* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*/
-#define _XOPEN_SOURCE 600
-#define __USE_XOPEN2K
#include <stdio.h>
#include <stdlib.h>
#include <sys/types.h>
#include "transaction.h"
#include "print-tree.h"
#include "volumes.h"
+#include "utils.h"
struct stripe {
struct btrfs_device *dev;
u64 physical;
};
-struct map_lookup {
- struct cache_extent ce;
- u64 type;
- int io_align;
- int io_width;
- int stripe_len;
- int sector_size;
- int num_stripes;
- int sub_stripes;
- struct btrfs_bio_stripe stripes[];
-};
+static inline int nr_parity_stripes(struct map_lookup *map)
+{
+ if (map->type & BTRFS_BLOCK_GROUP_RAID5)
+ return 1;
+ else if (map->type & BTRFS_BLOCK_GROUP_RAID6)
+ return 2;
+ else
+ return 0;
+}
-#define map_lookup_size(n) (sizeof(struct map_lookup) + \
- (sizeof(struct btrfs_bio_stripe) * (n)))
+static inline int nr_data_stripes(struct map_lookup *map)
+{
+ return map->num_stripes - nr_parity_stripes(map);
+}
+
+#define is_parity_stripe(x) ( ((x) == BTRFS_RAID5_P_STRIPE) || ((x) == BTRFS_RAID6_Q_STRIPE) )
static LIST_HEAD(fs_uuids);
/* we can safely leave the fs_devices entry around */
return -ENOMEM;
}
+ device->fd = -1;
device->devid = devid;
+ device->generation = found_transid;
memcpy(device->uuid, disk_super->dev_item.uuid,
BTRFS_UUID_SIZE);
device->name = kstrdup(path, GFP_NOFS);
return -ENOMEM;
}
device->label = kstrdup(disk_super->label, GFP_NOFS);
+ if (!device->label) {
+ kfree(device->name);
+ kfree(device);
+ return -ENOMEM;
+ }
device->total_devs = btrfs_super_num_devices(disk_super);
device->super_bytes_used = btrfs_super_bytes_used(disk_super);
device->total_bytes =
btrfs_stack_device_bytes_used(&disk_super->dev_item);
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);
+ if (!name)
+ return -ENOMEM;
+ kfree(device->name);
+ device->name = name;
+ }
+
if (found_transid > fs_devices->latest_trans) {
fs_devices->latest_devid = devid;
int btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
{
struct btrfs_fs_devices *seed_devices;
- struct list_head *cur;
struct btrfs_device *device;
+ int ret = 0;
+
again:
- list_for_each(cur, &fs_devices->devices) {
- device = list_entry(cur, struct btrfs_device, dev_list);
- close(device->fd);
- device->fd = -1;
+ 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) {
+ 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);
+ device->fd = -1;
+ }
device->writeable = 0;
+ list_del(&device->dev_list);
+ /* free the memory */
+ free(device->name);
+ free(device->label);
+ free(device);
}
seed_devices = fs_devices->seed;
fs_devices->seed = NULL;
if (seed_devices) {
+ struct btrfs_fs_devices *orig;
+
+ orig = fs_devices;
fs_devices = seed_devices;
+ list_del(&orig->list);
+ free(orig);
goto again;
+ } else {
+ list_del(&fs_devices->list);
+ 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)
list_for_each(cur, head) {
device = list_entry(cur, struct btrfs_device, 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;
}
+ if (posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED))
+ fprintf(stderr, "Warning, could not drop caches\n");
+
if (device->devid == fs_devices->latest_devid)
fs_devices->latest_bdev = fd;
if (device->devid == fs_devices->lowest_devid)
fs_devices->lowest_bdev = fd;
device->fd = fd;
- if (flags == O_RDWR)
+ if (flags & O_RDWR)
device->writeable = 1;
}
return 0;
int btrfs_scan_one_device(int fd, const char *path,
struct btrfs_fs_devices **fs_devices_ret,
- u64 *total_devs, u64 super_offset)
+ 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;
- char uuidbuf[37];
- 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);
- if (ret < 0) {
- ret = -EIO;
- goto error_brelse;
- }
- devid = le64_to_cpu(disk_super->dev_item.devid);
+ 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;
else
*total_devs = btrfs_super_num_devices(disk_super);
- uuid_unparse(disk_super->fsid, uuidbuf);
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 = 0;
+ 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((u64)1024 * 1024, search_start);
+ max_hole_start = search_start;
+ max_hole_size = 0;
- if (root->fs_info->alloc_start + num_bytes <= device->total_bytes)
- search_start = max(root->fs_info->alloc_start, search_start);
+ if (search_start >= search_end) {
+ ret = -ENOSPC;
+ 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(root, 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(root, 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 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)
+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, 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;
struct btrfs_key found_key;
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path)
+ return -ENOMEM;
key.objectid = objectid;
key.offset = (u64)-1;
}
ret = 0;
error:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return ret;
}
return ret;
}
-int btrfs_add_system_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
+int btrfs_add_system_chunk(struct btrfs_root *root,
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 = root->fs_info->super_copy;
struct btrfs_disk_key disk_key;
u32 array_size;
u8 *ptr;
array_size = btrfs_super_sys_array_size(super_copy);
- if (array_size + item_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE)
+ if (array_size + item_size + sizeof(disk_key)
+ > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE)
return -EFBIG;
ptr = super_copy->sys_chunk_array + array_size;
return 0;
}
-static u64 div_factor(u64 num, int factor)
-{
- if (factor == 10)
- return num;
- num *= factor;
- return num / 10;
-}
-
static u64 chunk_bytes_by_type(u64 type, u64 calc_size, int num_stripes,
int sub_stripes)
{
return calc_size;
else if (type & BTRFS_BLOCK_GROUP_RAID10)
return calc_size * (num_stripes / sub_stripes);
+ else if (type & BTRFS_BLOCK_GROUP_RAID5)
+ return calc_size * (num_stripes - 1);
+ else if (type & BTRFS_BLOCK_GROUP_RAID6)
+ return calc_size * (num_stripes - 2);
else
return calc_size * num_stripes;
}
+static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
+{
+ /* TODO, add a way to store the preferred stripe size */
+ return BTRFS_STRIPE_LEN;
+}
+
+/*
+ * btrfs_device_avail_bytes - count bytes available for alloc_chunk
+ *
+ * It is not equal to "device->total_bytes - device->bytes_used".
+ * We do not allocate any chunk in 1M at beginning of device, and not
+ * allowed to allocate any chunk before alloc_start if it is specified.
+ * So search holes from max(1M, alloc_start) to device->total_bytes.
+ */
+static int btrfs_device_avail_bytes(struct btrfs_trans_handle *trans,
+ struct btrfs_device *device,
+ u64 *avail_bytes)
+{
+ struct btrfs_path *path;
+ struct btrfs_root *root = device->dev_root;
+ struct btrfs_key key;
+ struct btrfs_dev_extent *dev_extent = NULL;
+ struct extent_buffer *l;
+ u64 search_start = root->fs_info->alloc_start;
+ u64 search_end = device->total_bytes;
+ u64 extent_end = 0;
+ u64 free_bytes = 0;
+ int ret;
+ int slot = 0;
+
+ search_start = max(BTRFS_BLOCK_RESERVED_1M_FOR_SUPER, search_start);
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ key.objectid = device->devid;
+ key.offset = root->fs_info->alloc_start;
+ key.type = BTRFS_DEV_EXTENT_KEY;
+
+ path->reada = 2;
+ ret = btrfs_search_slot(trans, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto error;
+ ret = btrfs_previous_item(root, path, 0, key.type);
+ if (ret < 0)
+ goto error;
+
+ while (1) {
+ l = path->nodes[0];
+ slot = path->slots[0];
+ if (slot >= btrfs_header_nritems(l)) {
+ ret = btrfs_next_leaf(root, path);
+ if (ret == 0)
+ continue;
+ if (ret < 0)
+ goto error;
+ break;
+ }
+ btrfs_item_key_to_cpu(l, &key, slot);
+
+ if (key.objectid < device->devid)
+ goto next;
+ if (key.objectid > device->devid)
+ break;
+ if (key.type != BTRFS_DEV_EXTENT_KEY)
+ goto next;
+ if (key.offset > search_end)
+ break;
+ if (key.offset > search_start)
+ free_bytes += key.offset - search_start;
+
+ dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
+ extent_end = key.offset + btrfs_dev_extent_length(l,
+ dev_extent);
+ if (extent_end > search_start)
+ search_start = extent_end;
+ if (search_start > search_end)
+ break;
+next:
+ path->slots[0]++;
+ cond_resched();
+ }
+
+ if (search_start < search_end)
+ free_bytes += search_end - search_start;
+
+ *avail_bytes = free_bytes;
+ ret = 0;
+error:
+ btrfs_free_path(path);
+ return ret;
+}
+
+#define BTRFS_MAX_DEVS(r) ((BTRFS_LEAF_DATA_SIZE(r) \
+ - sizeof(struct btrfs_item) \
+ - sizeof(struct btrfs_chunk)) \
+ / sizeof(struct btrfs_stripe) + 1)
+
+#define BTRFS_MAX_DEVS_SYS_CHUNK ((BTRFS_SYSTEM_CHUNK_ARRAY_SIZE \
+ - 2 * sizeof(struct btrfs_disk_key) \
+ - 2 * sizeof(struct btrfs_chunk)) \
+ / sizeof(struct btrfs_stripe) + 1)
+
int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
struct btrfs_root *extent_root, u64 *start,
u64 *num_bytes, u64 type)
{
u64 dev_offset;
struct btrfs_fs_info *info = extent_root->fs_info;
- struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
+ struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct btrfs_chunk *chunk;
struct list_head private_devs;
- struct list_head *dev_list = &extent_root->fs_info->fs_devices->devices;
+ 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;
+ u64 avail = 0;
u64 max_avail = 0;
u64 percent_max;
int num_stripes = 1;
+ int max_stripes = 0;
int min_stripes = 1;
int sub_stripes = 0;
int looped = 0;
int ret;
int index;
- int stripe_len = 64 * 1024;
+ int stripe_len = BTRFS_STRIPE_LEN;
struct btrfs_key key;
+ u64 offset;
if (list_empty(dev_list)) {
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_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);
}
}
if (type & BTRFS_BLOCK_GROUP_RAID1) {
num_stripes = min_t(u64, 2,
- btrfs_super_num_devices(&info->super_copy));
+ btrfs_super_num_devices(info->super_copy));
if (num_stripes < 2)
return -ENOSPC;
min_stripes = 2;
min_stripes = 2;
}
if (type & (BTRFS_BLOCK_GROUP_RAID0)) {
- num_stripes = btrfs_super_num_devices(&info->super_copy);
+ num_stripes = btrfs_super_num_devices(info->super_copy);
+ if (num_stripes > max_stripes)
+ num_stripes = max_stripes;
min_stripes = 2;
}
if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
- num_stripes = btrfs_super_num_devices(&info->super_copy);
+ num_stripes = btrfs_super_num_devices(info->super_copy);
+ if (num_stripes > max_stripes)
+ num_stripes = max_stripes;
if (num_stripes < 4)
return -ENOSPC;
num_stripes &= ~(u32)1;
sub_stripes = 2;
min_stripes = 4;
}
+ if (type & (BTRFS_BLOCK_GROUP_RAID5)) {
+ num_stripes = btrfs_super_num_devices(info->super_copy);
+ if (num_stripes > max_stripes)
+ num_stripes = max_stripes;
+ if (num_stripes < 2)
+ return -ENOSPC;
+ min_stripes = 2;
+ stripe_len = find_raid56_stripe_len(num_stripes - 1,
+ btrfs_super_stripesize(info->super_copy));
+ }
+ if (type & (BTRFS_BLOCK_GROUP_RAID6)) {
+ num_stripes = btrfs_super_num_devices(info->super_copy);
+ if (num_stripes > max_stripes)
+ num_stripes = max_stripes;
+ if (num_stripes < 3)
+ return -ENOSPC;
+ min_stripes = 3;
+ stripe_len = find_raid56_stripe_len(num_stripes - 2,
+ btrfs_super_stripesize(info->super_copy));
+ }
/* we don't want a chunk larger than 10% of the FS */
- percent_max = div_factor(btrfs_super_total_bytes(&info->super_copy), 1);
+ percent_max = div_factor(btrfs_super_total_bytes(info->super_copy), 1);
max_chunk_size = min(percent_max, max_chunk_size);
again:
/* build a private list of devices we will allocate from */
while(index < num_stripes) {
device = list_entry(cur, struct btrfs_device, dev_list);
- avail = device->total_bytes - device->bytes_used;
+ ret = btrfs_device_avail_bytes(trans, device, &avail);
+ if (ret)
+ return ret;
cur = cur->next;
if (avail >= min_free) {
list_move_tail(&device->dev_list, &private_devs);
}
return -ENOSPC;
}
- 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);
+ &offset);
if (ret)
return ret;
+ key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID;
+ key.type = BTRFS_CHUNK_ITEM_KEY;
+ key.offset = offset;
chunk = kmalloc(btrfs_chunk_item_size(num_stripes), GFP_NOFS);
if (!chunk)
return -ENOMEM;
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map) {
kfree(chunk);
return -ENOMEM;
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, 0);
BUG_ON(ret);
device->bytes_used += calc_size;
map->ce.start = key.offset;
map->ce.size = *num_bytes;
- ret = insert_existing_cache_extent(
- &extent_root->fs_info->mapping_tree.cache_tree,
- &map->ce);
+ ret = insert_cache_extent(&info->mapping_tree.cache_tree, &map->ce);
BUG_ON(ret);
if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
- ret = btrfs_add_system_chunk(trans, chunk_root, &key,
+ ret = btrfs_add_system_chunk(chunk_root, &key,
chunk, btrfs_chunk_item_size(num_stripes));
BUG_ON(ret);
}
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)
+ u64 num_bytes, u64 type, int convert)
{
u64 dev_offset;
struct btrfs_fs_info *info = extent_root->fs_info;
- struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
+ struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct btrfs_chunk *chunk;
- struct list_head *dev_list = &extent_root->fs_info->fs_devices->devices;
+ 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;
int index;
- int stripe_len = 64 * 1024;
+ int stripe_len = BTRFS_STRIPE_LEN;
struct btrfs_key key;
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, extent_root->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(chunk_root,
+ BTRFS_FIRST_CHUNK_TREE_OBJECTID,
+ &tmp);
+ key.offset = tmp;
+ if (ret)
+ return ret;
+ }
chunk = kmalloc(btrfs_chunk_item_size(num_stripes), GFP_NOFS);
if (!chunk)
return -ENOMEM;
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map) {
kfree(chunk);
return -ENOMEM;
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;
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;
- ret = insert_existing_cache_extent(
- &extent_root->fs_info->mapping_tree.cache_tree,
- &map->ce);
+ ret = insert_cache_extent(&info->mapping_tree.cache_tree, &map->ce);
BUG_ON(ret);
kfree(chunk);
return ret;
}
-void btrfs_mapping_init(struct btrfs_mapping_tree *tree)
-{
- cache_tree_init(&tree->cache_tree);
-}
-
int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len)
{
struct cache_extent *ce;
struct map_lookup *map;
int ret;
- ce = find_first_cache_extent(&map_tree->cache_tree, logical);
- BUG_ON(!ce);
- BUG_ON(ce->start > logical || ce->start + ce->size < logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
+ if (!ce) {
+ fprintf(stderr, "No mapping for %llu-%llu\n",
+ (unsigned long long)logical,
+ (unsigned long long)logical+len);
+ return 1;
+ }
+ if (ce->start > logical || ce->start + ce->size < logical) {
+ fprintf(stderr, "Invalid mapping for %llu-%llu, got "
+ "%llu-%llu\n", (unsigned long long)logical,
+ (unsigned long long)logical+len,
+ (unsigned long long)ce->start,
+ (unsigned long long)ce->start + ce->size);
+ return 1;
+ }
map = container_of(ce, struct map_lookup, ce);
if (map->type & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1))
ret = map->num_stripes;
else if (map->type & BTRFS_BLOCK_GROUP_RAID10)
ret = map->sub_stripes;
+ else if (map->type & BTRFS_BLOCK_GROUP_RAID5)
+ ret = 2;
+ else if (map->type & BTRFS_BLOCK_GROUP_RAID6)
+ ret = 3;
else
ret = 1;
return ret;
}
-int btrfs_next_metadata(struct btrfs_mapping_tree *map_tree, u64 *logical,
- u64 *size)
+int btrfs_next_bg(struct btrfs_mapping_tree *map_tree, u64 *logical,
+ u64 *size, u64 type)
{
struct cache_extent *ce;
struct map_lookup *map;
+ u64 cur = *logical;
- ce = find_first_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;
u64 bytenr;
u64 length;
u64 stripe_nr;
+ u64 rmap_len;
int i, j, nr = 0;
- ce = find_first_cache_extent(&map_tree->cache_tree, chunk_start);
+ ce = search_cache_extent(&map_tree->cache_tree, chunk_start);
BUG_ON(!ce);
map = container_of(ce, struct map_lookup, ce);
length = ce->size;
+ rmap_len = map->stripe_len;
if (map->type & BTRFS_BLOCK_GROUP_RAID10)
length = ce->size / (map->num_stripes / map->sub_stripes);
else if (map->type & BTRFS_BLOCK_GROUP_RAID0)
length = ce->size / map->num_stripes;
+ else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
+ BTRFS_BLOCK_GROUP_RAID6)) {
+ length = ce->size / nr_data_stripes(map);
+ rmap_len = map->stripe_len * nr_data_stripes(map);
+ }
buf = kzalloc(sizeof(u64) * map->num_stripes, GFP_NOFS);
map->sub_stripes;
} else if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
stripe_nr = stripe_nr * map->num_stripes + i;
- }
- bytenr = ce->start + stripe_nr * map->stripe_len;
+ } /* else if RAID[56], multiply by nr_data_stripes().
+ * Alternatively, just use rmap_len below instead of
+ * map->stripe_len */
+
+ bytenr = ce->start + stripe_nr * rmap_len;
for (j = 0; j < nr; j++) {
if (buf[j] == bytenr)
break;
*logical = buf;
*naddrs = nr;
- *stripe_len = map->stripe_len;
+ *stripe_len = rmap_len;
return 0;
}
+static inline int parity_smaller(u64 a, u64 b)
+{
+ return a > b;
+}
+
+/* Bubble-sort the stripe set to put the parity/syndrome stripes last */
+static void sort_parity_stripes(struct btrfs_multi_bio *bbio, u64 *raid_map)
+{
+ struct btrfs_bio_stripe s;
+ int i;
+ u64 l;
+ int again = 1;
+
+ while (again) {
+ again = 0;
+ for (i = 0; i < bbio->num_stripes - 1; i++) {
+ if (parity_smaller(raid_map[i], raid_map[i+1])) {
+ s = bbio->stripes[i];
+ l = raid_map[i];
+ bbio->stripes[i] = bbio->stripes[i+1];
+ raid_map[i] = raid_map[i+1];
+ bbio->stripes[i+1] = s;
+ raid_map[i+1] = l;
+ again = 1;
+ }
+ }
+ }
+}
+
int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
u64 logical, u64 *length,
- struct btrfs_multi_bio **multi_ret, int mirror_num)
+ struct btrfs_multi_bio **multi_ret, int mirror_num,
+ u64 **raid_map_ret)
{
return __btrfs_map_block(map_tree, rw, logical, length, NULL,
- multi_ret, mirror_num);
+ 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)
+ struct btrfs_multi_bio **multi_ret, int mirror_num,
+ u64 **raid_map_ret)
{
struct cache_extent *ce;
struct map_lookup *map;
u64 offset;
u64 stripe_offset;
u64 stripe_nr;
+ u64 *raid_map = NULL;
int stripes_allocated = 8;
int stripes_required = 1;
int stripe_index;
stripes_allocated = 1;
}
again:
- ce = find_first_cache_extent(&map_tree->cache_tree, logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
if (!ce) {
- if (multi)
- kfree(multi);
+ kfree(multi);
+ *length = (u64)-1;
return -ENOENT;
}
- if (ce->start > logical || ce->start + ce->size < logical) {
- if (multi)
- kfree(multi);
+ if (ce->start > logical) {
+ kfree(multi);
+ *length = ce->start - logical;
return -ENOENT;
}
stripes_required = map->sub_stripes;
}
}
+ if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)
+ && multi_ret && ((rw & WRITE) || mirror_num > 1) && raid_map_ret) {
+ /* RAID[56] write or recovery. Return all stripes */
+ stripes_required = map->num_stripes;
+
+ /* Only allocate the map if we've already got a large enough multi_ret */
+ if (stripes_allocated >= stripes_required) {
+ raid_map = kmalloc(sizeof(u64) * map->num_stripes, GFP_NOFS);
+ if (!raid_map) {
+ kfree(multi);
+ return -ENOMEM;
+ }
+ }
+ }
+
/* if our multi bio struct is too small, back off and try again */
- if (multi_ret && rw == WRITE &&
- stripes_allocated < stripes_required) {
- stripes_allocated = map->num_stripes;
+ if (multi_ret && stripes_allocated < stripes_required) {
+ stripes_allocated = stripes_required;
kfree(multi);
+ multi = NULL;
goto again;
}
stripe_nr = offset;
stripe_offset = offset - stripe_offset;
if (map->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)) {
/* we limit the length of each bio to what fits in a stripe */
multi->num_stripes = map->num_stripes;
else if (mirror_num)
stripe_index = mirror_num - 1;
+ } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 |
+ BTRFS_BLOCK_GROUP_RAID6)) {
+
+ if (raid_map) {
+ int rot;
+ u64 tmp;
+ u64 raid56_full_stripe_start;
+ u64 full_stripe_len = nr_data_stripes(map) * map->stripe_len;
+
+ /*
+ * align the start of our data stripe in the logical
+ * address space
+ */
+ raid56_full_stripe_start = offset / full_stripe_len;
+ raid56_full_stripe_start *= full_stripe_len;
+
+ /* get the data stripe number */
+ stripe_nr = raid56_full_stripe_start / map->stripe_len;
+ stripe_nr = stripe_nr / nr_data_stripes(map);
+
+ /* Work out the disk rotation on this stripe-set */
+ rot = stripe_nr % map->num_stripes;
+
+ /* Fill in the logical address of each stripe */
+ tmp = stripe_nr * nr_data_stripes(map);
+
+ for (i = 0; i < nr_data_stripes(map); i++)
+ raid_map[(i+rot) % map->num_stripes] =
+ ce->start + (tmp + i) * map->stripe_len;
+
+ raid_map[(i+rot) % map->num_stripes] = BTRFS_RAID5_P_STRIPE;
+ if (map->type & BTRFS_BLOCK_GROUP_RAID6)
+ raid_map[(i+rot+1) % map->num_stripes] = BTRFS_RAID6_Q_STRIPE;
+
+ *length = map->stripe_len;
+ stripe_index = 0;
+ stripe_offset = 0;
+ multi->num_stripes = map->num_stripes;
+ } else {
+ stripe_index = stripe_nr % nr_data_stripes(map);
+ stripe_nr = stripe_nr / nr_data_stripes(map);
+
+ /*
+ * Mirror #0 or #1 means the original data block.
+ * Mirror #2 is RAID5 parity block.
+ * Mirror #3 is RAID6 Q block.
+ */
+ if (mirror_num > 1)
+ stripe_index = nr_data_stripes(map) + mirror_num - 2;
+
+ /* We distribute the parity blocks across stripes */
+ stripe_index = (stripe_nr + stripe_index) % map->num_stripes;
+ }
} else {
/*
* after this do_div call, stripe_nr is the number of stripes
stripe_index++;
}
*multi_ret = multi;
+
if (type)
*type = map->type;
+
+ if (raid_map) {
+ sort_parity_stripes(multi, raid_map);
+ *raid_map_ret = raid_map;
+ }
out:
return 0;
}
cur_devices = root->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) ||
+ root->fs_info->ignore_fsid_mismatch)) {
device = __find_device(&cur_devices->devices,
devid, uuid);
if (device)
return NULL;
}
-int btrfs_bootstrap_super_map(struct btrfs_mapping_tree *map_tree,
- struct btrfs_fs_devices *fs_devices)
+struct btrfs_device *
+btrfs_find_device_by_devid(struct btrfs_fs_devices *fs_devices,
+ u64 devid, int instance)
{
- struct map_lookup *map;
- u64 logical = BTRFS_SUPER_INFO_OFFSET;
- u64 length = BTRFS_SUPER_INFO_SIZE;
- int num_stripes = 0;
- int sub_stripes = 0;
- int ret;
- int i;
- struct list_head *cur;
-
- list_for_each(cur, &fs_devices->devices) {
- num_stripes++;
- }
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
- if (!map)
- return -ENOMEM;
+ struct list_head *head = &fs_devices->devices;
+ struct btrfs_device *dev;
+ int num_found = 0;
- map->ce.start = logical;
- map->ce.size = length;
- map->num_stripes = num_stripes;
- map->sub_stripes = sub_stripes;
- map->io_width = length;
- map->io_align = length;
- map->sector_size = length;
- map->stripe_len = length;
- map->type = BTRFS_BLOCK_GROUP_RAID1;
-
- i = 0;
- list_for_each(cur, &fs_devices->devices) {
- struct btrfs_device *device = list_entry(cur,
- struct btrfs_device,
- dev_list);
- map->stripes[i].physical = logical;
- map->stripes[i].dev = device;
- i++;
- }
- ret = insert_existing_cache_extent(&map_tree->cache_tree, &map->ce);
- if (ret == -EEXIST) {
- struct cache_extent *old;
- struct map_lookup *old_map;
- old = find_cache_extent(&map_tree->cache_tree, logical, length);
- old_map = container_of(old, struct map_lookup, ce);
- remove_cache_extent(&map_tree->cache_tree, old);
- kfree(old_map);
- ret = insert_existing_cache_extent(&map_tree->cache_tree,
- &map->ce);
+ list_for_each_entry(dev, head, dev_list) {
+ if (dev->devid == devid && num_found++ == instance)
+ return dev;
}
- BUG_ON(ret);
- return 0;
+ return NULL;
}
int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
int readonly = 0;
int i;
- ce = find_first_cache_extent(&map_tree->cache_tree, chunk_offset);
- BUG_ON(!ce);
+ /*
+ * During chunk recovering, we may fail to find block group's
+ * corresponding chunk, we will rebuild it later
+ */
+ ce = search_cache_extent(&map_tree->cache_tree, chunk_offset);
+ if (!root->fs_info->is_chunk_recover)
+ BUG_ON(!ce);
+ else
+ return 0;
map = container_of(ce, struct map_lookup, ce);
for (i = 0; i < map->num_stripes; i++) {
return device;
}
+/*
+ * slot == -1: SYSTEM chunk
+ * return -EIO on error, otherwise return 0
+ */
+int btrfs_check_chunk_valid(struct btrfs_root *root,
+ 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;
+
+ 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, root->sectorsize)) {
+ error("invalid chunk logical %llu", logical);
+ return -EIO;
+ }
+ if (btrfs_chunk_sector_size(leaf, chunk) != root->sectorsize) {
+ error("invalid chunk sectorsize %llu",
+ (unsigned long long)btrfs_chunk_sector_size(leaf, chunk));
+ return -EIO;
+ }
+ if (!length || !IS_ALIGNED(length, root->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;
+ }
+ /*
+ * 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))) {
+ error("invalid num_stripes: %u", num_stripes);
+ return -EIO;
+ }
+ /*
+ * Device number check against profile
+ */
+ if ((type & BTRFS_BLOCK_GROUP_RAID10 && sub_stripes == 0) ||
+ (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,
struct extent_buffer *leaf,
- struct btrfs_chunk *chunk)
+ struct btrfs_chunk *chunk, int slot)
{
struct btrfs_mapping_tree *map_tree = &root->fs_info->mapping_tree;
struct map_lookup *map;
logical = key->offset;
length = btrfs_chunk_length(leaf, chunk);
+ num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
+ /* Validation check */
+ ret = btrfs_check_chunk_valid(root, 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 = find_first_cache_extent(&map_tree->cache_tree, logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
/* already mapped? */
if (ce && ce->start <= logical && ce->start + ce->size > logical) {
return 0;
}
- num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map)
return -ENOMEM;
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,
+ &root->fs_info->fs_devices->devices);
}
}
- ret = insert_existing_cache_extent(&map_tree->cache_tree, &map->ce);
+ ret = insert_cache_extent(&map_tree->cache_tree, &map->ce);
BUG_ON(ret);
return 0;
fs_devices = find_fsid(fsid);
if (!fs_devices) {
- ret = -ENOENT;
- goto out;
+ /* missing all seed devices */
+ fs_devices = kzalloc(sizeof(*fs_devices), GFP_NOFS);
+ if (!fs_devices) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ INIT_LIST_HEAD(&fs_devices->devices);
+ list_add(&fs_devices->list, &fs_uuids);
+ memcpy(fs_devices->fsid, fsid, BTRFS_FSID_SIZE);
}
ret = btrfs_open_devices(fs_devices, O_RDONLY);
device = btrfs_find_device(root, devid, dev_uuid, fs_uuid);
if (!device) {
- printk("warning devid %llu not found already\n",
- (unsigned long long)devid);
- device = kmalloc(sizeof(*device), GFP_NOFS);
+ device = kzalloc(sizeof(*device), GFP_NOFS);
if (!device)
return -ENOMEM;
- device->total_ios = 0;
+ device->fd = -1;
list_add(&device->dev_list,
&root->fs_info->fs_devices->devices);
}
return ret;
}
-int btrfs_read_super_device(struct btrfs_root *root, struct extent_buffer *buf)
-{
- struct btrfs_dev_item *dev_item;
-
- dev_item = (struct btrfs_dev_item *)offsetof(struct btrfs_super_block,
- dev_item);
- return read_one_dev(root, buf, dev_item);
-}
-
int btrfs_read_sys_array(struct btrfs_root *root)
{
- struct btrfs_super_block *super_copy = &root->fs_info->super_copy;
+ struct btrfs_super_block *super_copy = root->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;
- unsigned long sb_ptr;
- u32 cur;
- int ret = 0;
+ u32 cur_offset;
+ struct btrfs_key key;
- sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET,
+ sb = btrfs_find_create_tree_block(root->fs_info,
+ BTRFS_SUPER_INFO_OFFSET,
BTRFS_SUPER_INFO_SIZE);
if (!sb)
return -ENOMEM;
btrfs_set_buffer_uptodate(sb);
- write_extent_buffer(sb, super_copy, 0, BTRFS_SUPER_INFO_SIZE);
+ write_extent_buffer(sb, super_copy, 0, sizeof(*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;
- sb_ptr = offsetof(struct btrfs_super_block, sys_chunk_array);
- cur = 0;
+ 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 (cur < array_size) {
- disk_key = (struct btrfs_disk_key *)ptr;
btrfs_disk_key_to_cpu(&key, disk_key);
- len = sizeof(*disk_key);
- ptr += len;
- sb_ptr += len;
- cur += len;
+ array_ptr += len;
+ sb_array_offset += len;
+ cur_offset += len;
if (key.type == BTRFS_CHUNK_ITEM_KEY) {
- chunk = (struct btrfs_chunk *)sb_ptr;
- ret = read_one_chunk(root, &key, sb, chunk);
- 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(root, &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;
- sb_ptr += len;
- cur += 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)
if (!path)
return -ENOMEM;
- /* first we search for all of the device items, and then we
- * read in all of the chunk items. This way we can create chunk
- * mappings that reference all of the devices that are afound
+ /*
+ * Read all device items, and then all the chunk items. All
+ * device items are found before any chunk item (their object id
+ * is smaller than the lowest possible object id for a chunk
+ * item - BTRFS_FIRST_CHUNK_TREE_OBJECTID).
*/
key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
key.offset = 0;
key.type = 0;
-again:
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto error;
while(1) {
leaf = path->nodes[0];
slot = path->slots[0];
break;
}
btrfs_item_key_to_cpu(leaf, &found_key, slot);
- if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
- if (found_key.objectid != BTRFS_DEV_ITEMS_OBJECTID)
- break;
- if (found_key.type == BTRFS_DEV_ITEM_KEY) {
- struct btrfs_dev_item *dev_item;
- dev_item = btrfs_item_ptr(leaf, slot,
+ if (found_key.type == BTRFS_DEV_ITEM_KEY) {
+ 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);
- BUG_ON(ret);
- }
+ ret = read_one_dev(root, 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(root, &found_key, leaf, chunk,
+ slot);
BUG_ON(ret);
}
path->slots[0]++;
}
- if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
- key.objectid = 0;
- btrfs_release_path(root, path);
- goto again;
- }
- btrfs_free_path(path);
ret = 0;
error:
+ btrfs_free_path(path);
return ret;
}
{
return &fs_uuids;
}
+
+static int rmw_eb(struct btrfs_fs_info *info,
+ struct extent_buffer *eb, struct extent_buffer *orig_eb)
+{
+ int ret;
+ unsigned long orig_off = 0;
+ unsigned long dest_off = 0;
+ unsigned long copy_len = eb->len;
+
+ ret = read_whole_eb(info, eb, 0);
+ if (ret)
+ return ret;
+
+ if (eb->start + eb->len <= orig_eb->start ||
+ eb->start >= orig_eb->start + orig_eb->len)
+ return 0;
+ /*
+ * | ----- orig_eb ------- |
+ * | ----- stripe ------- |
+ * | ----- orig_eb ------- |
+ * | ----- orig_eb ------- |
+ */
+ if (eb->start > orig_eb->start)
+ orig_off = eb->start - orig_eb->start;
+ if (orig_eb->start > eb->start)
+ dest_off = orig_eb->start - eb->start;
+
+ if (copy_len > orig_eb->len - orig_off)
+ copy_len = orig_eb->len - orig_off;
+ if (copy_len > eb->len - dest_off)
+ copy_len = eb->len - dest_off;
+
+ memcpy(eb->data + dest_off, orig_eb->data + orig_off, copy_len);
+ return 0;
+}
+
+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 **tmp_ebs;
+ u64 start = orig_eb->start;
+ u64 this_eb_start;
+ int i;
+ 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;
+
+ 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;
+ eb->refs = 1;
+ eb->flags = 0;
+ eb->fd = -1;
+ eb->dev_bytenr = (u64)-1;
+
+ this_eb_start = raid_map[i];
+
+ if (start > this_eb_start ||
+ start + orig_eb->len < this_eb_start + stripe_len) {
+ ret = rmw_eb(info, eb, orig_eb);
+ if (ret)
+ goto clean_up;
+ } else {
+ 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 *eb,
+ struct btrfs_multi_bio *multi,
+ u64 stripe_len, u64 *raid_map)
+{
+ struct extent_buffer **ebs, *p_eb = NULL, *q_eb = NULL;
+ int i;
+ int ret;
+ int alloc_size = eb->len;
+ void **pointers;
+
+ 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;
+
+ 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;
+ if (raid_map[i] < BTRFS_RAID5_P_STRIPE) {
+ ebs[i]->dev_bytenr = multi->stripes[i].physical;
+ ebs[i]->fd = multi->stripes[i].dev->fd;
+ multi->stripes[i].dev->total_ios++;
+ if (ebs[i]->start != raid_map[i]) {
+ ret = -EINVAL;
+ goto out_free_split;
+ }
+ continue;
+ }
+ 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++;
+ new_eb->len = stripe_len;
+
+ if (raid_map[i] == BTRFS_RAID5_P_STRIPE)
+ p_eb = new_eb;
+ else if (raid_map[i] == BTRFS_RAID6_Q_STRIPE)
+ q_eb = new_eb;
+ }
+ if (q_eb) {
+ ebs[multi->num_stripes - 2] = p_eb;
+ ebs[multi->num_stripes - 1] = q_eb;
+
+ for (i = 0; i < multi->num_stripes; i++)
+ pointers[i] = ebs[i]->data;
+
+ raid6_gen_syndrome(multi->num_stripes, stripe_len, pointers);
+ } else {
+ ebs[multi->num_stripes - 1] = p_eb;
+ 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]);
+ if (ret < 0)
+ goto out_free_split;
+ }
+
+out_free_split:
+ for (i = 0; i < multi->num_stripes; i++) {
+ if (ebs[i] != eb)
+ free(ebs[i]);
+ }
+out:
+ free(ebs);
+ free(pointers);
+
+ return ret;
+}
projects / platform / upstream / btrfs-progs.git / blobdiff