_workers, "endio-meta",
+ fs_info->thread_pool_size);
btrfs_init_workers(&fs_info->endio_write_workers, "endio-write",
fs_info->thread_pool_size);
btrfs_start_workers(&fs_info->delalloc_workers, 1);
btrfs_start_workers(&fs_info->fixup_workers, 1);
btrfs_start_workers(&fs_info->endio_workers, fs_info->thread_pool_size);
+ btrfs_start_workers(&fs_info->endio_meta_workers,
+ fs_info->thread_pool_size);
btrfs_start_workers(&fs_info->endio_write_workers,
fs_info->thread_pool_size);
- err = -EINVAL;
- if (btrfs_super_num_devices(disk_super) > fs_devices->open_devices) {
- printk("Btrfs: wanted %llu devices, but found %llu\n",
- (unsigned long long)btrfs_super_num_devices(disk_super),
- (unsigned long long)fs_devices->open_devices);
- if (btrfs_test_opt(tree_root, DEGRADED))
- printk("continuing in degraded mode\n");
- else {
- goto fail_sb_buffer;
- }
- }
-
fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
4 * 1024 * 1024 / PAGE_CACHE_SIZE);
}
mutex_lock(&fs_info->chunk_mutex);
- ret = btrfs_read_sys_array(tree_root);
+ ret = btrfs_read_sys_array(tree_root, btrfs_super_bytenr(disk_super));
mutex_unlock(&fs_info->chunk_mutex);
if (ret) {
printk("btrfs: failed to read the system array on %s\n",
mutex_lock(&fs_info->chunk_mutex);
ret = btrfs_read_chunk_tree(chunk_root);
mutex_unlock(&fs_info->chunk_mutex);
- BUG_ON(ret);
+ if (ret) {
+ printk("btrfs: failed to read chunk tree on %s\n", sb->s_id);
+ goto fail_chunk_root;
+ }
btrfs_close_extra_devices(fs_devices);
btrfs_super_root(disk_super),
blocksize, generation);
if (!tree_root->node)
- goto fail_sb_buffer;
+ goto fail_chunk_root;
ret = find_and_setup_root(tree_root, fs_info,
if (ret)
goto fail_extent_root;
+ ret = find_and_setup_root(tree_root, fs_info,
+ BTRFS_CSUM_TREE_OBJECTID, csum_root);
+ if (ret)
+ goto fail_extent_root;
+
+ csum_root->track_dirty = 1;
+
btrfs_read_block_groups(extent_root);
- fs_info->generation = btrfs_super_generation(disk_super) + 1;
+ fs_info->generation = generation + 1;
+ fs_info->last_trans_committed = generation;
fs_info->data_alloc_profile = (u64)-1;
fs_info->metadata_alloc_profile = (u64)-1;
fs_info->system_alloc_profile = fs_info->metadata_alloc_profile;
fs_info->cleaner_kthread = kthread_run(cleaner_kthread, tree_root,
"btrfs-cleaner");
if (!fs_info->cleaner_kthread)
- goto fail_extent_root;
+ goto fail_csum_root;
fs_info->transaction_kthread = kthread_run(transaction_kthread,
tree_root,
goto fail_cleaner;
if (btrfs_super_log_root(disk_super) != 0) {
- u32 blocksize;
u64 bytenr = btrfs_super_log_root(disk_super);
+ if (fs_devices->rw_devices == 0) {
+ printk("Btrfs log replay required on RO media\n");
+ err = -EIO;
+ goto fail_trans_kthread;
+ }
blocksize =
btrfs_level_size(tree_root,
btrfs_super_log_root_level(disk_super));
generation + 1);
ret = btrfs_recover_log_trees(log_tree_root);
BUG_ON(ret);
+
+ if (sb->s_flags & MS_RDONLY) {
+ ret = btrfs_commit_super(tree_root);
+ BUG_ON(ret);
+ }
}
- fs_info->last_trans_committed = btrfs_super_generation(disk_super);
- ret = btrfs_cleanup_reloc_trees(tree_root);
- BUG_ON(ret);
+ if (!(sb->s_flags & MS_RDONLY)) {
+ ret = btrfs_cleanup_reloc_trees(tree_root);
+ BUG_ON(ret);
+ }
+
+ location.objectid = BTRFS_FS_TREE_OBJECTID;
+ location.type = BTRFS_ROOT_ITEM_KEY;
+ location.offset = (u64)-1;
+ fs_info->fs_root = btrfs_read_fs_root_no_name(fs_info, &location);
+ if (!fs_info->fs_root)
+ goto fail_trans_kthread;
return tree_root;
+fail_trans_kthread:
+ kthread_stop(fs_info->transaction_kthread);
fail_cleaner:
kthread_stop(fs_info->cleaner_kthread);
+
+ /*
+ * make sure we're done with the btree inode before we stop our
+ * kthreads
+ */
+ filemap_write_and_wait(fs_info->btree_inode->i_mapping);
+ invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
+
+fail_csum_root:
+ free_extent_buffer(csum_root->node);
fail_extent_root:
free_extent_buffer(extent_root->node);
fail_tree_root:
free_extent_buffer(tree_root->node);
+fail_chunk_root:
+ free_extent_buffer(chunk_root->node);
fail_sys_array:
+ free_extent_buffer(dev_root->node);
fail_sb_buffer:
btrfs_stop_workers(&fs_info->fixup_workers);
btrfs_stop_workers(&fs_info->delalloc_workers);
btrfs_stop_workers(&fs_info->workers);
btrfs_stop_workers(&fs_info->endio_workers);
+ btrfs_stop_workers(&fs_info->endio_meta_workers);
btrfs_stop_workers(&fs_info->endio_write_workers);
btrfs_stop_workers(&fs_info->submit_workers);
fail_iput:
+ invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
iput(fs_info->btree_inode);
fail:
btrfs_close_devices(fs_info->fs_devices);
kfree(fs_info);
kfree(chunk_root);
kfree(dev_root);
+ kfree(csum_root);
return ERR_PTR(err);
}
put_bh(bh);
}
-int write_all_supers(struct btrfs_root *root)
+struct buffer_head *btrfs_read_dev_super(struct block_device *bdev)
+{
+ struct buffer_head *bh;
+ struct buffer_head *latest = NULL;
+ struct btrfs_super_block *super;
+ int i;
+ u64 transid = 0;
+ u64 bytenr;
+
+ /* we would like to check all the supers, but that would make
+ * a btrfs mount succeed after a mkfs from a different FS.
+ * So, we need to add a special mount option to scan for
+ * later supers, using BTRFS_SUPER_MIRROR_MAX instead
+ */
+ for (i = 0; i < 1; i++) {
+ bytenr = btrfs_sb_offset(i);
+ if (bytenr + 4096 >= i_size_read(bdev->bd_inode))
+ break;
+ bh = __bread(bdev, bytenr / 4096, 4096);
+ if (!bh)
+ continue;
+
+ super = (struct btrfs_super_block *)bh->b_data;
+ if (btrfs_super_bytenr(super) != bytenr ||
+ strncmp((char *)(&super->magic), BTRFS_MAGIC,
+ sizeof(super->magic))) {
+ brelse(bh);
+ continue;
+ }
+
+ if (!latest || btrfs_super_generation(super) > transid) {
+ brelse(latest);
+ latest = bh;
+ transid = btrfs_super_generation(super);
+ } else {
+ brelse(bh);
+ }
+ }
+ return latest;
+}
+
+static int write_dev_supers(struct btrfs_device *device,
+ struct btrfs_super_block *sb,
+ int do_barriers, int wait, int max_mirrors)
+{
+ struct buffer_head *bh;
+ int i;
+ int ret;
+ int errors = 0;
+ u32 crc;
+ u64 bytenr;
+ int last_barrier = 0;
+
+ if (max_mirrors == 0)
+ max_mirrors = BTRFS_SUPER_MIRROR_MAX;
+
+ /* make sure only the last submit_bh does a barrier */
+ if (do_barriers) {
+ for (i = 0; i < max_mirrors; i++) {
+ bytenr = btrfs_sb_offset(i);
+ if (bytenr + BTRFS_SUPER_INFO_SIZE >=
+ device->total_bytes)
+ break;
+ last_barrier = i;
+ }
+ }
+
+ for (i = 0; i < max_mirrors; i++) {
+ bytenr = btrfs_sb_offset(i);
+ if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes)
+ break;
+
+ if (wait) {
+ bh = __find_get_block(device->bdev, bytenr / 4096,
+ BTRFS_SUPER_INFO_SIZE);
+ BUG_ON(!bh);
+ brelse(bh);
+ wait_on_buffer(bh);
+ if (buffer_uptodate(bh)) {
+ brelse(bh);
+ continue;
+ }
+ } else {
+ btrfs_set_super_bytenr(sb, bytenr);
+
+ crc = ~(u32)0;
+ crc = btrfs_csum_data(NULL, (char *)sb +
+ BTRFS_CSUM_SIZE, crc,
+ BTRFS_SUPER_INFO_SIZE -
+ BTRFS_CSUM_SIZE);
+ btrfs_csum_final(crc, sb->csum);
+
+ bh = __getblk(device->bdev, bytenr / 4096,
+ BTRFS_SUPER_INFO_SIZE);
+ memcpy(bh->b_data, sb, BTRFS_SUPER_INFO_SIZE);
+
+ set_buffer_uptodate(bh);
+ get_bh(bh);
+ lock_buffer(bh);
+ bh->b_end_io = btrfs_end_buffer_write_sync;
+ }
+
+ if (i == last_barrier && do_barriers && device->barriers) {
+ ret = submit_bh(WRITE_BARRIER, bh);
+ if (ret == -EOPNOTSUPP) {
+ printk("btrfs: disabling barriers on dev %s\n",
+ device->name);
+ set_buffer_uptodate(bh);
+ device->barriers = 0;
+ get_bh(bh);
+ lock_buffer(bh);
+ ret = submit_bh(WRITE, bh);
+ }
+ } else {
+ ret = submit_bh(WRITE, bh);
+ }
+
+ if (!ret && wait) {
+ wait_on_buffer(bh);
+ if (!buffer_uptodate(bh))
+ errors++;
+ } else if (ret) {
+ errors++;
+ }
+ if (wait)
+ brelse(bh);
+ }
+ return errors < i ? 0 : -1;
+}
+
+int write_all_supers(struct btrfs_root *root, int max_mirrors)
{
struct list_head *cur;
struct list_head *head = &root->fs_info->fs_devices->devices;
struct btrfs_device *dev;
struct btrfs_super_block *sb;
struct btrfs_dev_item *dev_item;
- struct buffer_head *bh;
int ret;
int do_barriers;
int max_errors;
int total_errors = 0;
- u32 crc;
u64 flags;
max_errors = btrfs_super_num_devices(&root->fs_info->super_copy) - 1;
total_errors++;
continue;
}
- if (!dev->in_fs_metadata)
+ if (!dev->in_fs_metadata || !dev->writeable)
continue;
+ btrfs_set_stack_device_generation(dev_item, 0);
btrfs_set_stack_device_type(dev_item, dev->type);
btrfs_set_stack_device_id(dev_item, dev->devid);
btrfs_set_stack_device_total_bytes(dev_item, dev->total_bytes);
btrfs_set_stack_device_io_width(dev_item, dev->io_width);
btrfs_set_stack_device_sector_size(dev_item, dev->sector_size);
memcpy(dev_item->uuid, dev->uuid, BTRFS_UUID_SIZE);
+ memcpy(dev_item->fsid, dev->fs_devices->fsid, BTRFS_UUID_SIZE);
+
flags = btrfs_super_flags(sb);
btrfs_set_super_flags(sb, flags | BTRFS_HEADER_FLAG_WRITTEN);
-
- crc = ~(u32)0;
- crc = btrfs_csum_data(root, (char *)sb + BTRFS_CSUM_SIZE, crc,
- BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE);
- btrfs_csum_final(crc, sb->csum);
-
- bh = __getblk(dev->bdev, BTRFS_SUPER_INFO_OFFSET / 4096,
- BTRFS_SUPER_INFO_SIZE);
-
- memcpy(bh->b_data, sb, BTRFS_SUPER_INFO_SIZE);
- dev->pending_io = bh;
-
- get_bh(bh);
- set_buffer_uptodate(bh);
- lock_buffer(bh);
- bh->b_end_io = btrfs_end_buffer_write_sync;
-
- if (do_barriers && dev->barriers) {
- ret = submit_bh(WRITE_BARRIER, bh);
- if (ret == -EOPNOTSUPP) {
- printk("btrfs: disabling barriers on dev %s\n",
- dev->name);
- set_buffer_uptodate(bh);
- dev->barriers = 0;
- get_bh(bh);
- lock_buffer(bh);
- ret = submit_bh(WRITE, bh);
- }
- } else {
- ret = submit_bh(WRITE, bh);
- }
+ ret = write_dev_supers(dev, sb, do_barriers, 0, max_mirrors);
if (ret)
total_errors++;
}
printk("btrfs: %d errors while writing supers\n", total_errors);
BUG();
}
- total_errors = 0;
+ total_errors = 0;
list_for_each(cur, head) {
dev = list_entry(cur, struct btrfs_device, dev_list);
if (!dev->bdev)
continue;
- if (!dev->in_fs_metadata)
+ if (!dev->in_fs_metadata || !dev->writeable)
continue;
- BUG_ON(!dev->pending_io);
- bh = dev->pending_io;
- wait_on_buffer(bh);
- if (!buffer_uptodate(dev->pending_io)) {
- if (do_barriers && dev->barriers) {
- printk("btrfs: disabling barriers on dev %s\n",
- dev->name);
- set_buffer_uptodate(bh);
- get_bh(bh);
- lock_buffer(bh);
- dev->barriers = 0;
- ret = submit_bh(WRITE, bh);
- BUG_ON(ret);
- wait_on_buffer(bh);
- if (!buffer_uptodate(bh))
- total_errors++;
- } else {
- total_errors++;
- }
-
- }
- dev->pending_io = NULL;
- brelse(bh);
+ ret = write_dev_supers(dev, sb, do_barriers, 1, max_mirrors);
+ if (ret)
+ total_errors++;
}
if (total_errors > max_errors) {
printk("btrfs: %d errors while writing supers\n", total_errors);
return 0;
}
-int write_ctree_super(struct btrfs_trans_handle *trans, struct btrfs_root
- *root)
+int write_ctree_super(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root, int max_mirrors)
{
int ret;
- ret = write_all_supers(root);
+ ret = write_all_supers(root, max_mirrors);
return ret;
}
{
radix_tree_delete(&fs_info->fs_roots_radix,
(unsigned long)root->root_key.objectid);
+ if (root->anon_super.s_dev) {
+ down_write(&root->anon_super.s_umount);
+ kill_anon_super(&root->anon_super);
+ }
+#if 0
if (root->in_sysfs)
btrfs_sysfs_del_root(root);
- if (root->inode)
- iput(root->inode);
+#endif
if (root->node)
free_extent_buffer(root->node);
if (root->commit_root)
return 0;
}
-int close_ctree(struct btrfs_root *root)
+int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info)
{
+ u64 root_objectid = 0;
+ struct btrfs_root *gang[8];
+ int i;
int ret;
- struct btrfs_trans_handle *trans;
- struct btrfs_fs_info *fs_info = root->fs_info;
- fs_info->closing = 1;
- smp_mb();
+ while (1) {
+ ret = radix_tree_gang_lookup(&fs_info->fs_roots_radix,
+ (void **)gang, root_objectid,
+ ARRAY_SIZE(gang));
+ if (!ret)
+ break;
+ for (i = 0; i < ret; i++) {
+ root_objectid = gang[i]->root_key.objectid;
+ ret = btrfs_find_dead_roots(fs_info->tree_root,
+ root_objectid, gang[i]);
+ BUG_ON(ret);
+ btrfs_orphan_cleanup(gang[i]);
+ }
+ root_objectid++;
+ }
+ return 0;
+}
- kthread_stop(root->fs_info->transaction_kthread);
- kthread_stop(root->fs_info->cleaner_kthread);
+int btrfs_commit_super(struct btrfs_root *root)
+{
+ struct btrfs_trans_handle *trans;
+ int ret;
+ mutex_lock(&root->fs_info->cleaner_mutex);
btrfs_clean_old_snapshots(root);
+ mutex_unlock(&root->fs_info->cleaner_mutex);
trans = btrfs_start_transaction(root, 1);
ret = btrfs_commit_transaction(trans, root);
- /* run commit again to drop the original snapshot */
+ BUG_ON(ret);
+ /* run commit again to drop the original snapshot */
trans = btrfs_start_transaction(root, 1);
btrfs_commit_transaction(trans, root);
ret = btrfs_write_and_wait_transaction(NULL, root);
BUG_ON(ret);
- write_ctree_super(NULL, root);
+ ret = write_ctree_super(NULL, root, 0);
+ return ret;
+}
+
+int close_ctree(struct btrfs_root *root)
+{
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ int ret;
+
+ fs_info->closing = 1;
+ smp_mb();
+
+ kthread_stop(root->fs_info->transaction_kthread);
+ kthread_stop(root->fs_info->cleaner_kthread);
+
+ if (!(fs_info->sb->s_flags & MS_RDONLY)) {
+ ret = btrfs_commit_super(root);
+ if (ret) {
+ printk("btrfs: commit super returns %d\n", ret);
+ }
+ }
if (fs_info->delalloc_bytes) {
printk("btrfs: at unmount delalloc count %Lu\n",
if (root->fs_info->dev_root->node);
free_extent_buffer(root->fs_info->dev_root->node);
+ if (root->fs_info->csum_root->node);
+ free_extent_buffer(root->fs_info->csum_root->node);
+
btrfs_free_block_groups(root->fs_info);
- fs_info->closing = 2;
- del_fs_roots(fs_info);
- filemap_write_and_wait(fs_info->btree_inode->i_mapping);
+ del_fs_roots(fs_info);
- truncate_inode_pages(fs_info->btree_inode->i_mapping, 0);
+ iput(fs_info->btree_inode);
btrfs_stop_workers(&fs_info->fixup_workers);
btrfs_stop_workers(&fs_info->delalloc_workers);
btrfs_stop_workers(&fs_info->workers);
btrfs_stop_workers(&fs_info->endio_workers);
+ btrfs_stop_workers(&fs_info->endio_meta_workers);
btrfs_stop_workers(&fs_info->endio_write_workers);
btrfs_stop_workers(&fs_info->submit_workers);
- iput(fs_info->btree_inode);
#if 0
while(!list_empty(&fs_info->hashers)) {
struct btrfs_hasher *hasher;
kfree(fs_info->tree_root);
kfree(fs_info->chunk_root);
kfree(fs_info->dev_root);
+ kfree(fs_info->csum_root);
return 0;
}
projects / linux-2.6-block.git / blobdiff