#include "print-tree.h"
#include "volumes.h"
#include "utils.h"
+#include "kernel-lib/raid56.h"
struct stripe {
struct btrfs_device *dev;
u8 *uuid)
{
struct btrfs_device *dev;
- struct list_head *cur;
- list_for_each(cur, head) {
- dev = list_entry(cur, struct btrfs_device, dev_list);
+ list_for_each_entry(dev, head, dev_list) {
if (dev->devid == devid &&
!memcmp(dev->uuid, uuid, BTRFS_UUID_SIZE)) {
return dev;
static struct btrfs_fs_devices *find_fsid(u8 *fsid)
{
- struct list_head *cur;
struct btrfs_fs_devices *fs_devices;
- list_for_each(cur, &fs_uuids) {
- fs_devices = list_entry(cur, struct btrfs_fs_devices, list);
+ list_for_each_entry(fs_devices, &fs_uuids, list) {
if (memcmp(fsid, fs_devices->fsid, BTRFS_FSID_SIZE) == 0)
return fs_devices;
}
list_add(&device->dev_list, &fs_devices->devices);
device->fs_devices = fs_devices;
} else if (!device->name || strcmp(device->name, path)) {
- char *name = strdup(path);
+ char *name;
+
+ /*
+ * The existing device has newer generation, so this one could
+ * be a stale one, don't add it.
+ */
+ if (found_transid < device->generation) {
+ warning(
+ "adding device %s gen %llu but found an existing device %s gen %llu",
+ path, found_transid, device->name,
+ device->generation);
+ return -EEXIST;
+ }
+
+ name = strdup(path);
if (!name)
return -ENOMEM;
kfree(device->name);
{
struct btrfs_fs_devices *seed_devices;
struct btrfs_device *device;
+ int ret = 0;
again:
+ if (!fs_devices)
+ return 0;
while (!list_empty(&fs_devices->devices)) {
device = list_entry(fs_devices->devices.next,
struct btrfs_device, dev_list);
if (device->fd != -1) {
- fsync(device->fd);
+ if (fsync(device->fd) == -1) {
+ warning("fsync on device %llu failed: %s",
+ device->devid, strerror(errno));
+ ret = -errno;
+ }
if (posix_fadvise(device->fd, 0, 0, POSIX_FADV_DONTNEED))
fprintf(stderr, "Warning, could not drop caches\n");
close(device->fd);
free(fs_devices);
}
- return 0;
+ return ret;
+}
+
+void btrfs_close_all_devices(void)
+{
+ struct btrfs_fs_devices *fs_devices;
+
+ while (!list_empty(&fs_uuids)) {
+ fs_devices = list_entry(fs_uuids.next, struct btrfs_fs_devices,
+ list);
+ btrfs_close_devices(fs_devices);
+ }
}
int btrfs_open_devices(struct btrfs_fs_devices *fs_devices, int flags)
{
int fd;
- struct list_head *head = &fs_devices->devices;
- struct list_head *cur;
struct btrfs_device *device;
int ret;
- list_for_each(cur, head) {
- device = list_entry(cur, struct btrfs_device, dev_list);
+ list_for_each_entry(device, &fs_devices->devices, dev_list) {
if (!device->name) {
printk("no name for device %llu, skip it now\n", device->devid);
continue;
fd = open(device->name, flags);
if (fd < 0) {
ret = -errno;
+ error("cannot open device '%s': %s", device->name,
+ strerror(errno));
goto fail;
}
int btrfs_scan_one_device(int fd, const char *path,
struct btrfs_fs_devices **fs_devices_ret,
- u64 *total_devs, u64 super_offset, int super_recover)
+ u64 *total_devs, u64 super_offset, unsigned sbflags)
{
struct btrfs_super_block *disk_super;
- char *buf;
+ char buf[BTRFS_SUPER_INFO_SIZE];
int ret;
u64 devid;
- buf = malloc(4096);
- if (!buf) {
- ret = -ENOMEM;
- goto error;
- }
disk_super = (struct btrfs_super_block *)buf;
- ret = btrfs_read_dev_super(fd, disk_super, super_offset, super_recover);
- if (ret < 0) {
- ret = -EIO;
- goto error_brelse;
- }
+ ret = btrfs_read_dev_super(fd, disk_super, super_offset, sbflags);
+ if (ret < 0)
+ return -EIO;
devid = btrfs_stack_device_id(&disk_super->dev_item);
if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_METADUMP)
*total_devs = 1;
ret = device_list_add(path, disk_super, devid, fs_devices_ret);
-error_brelse:
- free(buf);
-error:
return ret;
}
/*
+ * find_free_dev_extent_start - find free space in the specified device
+ * @device: the device which we search the free space in
+ * @num_bytes: the size of the free space that we need
+ * @search_start: the position from which to begin the search
+ * @start: store the start of the free space.
+ * @len: the size of the free space. that we find, or the size
+ * of the max free space if we don't find suitable free space
+ *
* this uses a pretty simple search, the expectation is that it is
* called very infrequently and that a given device has a small number
* of extents
+ *
+ * @start is used to store the start of the free space if we find. But if we
+ * don't find suitable free space, it will be used to store the start position
+ * of the max free space.
+ *
+ * @len is used to store the size of the free space that we find.
+ * But if we don't find suitable free space, it is used to store the size of
+ * the max free space.
*/
-static int find_free_dev_extent(struct btrfs_trans_handle *trans,
- struct btrfs_device *device,
- struct btrfs_path *path,
- u64 num_bytes, u64 *start)
+static int find_free_dev_extent_start(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device, u64 num_bytes,
+ u64 search_start, u64 *start, u64 *len)
{
struct btrfs_key key;
struct btrfs_root *root = device->dev_root;
- struct btrfs_dev_extent *dev_extent = NULL;
- u64 hole_size = 0;
- u64 last_byte = 0;
- u64 search_start = root->fs_info->alloc_start;
+ struct btrfs_dev_extent *dev_extent;
+ struct btrfs_path *path;
+ u64 hole_size;
+ u64 max_hole_start;
+ u64 max_hole_size;
+ u64 extent_end;
u64 search_end = device->total_bytes;
int ret;
- int slot = 0;
- int start_found;
+ int slot;
struct extent_buffer *l;
+ u64 min_search_start;
- start_found = 0;
- path->reada = 2;
+ /*
+ * We don't want to overwrite the superblock on the drive nor any area
+ * used by the boot loader (grub for example), so we make sure to start
+ * at an offset of at least 1MB.
+ */
+ min_search_start = max(root->fs_info->alloc_start, (u64)SZ_1M);
+ search_start = max(search_start, min_search_start);
- /* FIXME use last free of some kind */
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
- /* we don't want to overwrite the superblock on the drive,
- * so we make sure to start at an offset of at least 1MB
- */
- search_start = max(BTRFS_BLOCK_RESERVED_1M_FOR_SUPER, search_start);
+ max_hole_start = search_start;
+ max_hole_size = 0;
if (search_start >= search_end) {
ret = -ENOSPC;
- goto error;
+ goto out;
}
+ path->reada = 2;
+
key.objectid = device->devid;
key.offset = search_start;
key.type = BTRFS_DEV_EXTENT_KEY;
- ret = btrfs_search_slot(trans, root, &key, path, 0, 0);
- if (ret < 0)
- goto error;
- ret = btrfs_previous_item(root, path, 0, key.type);
+
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
- goto error;
- l = path->nodes[0];
- btrfs_item_key_to_cpu(l, &key, path->slots[0]);
+ goto out;
+ if (ret > 0) {
+ ret = btrfs_previous_item(root, path, key.objectid, key.type);
+ if (ret < 0)
+ goto out;
+ }
+
while (1) {
l = path->nodes[0];
slot = path->slots[0];
if (ret == 0)
continue;
if (ret < 0)
- goto error;
-no_more_items:
- if (!start_found) {
- if (search_start >= search_end) {
- ret = -ENOSPC;
- goto error;
- }
- *start = search_start;
- start_found = 1;
- goto check_pending;
- }
- *start = last_byte > search_start ?
- last_byte : search_start;
- if (search_end <= *start) {
- ret = -ENOSPC;
- goto error;
- }
- goto check_pending;
+ goto out;
+
+ break;
}
btrfs_item_key_to_cpu(l, &key, slot);
goto next;
if (key.objectid > device->devid)
- goto no_more_items;
-
- if (key.offset >= search_start && key.offset > last_byte &&
- start_found) {
- if (last_byte < search_start)
- last_byte = search_start;
- hole_size = key.offset - last_byte;
- if (key.offset > last_byte &&
- hole_size >= num_bytes) {
- *start = last_byte;
- goto check_pending;
- }
- }
- if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY) {
+ break;
+
+ if (key.type != BTRFS_DEV_EXTENT_KEY)
goto next;
+
+ if (key.offset > search_start) {
+ hole_size = key.offset - search_start;
+
+ /*
+ * Have to check before we set max_hole_start, otherwise
+ * we could end up sending back this offset anyway.
+ */
+ if (hole_size > max_hole_size) {
+ max_hole_start = search_start;
+ max_hole_size = hole_size;
+ }
+
+ /*
+ * If this free space is greater than which we need,
+ * it must be the max free space that we have found
+ * until now, so max_hole_start must point to the start
+ * of this free space and the length of this free space
+ * is stored in max_hole_size. Thus, we return
+ * max_hole_start and max_hole_size and go back to the
+ * caller.
+ */
+ if (hole_size >= num_bytes) {
+ ret = 0;
+ goto out;
+ }
}
- start_found = 1;
dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
- last_byte = key.offset + btrfs_dev_extent_length(l, dev_extent);
+ extent_end = key.offset + btrfs_dev_extent_length(l,
+ dev_extent);
+ if (extent_end > search_start)
+ search_start = extent_end;
next:
path->slots[0]++;
cond_resched();
}
-check_pending:
- /* we have to make sure we didn't find an extent that has already
- * been allocated by the map tree or the original allocation
+
+ /*
+ * At this point, search_start should be the end of
+ * allocated dev extents, and when shrinking the device,
+ * search_end may be smaller than search_start.
*/
- btrfs_release_path(path);
- BUG_ON(*start < search_start);
+ if (search_end > search_start) {
+ hole_size = search_end - search_start;
- if (*start + num_bytes > search_end) {
- ret = -ENOSPC;
- goto error;
+ if (hole_size > max_hole_size) {
+ max_hole_start = search_start;
+ max_hole_size = hole_size;
+ }
}
- /* check for pending inserts here */
- return 0;
-error:
- btrfs_release_path(path);
+ /* See above. */
+ if (max_hole_size < num_bytes)
+ ret = -ENOSPC;
+ else
+ ret = 0;
+
+out:
+ btrfs_free_path(path);
+ *start = max_hole_start;
+ if (len)
+ *len = max_hole_size;
return ret;
}
+int find_free_dev_extent(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device, u64 num_bytes,
+ u64 *start)
+{
+ /* FIXME use last free of some kind */
+ return find_free_dev_extent_start(trans, device,
+ num_bytes, 0, start, NULL);
+}
+
static int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
struct btrfs_device *device,
u64 chunk_tree, u64 chunk_objectid,
u64 chunk_offset,
- u64 num_bytes, u64 *start)
+ u64 num_bytes, u64 *start, int convert)
{
int ret;
struct btrfs_path *path;
if (!path)
return -ENOMEM;
- ret = find_free_dev_extent(trans, device, path, num_bytes, start);
- if (ret) {
- goto err;
+ /*
+ * For convert case, just skip search free dev_extent, as caller
+ * is responsible to make sure it's free.
+ */
+ if (!convert) {
+ ret = find_free_dev_extent(trans, device, num_bytes,
+ start);
+ if (ret)
+ goto err;
}
key.objectid = device->devid;
return ret;
}
-static int find_next_chunk(struct btrfs_root *root, u64 objectid, u64 *offset)
+static int find_next_chunk(struct btrfs_fs_info *fs_info, u64 *offset)
{
+ struct btrfs_root *root = fs_info->chunk_root;
struct btrfs_path *path;
int ret;
struct btrfs_key key;
struct btrfs_key found_key;
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path)
+ return -ENOMEM;
- key.objectid = objectid;
+ key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
key.offset = (u64)-1;
key.type = BTRFS_CHUNK_ITEM_KEY;
} else {
btrfs_item_key_to_cpu(path->nodes[0], &found_key,
path->slots[0]);
- if (found_key.objectid != objectid)
+ if (found_key.objectid != BTRFS_FIRST_CHUNK_TREE_OBJECTID)
*offset = 0;
else {
chunk = btrfs_item_ptr(path->nodes[0], path->slots[0],
* the btrfs_device struct should be fully filled in
*/
int btrfs_add_device(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
+ struct btrfs_fs_info *fs_info,
struct btrfs_device *device)
{
int ret;
struct btrfs_dev_item *dev_item;
struct extent_buffer *leaf;
struct btrfs_key key;
+ struct btrfs_root *root = fs_info->chunk_root;
unsigned long ptr;
u64 free_devid = 0;
- root = root->fs_info->chunk_root;
-
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
ptr = (unsigned long)btrfs_device_uuid(dev_item);
write_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
ptr = (unsigned long)btrfs_device_fsid(dev_item);
- write_extent_buffer(leaf, root->fs_info->fsid, ptr, BTRFS_UUID_SIZE);
+ write_extent_buffer(leaf, fs_info->fsid, ptr, BTRFS_UUID_SIZE);
btrfs_mark_buffer_dirty(leaf);
ret = 0;
return ret;
}
-int btrfs_add_system_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
- struct btrfs_key *key,
+int btrfs_add_system_chunk(struct btrfs_fs_info *fs_info, struct btrfs_key *key,
struct btrfs_chunk *chunk, int item_size)
{
- struct btrfs_super_block *super_copy = root->fs_info->super_copy;
+ struct btrfs_super_block *super_copy = fs_info->super_copy;
struct btrfs_disk_key disk_key;
u32 array_size;
u8 *ptr;
goto next;
if (key.objectid > device->devid)
break;
- if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
+ if (key.type != BTRFS_DEV_EXTENT_KEY)
goto next;
if (key.offset > search_end)
break;
/ sizeof(struct btrfs_stripe) + 1)
int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *extent_root, u64 *start,
+ struct btrfs_fs_info *info, u64 *start,
u64 *num_bytes, u64 type)
{
u64 dev_offset;
- struct btrfs_fs_info *info = extent_root->fs_info;
+ struct btrfs_root *extent_root = info->extent_root;
struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct list_head *dev_list = &info->fs_devices->devices;
struct list_head *cur;
struct map_lookup *map;
- int min_stripe_size = 1 * 1024 * 1024;
- u64 calc_size = 8 * 1024 * 1024;
+ int min_stripe_size = SZ_1M;
+ u64 calc_size = SZ_8M;
u64 min_free;
u64 max_chunk_size = 4 * calc_size;
u64 avail = 0;
return -ENOSPC;
}
- if (type & (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1 |
- BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6 |
- BTRFS_BLOCK_GROUP_RAID10 |
- BTRFS_BLOCK_GROUP_DUP)) {
+ if (type & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
- calc_size = 8 * 1024 * 1024;
+ calc_size = SZ_8M;
max_chunk_size = calc_size * 2;
- min_stripe_size = 1 * 1024 * 1024;
+ min_stripe_size = SZ_1M;
max_stripes = BTRFS_MAX_DEVS_SYS_CHUNK;
} else if (type & BTRFS_BLOCK_GROUP_DATA) {
- calc_size = 1024 * 1024 * 1024;
+ calc_size = SZ_1G;
max_chunk_size = 10 * calc_size;
- min_stripe_size = 64 * 1024 * 1024;
+ min_stripe_size = SZ_64M;
max_stripes = BTRFS_MAX_DEVS(chunk_root);
} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
- calc_size = 1024 * 1024 * 1024;
+ calc_size = SZ_1G;
max_chunk_size = 4 * calc_size;
- min_stripe_size = 32 * 1024 * 1024;
+ min_stripe_size = SZ_32M;
max_stripes = BTRFS_MAX_DEVS(chunk_root);
}
}
}
return -ENOSPC;
}
- ret = find_next_chunk(chunk_root, BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- &offset);
+ ret = find_next_chunk(info, &offset);
if (ret)
return ret;
key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
ret = btrfs_alloc_dev_extent(trans, device,
info->chunk_root->root_key.objectid,
BTRFS_FIRST_CHUNK_TREE_OBJECTID, key.offset,
- calc_size, &dev_offset);
- BUG_ON(ret);
+ calc_size, &dev_offset, 0);
+ if (ret < 0)
+ goto out_chunk_map;
device->bytes_used += calc_size;
ret = btrfs_update_device(trans, device);
- BUG_ON(ret);
+ if (ret < 0)
+ goto out_chunk_map;
map->stripes[index].dev = device;
map->stripes[index].physical = dev_offset;
btrfs_set_stack_chunk_num_stripes(chunk, num_stripes);
btrfs_set_stack_chunk_io_align(chunk, stripe_len);
btrfs_set_stack_chunk_io_width(chunk, stripe_len);
- btrfs_set_stack_chunk_sector_size(chunk, extent_root->sectorsize);
+ btrfs_set_stack_chunk_sector_size(chunk, info->sectorsize);
btrfs_set_stack_chunk_sub_stripes(chunk, sub_stripes);
- map->sector_size = extent_root->sectorsize;
+ map->sector_size = info->sectorsize;
map->stripe_len = stripe_len;
map->io_align = stripe_len;
map->io_width = stripe_len;
map->ce.size = *num_bytes;
ret = insert_cache_extent(&info->mapping_tree.cache_tree, &map->ce);
- BUG_ON(ret);
+ if (ret < 0)
+ goto out_chunk_map;
if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
- ret = btrfs_add_system_chunk(trans, chunk_root, &key,
+ ret = btrfs_add_system_chunk(info, &key,
chunk, btrfs_chunk_item_size(num_stripes));
- BUG_ON(ret);
+ if (ret < 0)
+ goto out_chunk;
}
kfree(chunk);
return ret;
+
+out_chunk_map:
+ kfree(map);
+out_chunk:
+ kfree(chunk);
+ return ret;
}
+/*
+ * Alloc a DATA chunk with SINGLE profile.
+ *
+ * If 'convert' is set, it will alloc a chunk with 1:1 mapping
+ * (btrfs logical bytenr == on-disk bytenr)
+ * For that case, caller must make sure the chunk and dev_extent are not
+ * occupied.
+ */
int btrfs_alloc_data_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *extent_root, u64 *start,
- u64 num_bytes, u64 type)
+ struct btrfs_fs_info *info, u64 *start,
+ u64 num_bytes, u64 type, int convert)
{
u64 dev_offset;
- struct btrfs_fs_info *info = extent_root->fs_info;
+ struct btrfs_root *extent_root = info->extent_root;
struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct list_head *dev_list = &info->fs_devices->devices;
struct list_head *cur;
struct map_lookup *map;
- u64 calc_size = 8 * 1024 * 1024;
+ u64 calc_size = SZ_8M;
int num_stripes = 1;
int sub_stripes = 0;
int ret;
key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
key.type = BTRFS_CHUNK_ITEM_KEY;
- ret = find_next_chunk(chunk_root, BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- &key.offset);
- if (ret)
- return ret;
+ if (convert) {
+ if (*start != round_down(*start, info->sectorsize)) {
+ error("DATA chunk start not sectorsize aligned: %llu",
+ (unsigned long long)*start);
+ return -EINVAL;
+ }
+ key.offset = *start;
+ dev_offset = *start;
+ } else {
+ u64 tmp;
+
+ ret = find_next_chunk(info, &tmp);
+ key.offset = tmp;
+ if (ret)
+ return ret;
+ }
chunk = kmalloc(btrfs_chunk_item_size(num_stripes), GFP_NOFS);
if (!chunk)
ret = btrfs_alloc_dev_extent(trans, device,
info->chunk_root->root_key.objectid,
BTRFS_FIRST_CHUNK_TREE_OBJECTID, key.offset,
- calc_size, &dev_offset);
+ calc_size, &dev_offset, convert);
BUG_ON(ret);
device->bytes_used += calc_size;
btrfs_set_stack_chunk_num_stripes(chunk, num_stripes);
btrfs_set_stack_chunk_io_align(chunk, stripe_len);
btrfs_set_stack_chunk_io_width(chunk, stripe_len);
- btrfs_set_stack_chunk_sector_size(chunk, extent_root->sectorsize);
+ btrfs_set_stack_chunk_sector_size(chunk, info->sectorsize);
btrfs_set_stack_chunk_sub_stripes(chunk, sub_stripes);
- map->sector_size = extent_root->sectorsize;
+ map->sector_size = info->sectorsize;
map->stripe_len = stripe_len;
map->io_align = stripe_len;
map->io_width = stripe_len;
ret = btrfs_insert_item(trans, chunk_root, &key, chunk,
btrfs_chunk_item_size(num_stripes));
BUG_ON(ret);
- *start = key.offset;
+ if (!convert)
+ *start = key.offset;
map->ce.start = key.offset;
map->ce.size = num_bytes;
return ret;
}
-int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len)
+int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len)
{
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
struct cache_extent *ce;
struct map_lookup *map;
int ret;
return ret;
}
-int btrfs_next_metadata(struct btrfs_mapping_tree *map_tree, u64 *logical,
- u64 *size)
+int btrfs_next_bg(struct btrfs_fs_info *fs_info, u64 *logical,
+ u64 *size, u64 type)
{
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
struct cache_extent *ce;
struct map_lookup *map;
+ u64 cur = *logical;
- ce = search_cache_extent(&map_tree->cache_tree, *logical);
+ ce = search_cache_extent(&map_tree->cache_tree, cur);
while (ce) {
- ce = next_cache_extent(ce);
- if (!ce)
- return -ENOENT;
+ /*
+ * only jump to next bg if our cur is not 0
+ * As the initial logical for btrfs_next_bg() is 0, and
+ * if we jump to next bg, we skipped a valid bg.
+ */
+ if (cur) {
+ ce = next_cache_extent(ce);
+ if (!ce)
+ return -ENOENT;
+ }
+ cur = ce->start;
map = container_of(ce, struct map_lookup, ce);
- if (map->type & BTRFS_BLOCK_GROUP_METADATA) {
+ if (map->type & type) {
*logical = ce->start;
*size = ce->size;
return 0;
}
+ if (!cur)
+ ce = next_cache_extent(ce);
}
return -ENOENT;
}
-int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
+int btrfs_rmap_block(struct btrfs_fs_info *fs_info,
u64 chunk_start, u64 physical, u64 devid,
u64 **logical, int *naddrs, int *stripe_len)
{
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
struct cache_extent *ce;
struct map_lookup *map;
u64 *buf;
}
}
-int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
+int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
u64 logical, u64 *length,
struct btrfs_multi_bio **multi_ret, int mirror_num,
u64 **raid_map_ret)
{
- return __btrfs_map_block(map_tree, rw, logical, length, NULL,
+ return __btrfs_map_block(fs_info, rw, logical, length, NULL,
multi_ret, mirror_num, raid_map_ret);
}
-int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
- u64 logical, u64 *length, u64 *type,
- struct btrfs_multi_bio **multi_ret, int mirror_num,
- u64 **raid_map_ret)
+int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw,
+ u64 logical, u64 *length, u64 *type,
+ struct btrfs_multi_bio **multi_ret, int mirror_num,
+ u64 **raid_map_ret)
{
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
struct cache_extent *ce;
struct map_lookup *map;
u64 offset;
return 0;
}
-struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid,
+struct btrfs_device *btrfs_find_device(struct btrfs_fs_info *fs_info, u64 devid,
u8 *uuid, u8 *fsid)
{
struct btrfs_device *device;
struct btrfs_fs_devices *cur_devices;
- cur_devices = root->fs_info->fs_devices;
+ cur_devices = fs_info->fs_devices;
while (cur_devices) {
if (!fsid ||
- !memcmp(cur_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
+ (!memcmp(cur_devices->fsid, fsid, BTRFS_UUID_SIZE) ||
+ fs_info->ignore_fsid_mismatch)) {
device = __find_device(&cur_devices->devices,
devid, uuid);
if (device)
return NULL;
}
-int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
+int btrfs_chunk_readonly(struct btrfs_fs_info *fs_info, u64 chunk_offset)
{
struct cache_extent *ce;
struct map_lookup *map;
- struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
int readonly = 0;
int i;
* corresponding chunk, we will rebuild it later
*/
ce = search_cache_extent(&map_tree->cache_tree, chunk_offset);
- if (!root->fs_info->is_chunk_recover)
+ if (!fs_info->is_chunk_recover)
BUG_ON(!ce);
else
return 0;
}
/*
- * Slot is used to verfy the chunk item is valid
+ * slot == -1: SYSTEM chunk
+ * return -EIO on error, otherwise return 0
+ */
+int btrfs_check_chunk_valid(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *leaf,
+ struct btrfs_chunk *chunk,
+ int slot, u64 logical)
+{
+ u64 length;
+ u64 stripe_len;
+ u16 num_stripes;
+ u16 sub_stripes;
+ u64 type;
+ u32 chunk_ondisk_size;
+ u32 sectorsize = fs_info->sectorsize;
+
+ length = btrfs_chunk_length(leaf, chunk);
+ stripe_len = btrfs_chunk_stripe_len(leaf, chunk);
+ num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
+ sub_stripes = btrfs_chunk_sub_stripes(leaf, chunk);
+ type = btrfs_chunk_type(leaf, chunk);
+
+ /*
+ * These valid checks may be insufficient to cover every corner cases.
+ */
+ if (!IS_ALIGNED(logical, sectorsize)) {
+ error("invalid chunk logical %llu", logical);
+ return -EIO;
+ }
+ if (btrfs_chunk_sector_size(leaf, chunk) != sectorsize) {
+ error("invalid chunk sectorsize %llu",
+ (unsigned long long)btrfs_chunk_sector_size(leaf, chunk));
+ return -EIO;
+ }
+ if (!length || !IS_ALIGNED(length, sectorsize)) {
+ error("invalid chunk length %llu", length);
+ return -EIO;
+ }
+ if (stripe_len != BTRFS_STRIPE_LEN) {
+ error("invalid chunk stripe length: %llu", stripe_len);
+ return -EIO;
+ }
+ /* Check on chunk item type */
+ if (slot == -1 && (type & BTRFS_BLOCK_GROUP_SYSTEM) == 0) {
+ error("invalid chunk type %llu", type);
+ return -EIO;
+ }
+ if (type & ~(BTRFS_BLOCK_GROUP_TYPE_MASK |
+ BTRFS_BLOCK_GROUP_PROFILE_MASK)) {
+ error("unrecognized chunk type: %llu",
+ ~(BTRFS_BLOCK_GROUP_TYPE_MASK |
+ BTRFS_BLOCK_GROUP_PROFILE_MASK) & type);
+ return -EIO;
+ }
+ if (!(type & BTRFS_BLOCK_GROUP_TYPE_MASK)) {
+ error("missing chunk type flag: %llu", type);
+ return -EIO;
+ }
+ if (!(is_power_of_2(type & BTRFS_BLOCK_GROUP_PROFILE_MASK) ||
+ (type & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0)) {
+ error("conflicting chunk type detected: %llu", type);
+ return -EIO;
+ }
+ if ((type & BTRFS_BLOCK_GROUP_PROFILE_MASK) &&
+ !is_power_of_2(type & BTRFS_BLOCK_GROUP_PROFILE_MASK)) {
+ error("conflicting chunk profile detected: %llu", type);
+ return -EIO;
+ }
+
+ chunk_ondisk_size = btrfs_chunk_item_size(num_stripes);
+ /*
+ * Btrfs_chunk contains at least one stripe, and for sys_chunk
+ * it can't exceed the system chunk array size
+ * For normal chunk, it should match its chunk item size.
+ */
+ if (num_stripes < 1 ||
+ (slot == -1 && chunk_ondisk_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) ||
+ (slot >= 0 && chunk_ondisk_size > btrfs_item_size_nr(leaf, slot))) {
+ error("invalid num_stripes: %u", num_stripes);
+ return -EIO;
+ }
+ /*
+ * Device number check against profile
+ */
+ if ((type & BTRFS_BLOCK_GROUP_RAID10 && (sub_stripes != 2 ||
+ !IS_ALIGNED(num_stripes, sub_stripes))) ||
+ (type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) ||
+ (type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) ||
+ (type & BTRFS_BLOCK_GROUP_RAID6 && num_stripes < 3) ||
+ (type & BTRFS_BLOCK_GROUP_DUP && num_stripes > 2) ||
+ ((type & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0 &&
+ num_stripes != 1)) {
+ error("Invalid num_stripes:sub_stripes %u:%u for profile %llu",
+ num_stripes, sub_stripes,
+ type & BTRFS_BLOCK_GROUP_PROFILE_MASK);
+ return -EIO;
+ }
+
+ return 0;
+}
+
+/*
+ * Slot is used to verify the chunk item is valid
*
* For sys chunk in superblock, pass -1 to indicate sys chunk.
*/
-static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
+static int read_one_chunk(struct btrfs_fs_info *fs_info, struct btrfs_key *key,
struct extent_buffer *leaf,
struct btrfs_chunk *chunk, int slot)
{
- struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
struct map_lookup *map;
struct cache_extent *ce;
u64 logical;
logical = key->offset;
length = btrfs_chunk_length(leaf, chunk);
+ num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
+ /* Validation check */
+ ret = btrfs_check_chunk_valid(fs_info, leaf, chunk, slot, logical);
+ if (ret) {
+ error("%s checksums match, but it has an invalid chunk, %s",
+ (slot == -1) ? "Superblock" : "Metadata",
+ (slot == -1) ? "try btrfsck --repair -s <superblock> ie, 0,1,2" : "");
+ return ret;
+ }
ce = search_cache_extent(&map_tree->cache_tree, logical);
return 0;
}
- num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map)
return -ENOMEM;
map->type = btrfs_chunk_type(leaf, chunk);
map->sub_stripes = btrfs_chunk_sub_stripes(leaf, chunk);
- /* Check on chunk item type */
- if (map->type & ~(BTRFS_BLOCK_GROUP_TYPE_MASK |
- BTRFS_BLOCK_GROUP_PROFILE_MASK)) {
- fprintf(stderr, "Unknown chunk type bits: %llu\n",
- map->type & ~(BTRFS_BLOCK_GROUP_TYPE_MASK |
- BTRFS_BLOCK_GROUP_PROFILE_MASK));
- ret = -EIO;
- goto out;
- }
-
- /*
- * Btrfs_chunk contains at least one stripe, and for sys_chunk
- * it can't exceed the system chunk array size
- * For normal chunk, it should match its chunk item size.
- */
- if (num_stripes < 1 ||
- (slot == -1 && sizeof(struct btrfs_stripe) * num_stripes >
- BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) ||
- (slot >= 0 && sizeof(struct btrfs_stripe) * (num_stripes - 1) >
- btrfs_item_size_nr(leaf, slot))) {
- fprintf(stderr, "Invalid num_stripes: %u\n",
- num_stripes);
- ret = -EIO;
- goto out;
- }
-
- /*
- * Device number check against profile
- */
- if ((map->type & BTRFS_BLOCK_GROUP_RAID10 && map->sub_stripes == 0) ||
- (map->type & BTRFS_BLOCK_GROUP_RAID1 && num_stripes < 1) ||
- (map->type & BTRFS_BLOCK_GROUP_RAID5 && num_stripes < 2) ||
- (map->type & BTRFS_BLOCK_GROUP_RAID6 && num_stripes < 3) ||
- (map->type & BTRFS_BLOCK_GROUP_DUP && num_stripes > 2) ||
- ((map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0 &&
- num_stripes != 1)) {
- fprintf(stderr,
- "Invalid num_stripes:sub_stripes %u:%u for profile %llu\n",
- num_stripes, map->sub_stripes,
- map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK);
- ret = -EIO;
- goto out;
- }
-
for (i = 0; i < num_stripes; i++) {
map->stripes[i].physical =
btrfs_stripe_offset_nr(leaf, chunk, i);
read_extent_buffer(leaf, uuid, (unsigned long)
btrfs_stripe_dev_uuid_nr(chunk, i),
BTRFS_UUID_SIZE);
- map->stripes[i].dev = btrfs_find_device(root, devid, uuid,
+ map->stripes[i].dev = btrfs_find_device(fs_info, devid, uuid,
NULL);
if (!map->stripes[i].dev) {
map->stripes[i].dev = fill_missing_device(devid);
printf("warning, device %llu is missing\n",
(unsigned long long)devid);
+ list_add(&map->stripes[i].dev->dev_list,
+ &fs_info->fs_devices->devices);
}
}
BUG_ON(ret);
return 0;
-out:
- free(map);
- return ret;
}
static int fill_device_from_item(struct extent_buffer *leaf,
return 0;
}
-static int open_seed_devices(struct btrfs_root *root, u8 *fsid)
+static int open_seed_devices(struct btrfs_fs_info *fs_info, u8 *fsid)
{
struct btrfs_fs_devices *fs_devices;
int ret;
- fs_devices = root->fs_info->fs_devices->seed;
+ fs_devices = fs_info->fs_devices->seed;
while (fs_devices) {
if (!memcmp(fs_devices->fsid, fsid, BTRFS_UUID_SIZE)) {
ret = 0;
if (ret)
goto out;
- fs_devices->seed = root->fs_info->fs_devices->seed;
- root->fs_info->fs_devices->seed = fs_devices;
+ fs_devices->seed = fs_info->fs_devices->seed;
+ fs_info->fs_devices->seed = fs_devices;
out:
return ret;
}
-static int read_one_dev(struct btrfs_root *root,
+static int read_one_dev(struct btrfs_fs_info *fs_info,
struct extent_buffer *leaf,
struct btrfs_dev_item *dev_item)
{
(unsigned long)btrfs_device_fsid(dev_item),
BTRFS_UUID_SIZE);
- if (memcmp(fs_uuid, root->fs_info->fsid, BTRFS_UUID_SIZE)) {
- ret = open_seed_devices(root, fs_uuid);
+ if (memcmp(fs_uuid, fs_info->fsid, BTRFS_UUID_SIZE)) {
+ ret = open_seed_devices(fs_info, fs_uuid);
if (ret)
return ret;
}
- device = btrfs_find_device(root, devid, dev_uuid, fs_uuid);
+ device = btrfs_find_device(fs_info, devid, dev_uuid, fs_uuid);
if (!device) {
- printk("warning devid %llu not found already\n",
- (unsigned long long)devid);
device = kzalloc(sizeof(*device), GFP_NOFS);
if (!device)
return -ENOMEM;
device->fd = -1;
list_add(&device->dev_list,
- &root->fs_info->fs_devices->devices);
+ &fs_info->fs_devices->devices);
}
fill_device_from_item(leaf, dev_item, device);
- device->dev_root = root->fs_info->dev_root;
+ device->dev_root = fs_info->dev_root;
return ret;
}
-int btrfs_read_sys_array(struct btrfs_root *root)
+int btrfs_read_sys_array(struct btrfs_fs_info *fs_info)
{
- struct btrfs_super_block *super_copy = root->fs_info->super_copy;
+ struct btrfs_super_block *super_copy = fs_info->super_copy;
struct extent_buffer *sb;
struct btrfs_disk_key *disk_key;
struct btrfs_chunk *chunk;
- struct btrfs_key key;
+ u8 *array_ptr;
+ unsigned long sb_array_offset;
+ int ret = 0;
u32 num_stripes;
+ u32 array_size;
u32 len = 0;
- u8 *ptr;
- u8 *array_end;
- int ret = 0;
+ u32 cur_offset;
+ struct btrfs_key key;
- sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET,
- BTRFS_SUPER_INFO_SIZE);
+ if (fs_info->nodesize < BTRFS_SUPER_INFO_SIZE) {
+ printf("ERROR: nodesize %u too small to read superblock\n",
+ fs_info->nodesize);
+ return -EINVAL;
+ }
+ sb = btrfs_find_create_tree_block(fs_info, BTRFS_SUPER_INFO_OFFSET);
if (!sb)
return -ENOMEM;
btrfs_set_buffer_uptodate(sb);
write_extent_buffer(sb, super_copy, 0, sizeof(*super_copy));
- array_end = ((u8 *)super_copy->sys_chunk_array) +
- btrfs_super_sys_array_size(super_copy);
+ array_size = btrfs_super_sys_array_size(super_copy);
- /*
- * we do this loop twice, once for the device items and
- * once for all of the chunks. This way there are device
- * structs filled in for every chunk
- */
- ptr = super_copy->sys_chunk_array;
+ array_ptr = super_copy->sys_chunk_array;
+ sb_array_offset = offsetof(struct btrfs_super_block, sys_chunk_array);
+ cur_offset = 0;
+
+ while (cur_offset < array_size) {
+ disk_key = (struct btrfs_disk_key *)array_ptr;
+ len = sizeof(*disk_key);
+ if (cur_offset + len > array_size)
+ goto out_short_read;
- while (ptr < array_end) {
- disk_key = (struct btrfs_disk_key *)ptr;
btrfs_disk_key_to_cpu(&key, disk_key);
- len = sizeof(*disk_key);
- ptr += len;
+ array_ptr += len;
+ sb_array_offset += len;
+ cur_offset += len;
if (key.type == BTRFS_CHUNK_ITEM_KEY) {
- chunk = (struct btrfs_chunk *)(ptr - (u8 *)super_copy);
- ret = read_one_chunk(root, &key, sb, chunk, -1);
- if (ret)
- break;
+ chunk = (struct btrfs_chunk *)sb_array_offset;
+ /*
+ * At least one btrfs_chunk with one stripe must be
+ * present, exact stripe count check comes afterwards
+ */
+ len = btrfs_chunk_item_size(1);
+ if (cur_offset + len > array_size)
+ goto out_short_read;
+
num_stripes = btrfs_chunk_num_stripes(sb, chunk);
+ if (!num_stripes) {
+ printk(
+ "ERROR: invalid number of stripes %u in sys_array at offset %u\n",
+ num_stripes, cur_offset);
+ ret = -EIO;
+ break;
+ }
+
len = btrfs_chunk_item_size(num_stripes);
+ if (cur_offset + len > array_size)
+ goto out_short_read;
+
+ ret = read_one_chunk(fs_info, &key, sb, chunk, -1);
+ if (ret)
+ break;
} else {
- BUG();
+ printk(
+ "ERROR: unexpected item type %u in sys_array at offset %u\n",
+ (u32)key.type, cur_offset);
+ ret = -EIO;
+ break;
}
- ptr += len;
+ array_ptr += len;
+ sb_array_offset += len;
+ cur_offset += len;
}
free_extent_buffer(sb);
return ret;
+
+out_short_read:
+ printk("ERROR: sys_array too short to read %u bytes at offset %u\n",
+ len, cur_offset);
+ free_extent_buffer(sb);
+ return -EIO;
}
-int btrfs_read_chunk_tree(struct btrfs_root *root)
+int btrfs_read_chunk_tree(struct btrfs_fs_info *fs_info)
{
struct btrfs_path *path;
struct extent_buffer *leaf;
struct btrfs_key key;
struct btrfs_key found_key;
+ struct btrfs_root *root = fs_info->chunk_root;
int ret;
int slot;
- root = root->fs_info->chunk_root;
-
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
struct btrfs_dev_item *dev_item;
dev_item = btrfs_item_ptr(leaf, slot,
struct btrfs_dev_item);
- ret = read_one_dev(root, leaf, dev_item);
+ ret = read_one_dev(fs_info, leaf, dev_item);
BUG_ON(ret);
} else if (found_key.type == BTRFS_CHUNK_ITEM_KEY) {
struct btrfs_chunk *chunk;
chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk);
- ret = read_one_chunk(root, &found_key, leaf, chunk,
+ ret = read_one_chunk(fs_info, &found_key, leaf, chunk,
slot);
BUG_ON(ret);
}
return 0;
}
-static void split_eb_for_raid56(struct btrfs_fs_info *info,
- struct extent_buffer *orig_eb,
+static int split_eb_for_raid56(struct btrfs_fs_info *info,
+ struct extent_buffer *orig_eb,
struct extent_buffer **ebs,
u64 stripe_len, u64 *raid_map,
int num_stripes)
{
- struct extent_buffer *eb;
+ struct extent_buffer **tmp_ebs;
u64 start = orig_eb->start;
u64 this_eb_start;
int i;
- int ret;
+ int ret = 0;
+
+ tmp_ebs = calloc(num_stripes, sizeof(*tmp_ebs));
+ if (!tmp_ebs)
+ return -ENOMEM;
+ /* Alloc memory in a row for data stripes */
for (i = 0; i < num_stripes; i++) {
if (raid_map[i] >= BTRFS_RAID5_P_STRIPE)
break;
- eb = malloc(sizeof(struct extent_buffer) + stripe_len);
- if (!eb)
- BUG();
- memset(eb, 0, sizeof(struct extent_buffer) + stripe_len);
+ tmp_ebs[i] = calloc(1, sizeof(**tmp_ebs) + stripe_len);
+ if (!tmp_ebs[i]) {
+ ret = -ENOMEM;
+ goto clean_up;
+ }
+ }
+
+ for (i = 0; i < num_stripes; i++) {
+ struct extent_buffer *eb = tmp_ebs[i];
+
+ if (raid_map[i] >= BTRFS_RAID5_P_STRIPE)
+ break;
eb->start = raid_map[i];
eb->len = stripe_len;
if (start > this_eb_start ||
start + orig_eb->len < this_eb_start + stripe_len) {
ret = rmw_eb(info, eb, orig_eb);
- BUG_ON(ret);
+ if (ret)
+ goto clean_up;
} else {
- memcpy(eb->data, orig_eb->data + eb->start - start, stripe_len);
+ memcpy(eb->data, orig_eb->data + eb->start - start,
+ stripe_len);
}
ebs[i] = eb;
}
+ free(tmp_ebs);
+ return ret;
+clean_up:
+ for (i = 0; i < num_stripes; i++)
+ free(tmp_ebs[i]);
+ free(tmp_ebs);
+ return ret;
}
int write_raid56_with_parity(struct btrfs_fs_info *info,
{
struct extent_buffer **ebs, *p_eb = NULL, *q_eb = NULL;
int i;
- int j;
int ret;
int alloc_size = eb->len;
+ void **pointers;
- ebs = kmalloc(sizeof(*ebs) * multi->num_stripes, GFP_NOFS);
- BUG_ON(!ebs);
+ ebs = malloc(sizeof(*ebs) * multi->num_stripes);
+ pointers = malloc(sizeof(*pointers) * multi->num_stripes);
+ if (!ebs || !pointers) {
+ free(ebs);
+ free(pointers);
+ return -ENOMEM;
+ }
if (stripe_len > alloc_size)
alloc_size = stripe_len;
- split_eb_for_raid56(info, eb, ebs, stripe_len, raid_map,
- multi->num_stripes);
+ ret = split_eb_for_raid56(info, eb, ebs, stripe_len, raid_map,
+ multi->num_stripes);
+ if (ret)
+ goto out;
for (i = 0; i < multi->num_stripes; i++) {
struct extent_buffer *new_eb;
ebs[i]->dev_bytenr = multi->stripes[i].physical;
ebs[i]->fd = multi->stripes[i].dev->fd;
multi->stripes[i].dev->total_ios++;
- BUG_ON(ebs[i]->start != raid_map[i]);
+ if (ebs[i]->start != raid_map[i]) {
+ ret = -EINVAL;
+ goto out_free_split;
+ }
continue;
}
- new_eb = kmalloc(sizeof(*eb) + alloc_size, GFP_NOFS);
- BUG_ON(!new_eb);
+ new_eb = malloc(sizeof(*eb) + alloc_size);
+ if (!new_eb) {
+ ret = -ENOMEM;
+ goto out_free_split;
+ }
new_eb->dev_bytenr = multi->stripes[i].physical;
new_eb->fd = multi->stripes[i].dev->fd;
multi->stripes[i].dev->total_ios++;
q_eb = new_eb;
}
if (q_eb) {
- void **pointers;
-
- pointers = kmalloc(sizeof(*pointers) * multi->num_stripes,
- GFP_NOFS);
- BUG_ON(!pointers);
-
ebs[multi->num_stripes - 2] = p_eb;
ebs[multi->num_stripes - 1] = q_eb;
pointers[i] = ebs[i]->data;
raid6_gen_syndrome(multi->num_stripes, stripe_len, pointers);
- kfree(pointers);
} else {
ebs[multi->num_stripes - 1] = p_eb;
- memcpy(p_eb->data, ebs[0]->data, stripe_len);
- for (j = 1; j < multi->num_stripes - 1; j++) {
- for (i = 0; i < stripe_len; i += sizeof(unsigned long)) {
- *(unsigned long *)(p_eb->data + i) ^=
- *(unsigned long *)(ebs[j]->data + i);
- }
- }
+ for (i = 0; i < multi->num_stripes; i++)
+ pointers[i] = ebs[i]->data;
+ ret = raid5_gen_result(multi->num_stripes, stripe_len,
+ multi->num_stripes - 1, pointers);
+ if (ret < 0)
+ goto out_free_split;
}
for (i = 0; i < multi->num_stripes; i++) {
ret = write_extent_to_disk(ebs[i]);
- BUG_ON(ret);
+ if (ret < 0)
+ goto out_free_split;
+ }
+
+out_free_split:
+ for (i = 0; i < multi->num_stripes; i++) {
if (ebs[i] != eb)
- kfree(ebs[i]);
+ free(ebs[i]);
}
+out:
+ free(ebs);
+ free(pointers);
- kfree(ebs);
+ return ret;
+}
- return 0;
+/*
+ * Get stripe length from chunk item and its stripe items
+ *
+ * Caller should only call this function after validating the chunk item
+ * by using btrfs_check_chunk_valid().
+ */
+u64 btrfs_stripe_length(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *leaf,
+ struct btrfs_chunk *chunk)
+{
+ u64 stripe_len;
+ u64 chunk_len;
+ u32 num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
+ u64 profile = btrfs_chunk_type(leaf, chunk) &
+ BTRFS_BLOCK_GROUP_PROFILE_MASK;
+
+ chunk_len = btrfs_chunk_length(leaf, chunk);
+
+ switch (profile) {
+ case 0: /* Single profile */
+ case BTRFS_BLOCK_GROUP_RAID1:
+ case BTRFS_BLOCK_GROUP_DUP:
+ stripe_len = chunk_len;
+ break;
+ case BTRFS_BLOCK_GROUP_RAID0:
+ stripe_len = chunk_len / num_stripes;
+ break;
+ case BTRFS_BLOCK_GROUP_RAID5:
+ stripe_len = chunk_len / (num_stripes - 1);
+ break;
+ case BTRFS_BLOCK_GROUP_RAID6:
+ stripe_len = chunk_len / (num_stripes - 2);
+ break;
+ case BTRFS_BLOCK_GROUP_RAID10:
+ stripe_len = chunk_len / (num_stripes /
+ btrfs_chunk_sub_stripes(leaf, chunk));
+ break;
+ default:
+ /* Invalid chunk profile found */
+ BUG_ON(1);
+ }
+ return stripe_len;
+}
+
+/*
+ * Return 0 if size of @device is already good
+ * Return >0 if size of @device is not aligned but fixed without problems
+ * Return <0 if something wrong happened when aligning the size of @device
+ */
+int btrfs_fix_device_size(struct btrfs_fs_info *fs_info,
+ struct btrfs_device *device)
+{
+ struct btrfs_trans_handle *trans;
+ struct btrfs_key key;
+ struct btrfs_path path;
+ struct btrfs_root *chunk_root = fs_info->chunk_root;
+ struct btrfs_dev_item *di;
+ u64 old_bytes = device->total_bytes;
+ int ret;
+
+ if (IS_ALIGNED(old_bytes, fs_info->sectorsize))
+ return 0;
+
+ /* Align the in-memory total_bytes first, and use it as correct size */
+ device->total_bytes = round_down(device->total_bytes,
+ fs_info->sectorsize);
+
+ key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
+ key.type = BTRFS_DEV_ITEM_KEY;
+ key.offset = device->devid;
+
+ trans = btrfs_start_transaction(chunk_root, 1);
+ if (IS_ERR(trans)) {
+ ret = PTR_ERR(trans);
+ error("error starting transaction: %d (%s)",
+ ret, strerror(-ret));
+ return ret;
+ }
+
+ btrfs_init_path(&path);
+ ret = btrfs_search_slot(trans, chunk_root, &key, &path, 0, 1);
+ if (ret > 0) {
+ error("failed to find DEV_ITEM for devid %llu", device->devid);
+ ret = -ENOENT;
+ goto err;
+ }
+ if (ret < 0) {
+ error("failed to search chunk root: %d (%s)",
+ ret, strerror(-ret));
+ goto err;
+ }
+ di = btrfs_item_ptr(path.nodes[0], path.slots[0], struct btrfs_dev_item);
+ btrfs_set_device_total_bytes(path.nodes[0], di, device->total_bytes);
+ btrfs_mark_buffer_dirty(path.nodes[0]);
+ ret = btrfs_commit_transaction(trans, chunk_root);
+ if (ret < 0) {
+ error("failed to commit current transaction: %d (%s)",
+ ret, strerror(-ret));
+ btrfs_release_path(&path);
+ return ret;
+ }
+ btrfs_release_path(&path);
+ printf("Fixed device size for devid %llu, old size: %llu new size: %llu\n",
+ device->devid, old_bytes, device->total_bytes);
+ return 1;
+
+err:
+ /* We haven't modified anything, it's OK to commit current trans */
+ btrfs_commit_transaction(trans, chunk_root);
+ btrfs_release_path(&path);
+ return ret;
+}
+
+/*
+ * Return 0 if super block total_bytes matches all devices' total_bytes
+ * Return >0 if super block total_bytes mismatch but fixed without problem
+ * Return <0 if we failed to fix super block total_bytes
+ */
+int btrfs_fix_super_size(struct btrfs_fs_info *fs_info)
+{
+ struct btrfs_trans_handle *trans;
+ struct btrfs_device *device;
+ struct list_head *dev_list = &fs_info->fs_devices->devices;
+ u64 total_bytes = 0;
+ u64 old_bytes = btrfs_super_total_bytes(fs_info->super_copy);
+ int ret;
+
+ list_for_each_entry(device, dev_list, dev_list) {
+ /*
+ * Caller should ensure this function is called after aligning
+ * all devices' total_bytes.
+ */
+ if (!IS_ALIGNED(device->total_bytes, fs_info->sectorsize)) {
+ error("device %llu total_bytes %llu not aligned to %u",
+ device->devid, device->total_bytes,
+ fs_info->sectorsize);
+ return -EUCLEAN;
+ }
+ total_bytes += device->total_bytes;
+ }
+
+ if (total_bytes == old_bytes)
+ return 0;
+
+ btrfs_set_super_total_bytes(fs_info->super_copy, total_bytes);
+
+ /* Commit transaction to update all super blocks */
+ trans = btrfs_start_transaction(fs_info->tree_root, 1);
+ if (IS_ERR(trans)) {
+ ret = PTR_ERR(trans);
+ error("error starting transaction: %d (%s)",
+ ret, strerror(-ret));
+ return ret;
+ }
+ ret = btrfs_commit_transaction(trans, fs_info->tree_root);
+ if (ret < 0) {
+ error("failed to commit current transaction: %d (%s)",
+ ret, strerror(-ret));
+ return ret;
+ }
+ printf("Fixed super total bytes, old size: %llu new size: %llu\n",
+ old_bytes, total_bytes);
+ return 1;
+}
+
+/*
+ * Return 0 if all devices and super block sizes are good
+ * Return >0 if any device/super size problem was found, but fixed
+ * Return <0 if something wrong happened during fixing
+ */
+int btrfs_fix_device_and_super_size(struct btrfs_fs_info *fs_info)
+{
+ struct btrfs_device *device;
+ struct list_head *dev_list = &fs_info->fs_devices->devices;
+ bool have_bad_value = false;
+ int ret;
+
+ /* Seed device is not supported yet */
+ if (fs_info->fs_devices->seed) {
+ error("fixing device size with seed device is not supported yet");
+ return -EOPNOTSUPP;
+ }
+
+ /* All devices must be set up before repairing */
+ if (list_empty(dev_list)) {
+ error("no device found");
+ return -ENODEV;
+ }
+ list_for_each_entry(device, dev_list, dev_list) {
+ if (device->fd == -1 || !device->writeable) {
+ error("devid %llu is missing or not writeable",
+ device->devid);
+ error(
+ "fixing device size needs all device(s) to be present and writeable");
+ return -ENODEV;
+ }
+ }
+
+ /* Repair total_bytes of each device */
+ list_for_each_entry(device, dev_list, dev_list) {
+ ret = btrfs_fix_device_size(fs_info, device);
+ if (ret < 0)
+ return ret;
+ if (ret > 0)
+ have_bad_value = true;
+ }
+
+ /* Repair super total_byte */
+ ret = btrfs_fix_super_size(fs_info);
+ if (ret > 0)
+ have_bad_value = true;
+ if (have_bad_value) {
+ printf(
+ "Fixed unaligned/mismatched total_bytes for super block and device items\n");
+ ret = 1;
+ } else {
+ printf("No device size related problem found\n");
+ ret = 0;
+ }
+ return ret;
}