diff options
Diffstat (limited to 'fs/btrfs/tree-log.c')
| -rw-r--r-- | fs/btrfs/tree-log.c | 1124 |
1 files changed, 704 insertions, 420 deletions
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 451fad96ecd..9e1f2cd5e67 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -18,15 +18,13 @@ #include <linux/sched.h> #include <linux/slab.h> +#include <linux/blkdev.h> #include <linux/list_sort.h> -#include "ctree.h" -#include "transaction.h" +#include "tree-log.h" #include "disk-io.h" #include "locking.h" #include "print-tree.h" #include "backref.h" -#include "compat.h" -#include "tree-log.h" #include "hash.h" /* magic values for the inode_only field in btrfs_log_inode: @@ -92,7 +90,8 @@ */ #define LOG_WALK_PIN_ONLY 0 #define LOG_WALK_REPLAY_INODES 1 -#define LOG_WALK_REPLAY_ALL 2 +#define LOG_WALK_REPLAY_DIR_INDEX 2 +#define LOG_WALK_REPLAY_ALL 3 static int btrfs_log_inode(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode, @@ -135,43 +134,61 @@ static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans, * syncing the tree wait for us to finish */ static int start_log_trans(struct btrfs_trans_handle *trans, - struct btrfs_root *root) + struct btrfs_root *root, + struct btrfs_log_ctx *ctx) { + int index; int ret; - int err = 0; mutex_lock(&root->log_mutex); if (root->log_root) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { + ret = -EAGAIN; + goto out; + } if (!root->log_start_pid) { root->log_start_pid = current->pid; - root->log_multiple_pids = false; + clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); } else if (root->log_start_pid != current->pid) { - root->log_multiple_pids = true; + set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); } atomic_inc(&root->log_batch); atomic_inc(&root->log_writers); + if (ctx) { + index = root->log_transid % 2; + list_add_tail(&ctx->list, &root->log_ctxs[index]); + ctx->log_transid = root->log_transid; + } mutex_unlock(&root->log_mutex); return 0; } - root->log_multiple_pids = false; - root->log_start_pid = current->pid; + + ret = 0; mutex_lock(&root->fs_info->tree_log_mutex); - if (!root->fs_info->log_root_tree) { + if (!root->fs_info->log_root_tree) ret = btrfs_init_log_root_tree(trans, root->fs_info); - if (ret) - err = ret; - } - if (err == 0 && !root->log_root) { + mutex_unlock(&root->fs_info->tree_log_mutex); + if (ret) + goto out; + + if (!root->log_root) { ret = btrfs_add_log_tree(trans, root); if (ret) - err = ret; + goto out; } - mutex_unlock(&root->fs_info->tree_log_mutex); + clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); + root->log_start_pid = current->pid; atomic_inc(&root->log_batch); atomic_inc(&root->log_writers); + if (ctx) { + index = root->log_transid % 2; + list_add_tail(&ctx->list, &root->log_ctxs[index]); + ctx->log_transid = root->log_transid; + } +out: mutex_unlock(&root->log_mutex); - return err; + return ret; } /* @@ -277,17 +294,31 @@ static int process_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, struct walk_control *wc, u64 gen) { + int ret = 0; + + /* + * If this fs is mixed then we need to be able to process the leaves to + * pin down any logged extents, so we have to read the block. + */ + if (btrfs_fs_incompat(log->fs_info, MIXED_GROUPS)) { + ret = btrfs_read_buffer(eb, gen); + if (ret) + return ret; + } + if (wc->pin) - btrfs_pin_extent_for_log_replay(log->fs_info->extent_root, - eb->start, eb->len); + ret = btrfs_pin_extent_for_log_replay(log->fs_info->extent_root, + eb->start, eb->len); - if (btrfs_buffer_uptodate(eb, gen, 0)) { + if (!ret && btrfs_buffer_uptodate(eb, gen, 0)) { + if (wc->pin && btrfs_header_level(eb) == 0) + ret = btrfs_exclude_logged_extents(log, eb); if (wc->write) btrfs_write_tree_block(eb); if (wc->wait) btrfs_wait_tree_block_writeback(eb); } - return 0; + return ret; } /* @@ -317,6 +348,7 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, unsigned long src_ptr; unsigned long dst_ptr; int overwrite_root = 0; + bool inode_item = key->type == BTRFS_INODE_ITEM_KEY; if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) overwrite_root = 1; @@ -326,6 +358,9 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, /* look for the key in the destination tree */ ret = btrfs_search_slot(NULL, root, key, path, 0, 0); + if (ret < 0) + return ret; + if (ret == 0) { char *src_copy; char *dst_copy; @@ -367,6 +402,50 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, return 0; } + /* + * We need to load the old nbytes into the inode so when we + * replay the extents we've logged we get the right nbytes. + */ + if (inode_item) { + struct btrfs_inode_item *item; + u64 nbytes; + u32 mode; + + item = btrfs_item_ptr(path->nodes[0], path->slots[0], + struct btrfs_inode_item); + nbytes = btrfs_inode_nbytes(path->nodes[0], item); + item = btrfs_item_ptr(eb, slot, + struct btrfs_inode_item); + btrfs_set_inode_nbytes(eb, item, nbytes); + + /* + * If this is a directory we need to reset the i_size to + * 0 so that we can set it up properly when replaying + * the rest of the items in this log. + */ + mode = btrfs_inode_mode(eb, item); + if (S_ISDIR(mode)) + btrfs_set_inode_size(eb, item, 0); + } + } else if (inode_item) { + struct btrfs_inode_item *item; + u32 mode; + + /* + * New inode, set nbytes to 0 so that the nbytes comes out + * properly when we replay the extents. + */ + item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item); + btrfs_set_inode_nbytes(eb, item, 0); + + /* + * If this is a directory we need to reset the i_size to 0 so + * that we can set it up properly when replaying the rest of + * the items in this log. + */ + mode = btrfs_inode_mode(eb, item); + if (S_ISDIR(mode)) + btrfs_set_inode_size(eb, item, 0); } insert: btrfs_release_path(path); @@ -380,9 +459,9 @@ insert: found_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]); if (found_size > item_size) - btrfs_truncate_item(trans, root, path, item_size, 1); + btrfs_truncate_item(root, path, item_size, 1); else if (found_size < item_size) - btrfs_extend_item(trans, root, path, + btrfs_extend_item(root, path, item_size - found_size); } else if (ret) { return ret; @@ -486,7 +565,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, int found_type; u64 extent_end; u64 start = key->offset; - u64 saved_nbytes; + u64 nbytes = 0; struct btrfs_file_extent_item *item; struct inode *inode = NULL; unsigned long size; @@ -496,10 +575,19 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, found_type = btrfs_file_extent_type(eb, item); if (found_type == BTRFS_FILE_EXTENT_REG || - found_type == BTRFS_FILE_EXTENT_PREALLOC) - extent_end = start + btrfs_file_extent_num_bytes(eb, item); - else if (found_type == BTRFS_FILE_EXTENT_INLINE) { - size = btrfs_file_extent_inline_len(eb, item); + found_type == BTRFS_FILE_EXTENT_PREALLOC) { + nbytes = btrfs_file_extent_num_bytes(eb, item); + extent_end = start + nbytes; + + /* + * We don't add to the inodes nbytes if we are prealloc or a + * hole. + */ + if (btrfs_file_extent_disk_bytenr(eb, item) == 0) + nbytes = 0; + } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { + size = btrfs_file_extent_inline_len(eb, slot, item); + nbytes = btrfs_file_extent_ram_bytes(eb, item); extent_end = ALIGN(start + size, root->sectorsize); } else { ret = 0; @@ -548,10 +636,10 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, } btrfs_release_path(path); - saved_nbytes = inode_get_bytes(inode); /* drop any overlapping extents */ ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1); - BUG_ON(ret); + if (ret) + goto out; if (found_type == BTRFS_FILE_EXTENT_REG || found_type == BTRFS_FILE_EXTENT_PREALLOC) { @@ -561,7 +649,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ret = btrfs_insert_empty_item(trans, root, path, key, sizeof(*item)); - BUG_ON(ret); + if (ret) + goto out; dest_offset = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]); copy_extent_buffer(path->nodes[0], eb, dest_offset, @@ -587,7 +676,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ins.objectid, ins.offset, 0, root->root_key.objectid, key->objectid, offset, 0); - BUG_ON(ret); + if (ret) + goto out; } else { /* * insert the extent pointer in the extent @@ -596,7 +686,8 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ret = btrfs_alloc_logged_file_extent(trans, root, root->root_key.objectid, key->objectid, offset, &ins); - BUG_ON(ret); + if (ret) + goto out; } btrfs_release_path(path); @@ -613,29 +704,33 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ret = btrfs_lookup_csums_range(root->log_root, csum_start, csum_end - 1, &ordered_sums, 0); - BUG_ON(ret); + if (ret) + goto out; while (!list_empty(&ordered_sums)) { struct btrfs_ordered_sum *sums; sums = list_entry(ordered_sums.next, struct btrfs_ordered_sum, list); - ret = btrfs_csum_file_blocks(trans, + if (!ret) + ret = btrfs_csum_file_blocks(trans, root->fs_info->csum_root, sums); - BUG_ON(ret); list_del(&sums->list); kfree(sums); } + if (ret) + goto out; } else { btrfs_release_path(path); } } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { /* inline extents are easy, we just overwrite them */ ret = overwrite_item(trans, root, path, eb, slot, key); - BUG_ON(ret); + if (ret) + goto out; } - inode_set_bytes(inode, saved_nbytes); + inode_add_bytes(inode, nbytes); ret = btrfs_update_inode(trans, root, inode); out: if (inode) @@ -677,20 +772,22 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, inode = read_one_inode(root, location.objectid); if (!inode) { - kfree(name); - return -EIO; + ret = -EIO; + goto out; } ret = link_to_fixup_dir(trans, root, path, location.objectid); - BUG_ON(ret); + if (ret) + goto out; ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len); - BUG_ON(ret); + if (ret) + goto out; + else + ret = btrfs_run_delayed_items(trans, root); +out: kfree(name); - iput(inode); - - btrfs_run_delayed_items(trans, root); return ret; } @@ -843,7 +940,8 @@ again: victim_name_len = btrfs_inode_ref_name_len(leaf, victim_ref); victim_name = kmalloc(victim_name_len, GFP_NOFS); - BUG_ON(!victim_name); + if (!victim_name) + return -ENOMEM; read_extent_buffer(leaf, victim_name, (unsigned long)(victim_ref + 1), @@ -853,15 +951,18 @@ again: parent_objectid, victim_name, victim_name_len)) { - btrfs_inc_nlink(inode); + inc_nlink(inode); btrfs_release_path(path); ret = btrfs_unlink_inode(trans, root, dir, inode, victim_name, victim_name_len); - BUG_ON(ret); - btrfs_run_delayed_items(trans, root); kfree(victim_name); + if (ret) + return ret; + ret = btrfs_run_delayed_items(trans, root); + if (ret) + return ret; *search_done = 1; goto again; } @@ -869,7 +970,6 @@ again: ptr = (unsigned long)(victim_ref + 1) + victim_name_len; } - BUG_ON(ret); /* * NOTE: we have searched root tree and checked the @@ -903,6 +1003,8 @@ again: goto next; victim_name = kmalloc(victim_name_len, GFP_NOFS); + if (!victim_name) + return -ENOMEM; read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name, victim_name_len); @@ -919,7 +1021,7 @@ again: victim_parent = read_one_inode(root, parent_objectid); if (victim_parent) { - btrfs_inc_nlink(inode); + inc_nlink(inode); btrfs_release_path(path); ret = btrfs_unlink_inode(trans, root, @@ -927,16 +1029,20 @@ again: inode, victim_name, victim_name_len); - btrfs_run_delayed_items(trans, root); + if (!ret) + ret = btrfs_run_delayed_items( + trans, root); } - BUG_ON(ret); iput(victim_parent); kfree(victim_name); + if (ret) + return ret; *search_done = 1; goto again; } kfree(victim_name); - BUG_ON(ret); + if (ret) + return ret; next: cur_offset += victim_name_len + sizeof(*extref); } @@ -949,7 +1055,8 @@ next: ref_index, name, namelen, 0); if (di && !IS_ERR(di)) { ret = drop_one_dir_item(trans, root, path, dir, di); - BUG_ON(ret); + if (ret) + return ret; } btrfs_release_path(path); @@ -958,7 +1065,8 @@ next: name, namelen, 0); if (di && !IS_ERR(di)) { ret = drop_one_dir_item(trans, root, path, dir, di); - BUG_ON(ret); + if (ret) + return ret; } btrfs_release_path(path); @@ -1020,11 +1128,11 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, struct extent_buffer *eb, int slot, struct btrfs_key *key) { - struct inode *dir; - struct inode *inode; + struct inode *dir = NULL; + struct inode *inode = NULL; unsigned long ref_ptr; unsigned long ref_end; - char *name; + char *name = NULL; int namelen; int ret; int search_done = 0; @@ -1057,13 +1165,15 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, * care of the rest */ dir = read_one_inode(root, parent_objectid); - if (!dir) - return -ENOENT; + if (!dir) { + ret = -ENOENT; + goto out; + } inode = read_one_inode(root, inode_objectid); if (!inode) { - iput(dir); - return -EIO; + ret = -EIO; + goto out; } while (ref_ptr < ref_end) { @@ -1076,14 +1186,16 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, */ if (!dir) dir = read_one_inode(root, parent_objectid); - if (!dir) - return -ENOENT; + if (!dir) { + ret = -ENOENT; + goto out; + } } else { ret = ref_get_fields(eb, ref_ptr, &namelen, &name, &ref_index); } if (ret) - return ret; + goto out; /* if we already have a perfect match, we're done */ if (!inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode), @@ -1103,21 +1215,25 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, parent_objectid, ref_index, name, namelen, &search_done); - if (ret == 1) + if (ret) { + if (ret == 1) + ret = 0; goto out; - BUG_ON(ret); + } } /* insert our name */ ret = btrfs_add_link(trans, dir, inode, name, namelen, 0, ref_index); - BUG_ON(ret); + if (ret) + goto out; btrfs_update_inode(trans, root, inode); } ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen; kfree(name); + name = NULL; if (log_ref_ver) { iput(dir); dir = NULL; @@ -1126,20 +1242,20 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, /* finally write the back reference in the inode */ ret = overwrite_item(trans, root, path, eb, slot, key); - BUG_ON(ret); - out: btrfs_release_path(path); + kfree(name); iput(dir); iput(inode); - return 0; + return ret; } static int insert_orphan_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 offset) { int ret; - ret = btrfs_find_orphan_item(root, offset); + ret = btrfs_find_item(root, NULL, BTRFS_ORPHAN_OBJECTID, + offset, BTRFS_ORPHAN_ITEM_KEY, NULL); if (ret > 0) ret = btrfs_insert_orphan_item(trans, root, offset); return ret; @@ -1212,6 +1328,7 @@ static int count_inode_refs(struct btrfs_root *root, break; path->slots[0]--; } +process_slot: btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); if (key.objectid != ino || @@ -1232,6 +1349,10 @@ static int count_inode_refs(struct btrfs_root *root, if (key.offset == 0) break; + if (path->slots[0] > 0) { + path->slots[0]--; + goto process_slot; + } key.offset--; btrfs_release_path(path); } @@ -1290,10 +1411,10 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, if (S_ISDIR(inode->i_mode)) { ret = replay_dir_deletes(trans, root, NULL, path, ino, 1); - BUG_ON(ret); + if (ret) + goto out; } ret = insert_orphan_item(trans, root, ino); - BUG_ON(ret); } out: @@ -1338,9 +1459,9 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans, return -EIO; ret = fixup_inode_link_count(trans, root, inode); - BUG_ON(ret); - iput(inode); + if (ret) + goto out; /* * fixup on a directory may create new entries, @@ -1385,12 +1506,12 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, if (!inode->i_nlink) set_nlink(inode, 1); else - btrfs_inc_nlink(inode); + inc_nlink(inode); ret = btrfs_update_inode(trans, root, inode); } else if (ret == -EEXIST) { ret = 0; } else { - BUG(); + BUG(); /* Logic Error */ } iput(inode); @@ -1422,6 +1543,7 @@ static noinline int insert_one_name(struct btrfs_trans_handle *trans, iput(inode); return -EIO; } + ret = btrfs_add_link(trans, dir, inode, name, name_len, 1, index); /* FIXME, put inode into FIXUP list */ @@ -1459,7 +1581,8 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, struct inode *dir; u8 log_type; int exists; - int ret; + int ret = 0; + bool update_size = (key->type == BTRFS_DIR_INDEX_KEY); dir = read_one_inode(root, key->objectid); if (!dir) @@ -1467,8 +1590,10 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, name_len = btrfs_dir_name_len(eb, di); name = kmalloc(name_len, GFP_NOFS); - if (!name) - return -ENOMEM; + if (!name) { + ret = -ENOMEM; + goto out; + } log_type = btrfs_dir_type(eb, di); read_extent_buffer(eb, name, (unsigned long)(di + 1), @@ -1491,7 +1616,9 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, key->offset, name, name_len, 1); } else { - BUG(); + /* Corruption */ + ret = -EINVAL; + goto out; } if (IS_ERR_OR_NULL(dst_di)) { /* we need a sequence number to insert, so we only @@ -1519,22 +1646,29 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, goto out; ret = drop_one_dir_item(trans, root, path, dir, dst_di); - BUG_ON(ret); + if (ret) + goto out; if (key->type == BTRFS_DIR_INDEX_KEY) goto insert; out: btrfs_release_path(path); + if (!ret && update_size) { + btrfs_i_size_write(dir, dir->i_size + name_len * 2); + ret = btrfs_update_inode(trans, root, dir); + } kfree(name); iput(dir); - return 0; + return ret; insert: btrfs_release_path(path); ret = insert_one_name(trans, root, path, key->objectid, key->offset, name, name_len, log_type, &log_key); - - BUG_ON(ret && ret != -ENOENT); + if (ret && ret != -ENOENT) + goto out; + update_size = false; + ret = 0; goto out; } @@ -1565,7 +1699,8 @@ static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans, return -EIO; name_len = btrfs_dir_name_len(eb, di); ret = replay_one_name(trans, root, path, eb, di, key); - BUG_ON(ret); + if (ret) + return ret; ptr = (unsigned long)(di + 1); ptr += name_len; } @@ -1714,7 +1849,7 @@ again: dir_key->offset, name, name_len, 0); } - if (IS_ERR_OR_NULL(log_di)) { + if (!log_di || (IS_ERR(log_di) && PTR_ERR(log_di) == -ENOENT)) { btrfs_dir_item_key_to_cpu(eb, di, &location); btrfs_release_path(path); btrfs_release_path(log_path); @@ -1726,16 +1861,21 @@ again: ret = link_to_fixup_dir(trans, root, path, location.objectid); - BUG_ON(ret); - btrfs_inc_nlink(inode); + if (ret) { + kfree(name); + iput(inode); + goto out; + } + + inc_nlink(inode); ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len); - BUG_ON(ret); - - btrfs_run_delayed_items(trans, root); - + if (!ret) + ret = btrfs_run_delayed_items(trans, root); kfree(name); iput(inode); + if (ret) + goto out; /* there might still be more names under this key * check and repeat if required @@ -1746,6 +1886,9 @@ again: goto again; ret = 0; goto out; + } else if (IS_ERR(log_di)) { + kfree(name); + return PTR_ERR(log_di); } btrfs_release_path(log_path); kfree(name); @@ -1839,7 +1982,8 @@ again: ret = check_item_in_log(trans, root, log, path, log_path, dir, &found_key); - BUG_ON(ret); + if (ret) + goto out; if (found_key.offset == (u64)-1) break; dir_key.offset = found_key.offset + 1; @@ -1916,11 +2060,13 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, if (S_ISDIR(mode)) { ret = replay_dir_deletes(wc->trans, root, log, path, key.objectid, 0); - BUG_ON(ret); + if (ret) + break; } ret = overwrite_item(wc->trans, root, path, eb, i, &key); - BUG_ON(ret); + if (ret) + break; /* for regular files, make sure corresponding * orhpan item exist. extents past the new EOF @@ -1929,13 +2075,24 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, if (S_ISREG(mode)) { ret = insert_orphan_item(wc->trans, root, key.objectid); - BUG_ON(ret); + if (ret) + break; } ret = link_to_fixup_dir(wc->trans, root, path, key.objectid); - BUG_ON(ret); + if (ret) + break; + } + + if (key.type == BTRFS_DIR_INDEX_KEY && + wc->stage == LOG_WALK_REPLAY_DIR_INDEX) { + ret = replay_one_dir_item(wc->trans, root, path, + eb, i, &key); + if (ret) + break; } + if (wc->stage < LOG_WALK_REPLAY_ALL) continue; @@ -1943,28 +2100,29 @@ static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb, if (key.type == BTRFS_XATTR_ITEM_KEY) { ret = overwrite_item(wc->trans, root, path, eb, i, &key); - BUG_ON(ret); - } else if (key.type == BTRFS_INODE_REF_KEY) { - ret = add_inode_ref(wc->trans, root, log, path, - eb, i, &key); - BUG_ON(ret && ret != -ENOENT); - } else if (key.type == BTRFS_INODE_EXTREF_KEY) { + if (ret) + break; + } else if (key.type == BTRFS_INODE_REF_KEY || + key.type == BTRFS_INODE_EXTREF_KEY) { ret = add_inode_ref(wc->trans, root, log, path, eb, i, &key); - BUG_ON(ret && ret != -ENOENT); + if (ret && ret != -ENOENT) + break; + ret = 0; } else if (key.type == BTRFS_EXTENT_DATA_KEY) { ret = replay_one_extent(wc->trans, root, path, eb, i, &key); - BUG_ON(ret); - } else if (key.type == BTRFS_DIR_ITEM_KEY || - key.type == BTRFS_DIR_INDEX_KEY) { + if (ret) + break; + } else if (key.type == BTRFS_DIR_ITEM_KEY) { ret = replay_one_dir_item(wc->trans, root, path, eb, i, &key); - BUG_ON(ret); + if (ret) + break; } } btrfs_free_path(path); - return 0; + return ret; } static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, @@ -1989,8 +2147,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, WARN_ON(*level >= BTRFS_MAX_LEVEL); cur = path->nodes[*level]; - if (btrfs_header_level(cur) != *level) - WARN_ON(1); + WARN_ON(btrfs_header_level(cur) != *level); if (path->slots[*level] >= btrfs_header_nritems(cur)) @@ -2009,8 +2166,10 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, if (*level == 1) { ret = wc->process_func(root, next, wc, ptr_gen); - if (ret) + if (ret) { + free_extent_buffer(next); return ret; + } path->slots[*level]++; if (wc->free) { @@ -2020,17 +2179,22 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, return ret; } - btrfs_tree_lock(next); - btrfs_set_lock_blocking(next); - clean_tree_block(trans, root, next); - btrfs_wait_tree_block_writeback(next); - btrfs_tree_unlock(next); + if (trans) { + btrfs_tree_lock(next); + btrfs_set_lock_blocking(next); + clean_tree_block(trans, root, next); + btrfs_wait_tree_block_writeback(next); + btrfs_tree_unlock(next); + } WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); ret = btrfs_free_and_pin_reserved_extent(root, bytenr, blocksize); - BUG_ON(ret); /* -ENOMEM or logic errors */ + if (ret) { + free_extent_buffer(next); + return ret; + } } free_extent_buffer(next); continue; @@ -2093,17 +2257,20 @@ static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans, next = path->nodes[*level]; - btrfs_tree_lock(next); - btrfs_set_lock_blocking(next); - clean_tree_block(trans, root, next); - btrfs_wait_tree_block_writeback(next); - btrfs_tree_unlock(next); + if (trans) { + btrfs_tree_lock(next); + btrfs_set_lock_blocking(next); + clean_tree_block(trans, root, next); + btrfs_wait_tree_block_writeback(next); + btrfs_tree_unlock(next); + } WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID); ret = btrfs_free_and_pin_reserved_extent(root, path->nodes[*level]->start, path->nodes[*level]->len); - BUG_ON(ret); + if (ret) + return ret; } free_extent_buffer(path->nodes[*level]); path->nodes[*level] = NULL; @@ -2125,7 +2292,6 @@ static int walk_log_tree(struct btrfs_trans_handle *trans, int wret; int level; struct btrfs_path *path; - int i; int orig_level; path = btrfs_alloc_path(); @@ -2167,27 +2333,24 @@ static int walk_log_tree(struct btrfs_trans_handle *trans, next = path->nodes[orig_level]; - btrfs_tree_lock(next); - btrfs_set_lock_blocking(next); - clean_tree_block(trans, log, next); - btrfs_wait_tree_block_writeback(next); - btrfs_tree_unlock(next); + if (trans) { + btrfs_tree_lock(next); + btrfs_set_lock_blocking(next); + clean_tree_block(trans, log, next); + btrfs_wait_tree_block_writeback(next); + btrfs_tree_unlock(next); + } WARN_ON(log->root_key.objectid != BTRFS_TREE_LOG_OBJECTID); ret = btrfs_free_and_pin_reserved_extent(log, next->start, next->len); - BUG_ON(ret); /* -ENOMEM or logic errors */ + if (ret) + goto out; } } out: - for (i = 0; i <= orig_level; i++) { - if (path->nodes[i]) { - free_extent_buffer(path->nodes[i]); - path->nodes[i] = NULL; - } - } btrfs_free_path(path); return ret; } @@ -2212,8 +2375,8 @@ static int update_log_root(struct btrfs_trans_handle *trans, return ret; } -static int wait_log_commit(struct btrfs_trans_handle *trans, - struct btrfs_root *root, unsigned long transid) +static void wait_log_commit(struct btrfs_trans_handle *trans, + struct btrfs_root *root, int transid) { DEFINE_WAIT(wait); int index = transid % 2; @@ -2228,36 +2391,63 @@ static int wait_log_commit(struct btrfs_trans_handle *trans, &wait, TASK_UNINTERRUPTIBLE); mutex_unlock(&root->log_mutex); - if (root->fs_info->last_trans_log_full_commit != - trans->transid && root->log_transid < transid + 2 && + if (root->log_transid_committed < transid && atomic_read(&root->log_commit[index])) schedule(); finish_wait(&root->log_commit_wait[index], &wait); mutex_lock(&root->log_mutex); - } while (root->fs_info->last_trans_log_full_commit != - trans->transid && root->log_transid < transid + 2 && + } while (root->log_transid_committed < transid && atomic_read(&root->log_commit[index])); - return 0; } static void wait_for_writer(struct btrfs_trans_handle *trans, struct btrfs_root *root) { DEFINE_WAIT(wait); - while (root->fs_info->last_trans_log_full_commit != - trans->transid && atomic_read(&root->log_writers)) { + + while (atomic_read(&root->log_writers)) { prepare_to_wait(&root->log_writer_wait, &wait, TASK_UNINTERRUPTIBLE); mutex_unlock(&root->log_mutex); - if (root->fs_info->last_trans_log_full_commit != - trans->transid && atomic_read(&root->log_writers)) + if (atomic_read(&root->log_writers)) schedule(); mutex_lock(&root->log_mutex); finish_wait(&root->log_writer_wait, &wait); } } +static inline void btrfs_remove_log_ctx(struct btrfs_root *root, + struct btrfs_log_ctx *ctx) +{ + if (!ctx) + return; + + mutex_lock(&root->log_mutex); + list_del_init(&ctx->list); + mutex_unlock(&root->log_mutex); +} + +/* + * Invoked in log mutex context, or be sure there is no other task which + * can access the list. + */ +static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root, + int index, int error) +{ + struct btrfs_log_ctx *ctx; + + if (!error) { + INIT_LIST_HEAD(&root->log_ctxs[index]); + return; + } + + list_for_each_entry(ctx, &root->log_ctxs[index], list) + ctx->log_ret = error; + + INIT_LIST_HEAD(&root->log_ctxs[index]); +} + /* * btrfs_sync_log does sends a given tree log down to the disk and * updates the super blocks to record it. When this call is done, @@ -2271,7 +2461,7 @@ static void wait_for_writer(struct btrfs_trans_handle *trans, * that has happened. */ int btrfs_sync_log(struct btrfs_trans_handle *trans, - struct btrfs_root *root) + struct btrfs_root *root, struct btrfs_log_ctx *ctx) { int index1; int index2; @@ -2279,25 +2469,35 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, int ret; struct btrfs_root *log = root->log_root; struct btrfs_root *log_root_tree = root->fs_info->log_root_tree; - unsigned long log_transid = 0; + int log_transid = 0; + struct btrfs_log_ctx root_log_ctx; + struct blk_plug plug; mutex_lock(&root->log_mutex); - log_transid = root->log_transid; - index1 = root->log_transid % 2; + log_transid = ctx->log_transid; + if (root->log_transid_committed >= log_transid) { + mutex_unlock(&root->log_mutex); + return ctx->log_ret; + } + + index1 = log_transid % 2; if (atomic_read(&root->log_commit[index1])) { - wait_log_commit(trans, root, root->log_transid); + wait_log_commit(trans, root, log_transid); mutex_unlock(&root->log_mutex); - return 0; + return ctx->log_ret; } + ASSERT(log_transid == root->log_transid); atomic_set(&root->log_commit[index1], 1); /* wait for previous tree log sync to complete */ if (atomic_read(&root->log_commit[(index1 + 1) % 2])) - wait_log_commit(trans, root, root->log_transid - 1); + wait_log_commit(trans, root, log_transid - 1); + while (1) { int batch = atomic_read(&root->log_batch); /* when we're on an ssd, just kick the log commit out */ - if (!btrfs_test_opt(root, SSD) && root->log_multiple_pids) { + if (!btrfs_test_opt(root, SSD) && + test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) { mutex_unlock(&root->log_mutex); schedule_timeout_uninterruptible(1); mutex_lock(&root->log_mutex); @@ -2308,7 +2508,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, } /* bail out if we need to do a full commit */ - if (root->fs_info->last_trans_log_full_commit == trans->transid) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { ret = -EAGAIN; btrfs_free_logged_extents(log, log_transid); mutex_unlock(&root->log_mutex); @@ -2323,10 +2523,13 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, /* we start IO on all the marked extents here, but we don't actually * wait for them until later. */ + blk_start_plug(&plug); ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark); if (ret) { + blk_finish_plug(&plug); btrfs_abort_transaction(trans, root, ret); btrfs_free_logged_extents(log, log_transid); + btrfs_set_log_full_commit(root->fs_info, trans); mutex_unlock(&root->log_mutex); goto out; } @@ -2336,7 +2539,6 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, root->log_transid++; log->log_transid = root->log_transid; root->log_start_pid = 0; - smp_mb(); /* * IO has been started, blocks of the log tree have WRITTEN flag set * in their headers. new modifications of the log will be written to @@ -2344,9 +2546,16 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, */ mutex_unlock(&root->log_mutex); + btrfs_init_log_ctx(&root_log_ctx); + mutex_lock(&log_root_tree->log_mutex); atomic_inc(&log_root_tree->log_batch); atomic_inc(&log_root_tree->log_writers); + + index2 = log_root_tree->log_transid % 2; + list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]); + root_log_ctx.log_transid = log_root_tree->log_transid; + mutex_unlock(&log_root_tree->log_mutex); ret = update_log_root(trans, log); @@ -2359,12 +2568,17 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, } if (ret) { + if (!list_empty(&root_log_ctx.list)) + list_del_init(&root_log_ctx.list); + + blk_finish_plug(&plug); + btrfs_set_log_full_commit(root->fs_info, trans); + if (ret != -ENOSPC) { btrfs_abort_transaction(trans, root, ret); mutex_unlock(&log_root_tree->log_mutex); goto out; } - root->fs_info->last_trans_log_full_commit = trans->transid; btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); @@ -2372,21 +2586,29 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, goto out; } - index2 = log_root_tree->log_transid % 2; + if (log_root_tree->log_transid_committed >= root_log_ctx.log_transid) { + mutex_unlock(&log_root_tree->log_mutex); + ret = root_log_ctx.log_ret; + goto out; + } + + index2 = root_log_ctx.log_transid % 2; if (atomic_read(&log_root_tree->log_commit[index2])) { + blk_finish_plug(&plug); btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); wait_log_commit(trans, log_root_tree, - log_root_tree->log_transid); + root_log_ctx.log_transid); btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); - ret = 0; + ret = root_log_ctx.log_ret; goto out; } + ASSERT(root_log_ctx.log_transid == log_root_tree->log_transid); atomic_set(&log_root_tree->log_commit[index2], 1); if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) { wait_log_commit(trans, log_root_tree, - log_root_tree->log_transid - 1); + root_log_ctx.log_transid - 1); } wait_for_writer(trans, log_root_tree); @@ -2395,7 +2617,8 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, * now that we've moved on to the tree of log tree roots, * check the full commit flag again */ - if (root->fs_info->last_trans_log_full_commit == trans->transid) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { + blk_finish_plug(&plug); btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); @@ -2403,16 +2626,21 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, goto out_wake_log_root; } - ret = btrfs_write_and_wait_marked_extents(log_root_tree, - &log_root_tree->dirty_log_pages, - EXTENT_DIRTY | EXTENT_NEW); + ret = btrfs_write_marked_extents(log_root_tree, + &log_root_tree->dirty_log_pages, + EXTENT_DIRTY | EXTENT_NEW); + blk_finish_plug(&plug); if (ret) { + btrfs_set_log_full_commit(root->fs_info, trans); btrfs_abort_transaction(trans, root, ret); btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); goto out_wake_log_root; } btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); + btrfs_wait_marked_extents(log_root_tree, + &log_root_tree->dirty_log_pages, + EXTENT_NEW | EXTENT_DIRTY); btrfs_wait_logged_extents(log, log_transid); btrfs_set_super_log_root(root->fs_info->super_for_commit, @@ -2421,8 +2649,6 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, btrfs_header_level(log_root_tree->node)); log_root_tree->log_transid++; - smp_mb(); - mutex_unlock(&log_root_tree->log_mutex); /* @@ -2432,10 +2658,9 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, * the running transaction open, so a full commit can't hop * in and cause problems either. */ - btrfs_scrub_pause_super(root); ret = write_ctree_super(trans, root->fs_info->tree_root, 1); - btrfs_scrub_continue_super(root); if (ret) { + btrfs_set_log_full_commit(root->fs_info, trans); btrfs_abort_transaction(trans, root, ret); goto out_wake_log_root; } @@ -2446,13 +2671,28 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, mutex_unlock(&root->log_mutex); out_wake_log_root: + /* + * We needn't get log_mutex here because we are sure all + * the other tasks are blocked. + */ + btrfs_remove_all_log_ctxs(log_root_tree, index2, ret); + + mutex_lock(&log_root_tree->log_mutex); + log_root_tree->log_transid_committed++; atomic_set(&log_root_tree->log_commit[index2], 0); - smp_mb(); + mutex_unlock(&log_root_tree->log_mutex); + if (waitqueue_active(&log_root_tree->log_commit_wait[index2])) wake_up(&log_root_tree->log_commit_wait[index2]); out: + /* See above. */ + btrfs_remove_all_log_ctxs(root, index1, ret); + + mutex_lock(&root->log_mutex); + root->log_transid_committed++; atomic_set(&root->log_commit[index1], 0); - smp_mb(); + mutex_unlock(&root->log_mutex); + if (waitqueue_active(&root->log_commit_wait[index1])) wake_up(&root->log_commit_wait[index1]); return ret; @@ -2469,10 +2709,10 @@ static void free_log_tree(struct btrfs_trans_handle *trans, .process_func = process_one_buffer }; - if (trans) { - ret = walk_log_tree(trans, log, &wc); - BUG_ON(ret); - } + ret = walk_log_tree(trans, log, &wc); + /* I don't think this can happen but just in case */ + if (ret) + btrfs_abort_transaction(trans, log, ret); while (1) { ret = find_first_extent_bit(&log->dirty_log_pages, @@ -2579,7 +2819,10 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, if (di) { ret = btrfs_delete_one_dir_name(trans, log, path, di); bytes_del += name_len; - BUG_ON(ret); + if (ret) { + err = ret; + goto fail; + } } btrfs_release_path(path); di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino, @@ -2591,7 +2834,10 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, if (di) { ret = btrfs_delete_one_dir_name(trans, log, path, di); bytes_del += name_len; - BUG_ON(ret); + if (ret) { + err = ret; + goto fail; + } } /* update the directory size in the log to reflect the names @@ -2632,7 +2878,7 @@ fail: out_unlock: mutex_unlock(&BTRFS_I(dir)->log_mutex); if (ret == -ENOSPC) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 0; } else if (ret < 0) btrfs_abort_transaction(trans, root, ret); @@ -2665,7 +2911,7 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans, dirid, &index); mutex_unlock(&BTRFS_I(inode)->log_mutex); if (ret == -ENOSPC) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 0; } else if (ret < 0 && ret != -ENOENT) btrfs_abort_transaction(trans, root, ret); @@ -2719,7 +2965,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, u64 min_offset, u64 *last_offset_ret) { struct btrfs_key min_key; - struct btrfs_key max_key; struct btrfs_root *log = root->log_root; struct extent_buffer *src; int err = 0; @@ -2731,9 +2976,6 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, u64 ino = btrfs_ino(inode); log = root->log_root; - max_key.objectid = ino; - max_key.offset = (u64)-1; - max_key.type = key_type; min_key.objectid = ino; min_key.type = key_type; @@ -2741,8 +2983,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, path->keep_locks = 1; - ret = btrfs_search_forward(root, &min_key, &max_key, - path, trans->transid); + ret = btrfs_search_forward(root, &min_key, path, trans->transid); /* * we didn't find anything from this transaction, see if there @@ -2795,10 +3036,8 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, /* find the first key from this transaction again */ ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); - if (ret != 0) { - WARN_ON(1); + if (WARN_ON(ret != 0)) goto done; - } /* * we have a block from this transaction, log every item in it @@ -2930,7 +3169,7 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans, while (1) { ret = btrfs_search_slot(trans, log, &key, path, -1, 1); - BUG_ON(ret == 0); + BUG_ON(ret == 0); /* Logic error */ if (ret < 0) break; @@ -3024,11 +3263,10 @@ static int log_inode_item(struct btrfs_trans_handle *trans, struct inode *inode) { struct btrfs_inode_item *inode_item; - struct btrfs_key key; int ret; - memcpy(&key, &BTRFS_I(inode)->location, sizeof(key)); - ret = btrfs_insert_empty_item(trans, log, path, &key, + ret = btrfs_insert_empty_item(trans, log, path, + &BTRFS_I(inode)->location, sizeof(*inode_item)); if (ret && ret != -EEXIST) return ret; @@ -3042,7 +3280,7 @@ static int log_inode_item(struct btrfs_trans_handle *trans, static noinline int copy_items(struct btrfs_trans_handle *trans, struct inode *inode, struct btrfs_path *dst_path, - struct extent_buffer *src, + struct btrfs_path *src_path, u64 *last_extent, int start_slot, int nr, int inode_only) { unsigned long src_offset; @@ -3050,6 +3288,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, struct btrfs_root *log = BTRFS_I(inode)->root->log_root; struct btrfs_file_extent_item *extent; struct btrfs_inode_item *inode_item; + struct extent_buffer *src = src_path->nodes[0]; + struct btrfs_key first_key, last_key, key; int ret; struct btrfs_key *ins_keys; u32 *ins_sizes; @@ -3057,6 +3297,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, int i; struct list_head ordered_sums; int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; + bool has_extents = false; + bool need_find_last_extent = (*last_extent == 0); + bool done = false; INIT_LIST_HEAD(&ordered_sums); @@ -3065,6 +3308,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, if (!ins_data) return -ENOMEM; + first_key.objectid = (u64)-1; + ins_sizes = (u32 *)ins_data; ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32)); @@ -3085,6 +3330,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset = btrfs_item_ptr_offset(src, start_slot + i); + if ((i == (nr - 1))) + last_key = ins_keys[i]; + if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) { inode_item = btrfs_item_ptr(dst_path->nodes[0], dst_path->slots[0], @@ -3096,6 +3344,21 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, src_offset, ins_sizes[i]); } + /* + * We set need_find_last_extent here in case we know we were + * processing other items and then walk into the first extent in + * the inode. If we don't hit an extent then nothing changes, + * we'll do the last search the next time around. + */ + if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY) { + has_extents = true; + if (need_find_last_extent && + first_key.objectid == (u64)-1) + first_key = ins_keys[i]; + } else { + need_find_last_extent = false; + } + /* take a reference on file data extents so that truncates * or deletes of this inode don't have to relog the inode * again @@ -3133,7 +3396,11 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, log->fs_info->csum_root, ds + cs, ds + cs + cl - 1, &ordered_sums, 0); - BUG_ON(ret); + if (ret) { + btrfs_release_path(dst_path); + kfree(ins_data); + return ret; + } } } } @@ -3156,135 +3423,149 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, list_del(&sums->list); kfree(sums); } - return ret; -} -static int extent_cmp(void *priv, struct list_head *a, struct list_head *b) -{ - struct extent_map *em1, *em2; - - em1 = list_entry(a, struct extent_map, list); - em2 = list_entry(b, struct extent_map, list); + if (!has_extents) + return ret; - if (em1->start < em2->start) - return -1; - else if (em1->start > em2->start) - return 1; - return 0; -} + /* + * Because we use btrfs_search_forward we could skip leaves that were + * not modified and then assume *last_extent is valid when it really + * isn't. So back up to the previous leaf and read the end of the last + * extent before we go and fill in holes. + */ + if (need_find_last_extent) { + u64 len; -static int drop_adjacent_extents(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct inode *inode, - struct extent_map *em, - struct btrfs_path *path) -{ - struct btrfs_file_extent_item *fi; - struct extent_buffer *leaf; - struct btrfs_key key, new_key; - struct btrfs_map_token token; - u64 extent_end; - u64 extent_offset = 0; - int extent_type; - int del_slot = 0; - int del_nr = 0; - int ret = 0; + ret = btrfs_prev_leaf(BTRFS_I(inode)->root, src_path); + if (ret < 0) + return ret; + if (ret) + goto fill_holes; + if (src_path->slots[0]) + src_path->slots[0]--; + src = src_path->nodes[0]; + btrfs_item_key_to_cpu(src, &key, src_path->slots[0]); + if (key.objectid != btrfs_ino(inode) || + key.type != BTRFS_EXTENT_DATA_KEY) + goto fill_holes; + extent = btrfs_item_ptr(src, src_path->slots[0], + struct btrfs_file_extent_item); + if (btrfs_file_extent_type(src, extent) == + BTRFS_FILE_EXTENT_INLINE) { + len = btrfs_file_extent_inline_len(src, + src_path->slots[0], + extent); + *last_extent = ALIGN(key.offset + len, + log->sectorsize); + } else { + len = btrfs_file_extent_num_bytes(src, extent); + *last_extent = key.offset + len; + } + } +fill_holes: + /* So we did prev_leaf, now we need to move to the next leaf, but a few + * things could have happened + * + * 1) A merge could have happened, so we could currently be on a leaf + * that holds what we were copying in the first place. + * 2) A split could have happened, and now not all of the items we want + * are on the same leaf. + * + * So we need to adjust how we search for holes, we need to drop the + * path and re-search for the first extent key we found, and then walk + * forward until we hit the last one we copied. + */ + if (need_find_last_extent) { + /* btrfs_prev_leaf could return 1 without releasing the path */ + btrfs_release_path(src_path); + ret = btrfs_search_slot(NULL, BTRFS_I(inode)->root, &first_key, + src_path, 0, 0); + if (ret < 0) + return ret; + ASSERT(ret == 0); + src = src_path->nodes[0]; + i = src_path->slots[0]; + } else { + i = start_slot; + } - while (1) { - btrfs_init_map_token(&token); - leaf = path->nodes[0]; - path->slots[0]++; - if (path->slots[0] >= btrfs_header_nritems(leaf)) { - if (del_nr) { - ret = btrfs_del_items(trans, root, path, - del_slot, del_nr); - if (ret) - return ret; - del_nr = 0; - } + /* + * Ok so here we need to go through and fill in any holes we may have + * to make sure that holes are punched for those areas in case they had + * extents previously. + */ + while (!done) { + u64 offset, len; + u64 extent_end; - ret = btrfs_next_leaf_write(trans, root, path, 1); + if (i >= btrfs_header_nritems(src_path->nodes[0])) { + ret = btrfs_next_leaf(BTRFS_I(inode)->root, src_path); if (ret < 0) return ret; - if (ret > 0) - return 0; - leaf = path->nodes[0]; + ASSERT(ret == 0); + src = src_path->nodes[0]; + i = 0; } - btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); + btrfs_item_key_to_cpu(src, &key, i); + if (!btrfs_comp_cpu_keys(&key, &last_key)) + done = true; if (key.objectid != btrfs_ino(inode) || - key.type != BTRFS_EXTENT_DATA_KEY || - key.offset >= em->start + em->len) - break; - - fi = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - extent_type = btrfs_token_file_extent_type(leaf, fi, &token); - if (extent_type == BTRFS_FILE_EXTENT_REG || - extent_type == BTRFS_FILE_EXTENT_PREALLOC) { - extent_offset = btrfs_token_file_extent_offset(leaf, - fi, &token); - extent_end = key.offset + - btrfs_token_file_extent_num_bytes(leaf, fi, - &token); - } else if (extent_type == BTRFS_FILE_EXTENT_INLINE) { - extent_end = key.offset + - btrfs_file_extent_inline_len(leaf, fi); + key.type != BTRFS_EXTENT_DATA_KEY) { + i++; + continue; + } + extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item); + if (btrfs_file_extent_type(src, extent) == + BTRFS_FILE_EXTENT_INLINE) { + len = btrfs_file_extent_inline_len(src, i, extent); + extent_end = ALIGN(key.offset + len, log->sectorsize); } else { - BUG(); + len = btrfs_file_extent_num_bytes(src, extent); + extent_end = key.offset + len; } + i++; - if (extent_end <= em->len + em->start) { - if (!del_nr) { - del_slot = path->slots[0]; - } - del_nr++; + if (*last_extent == key.offset) { + *last_extent = extent_end; continue; } - - /* - * Ok so we'll ignore previous items if we log a new extent, - * which can lead to overlapping extents, so if we have an - * existing extent we want to adjust we _have_ to check the next - * guy to make sure we even need this extent anymore, this keeps - * us from panicing in set_item_key_safe. - */ - if (path->slots[0] < btrfs_header_nritems(leaf) - 1) { - struct btrfs_key tmp_key; - - btrfs_item_key_to_cpu(leaf, &tmp_key, - path->slots[0] + 1); - if (tmp_key.objectid == btrfs_ino(inode) && - tmp_key.type == BTRFS_EXTENT_DATA_KEY && - tmp_key.offset <= em->start + em->len) { - if (!del_nr) - del_slot = path->slots[0]; - del_nr++; - continue; - } - } - - BUG_ON(extent_type == BTRFS_FILE_EXTENT_INLINE); - memcpy(&new_key, &key, sizeof(new_key)); - new_key.offset = em->start + em->len; - btrfs_set_item_key_safe(trans, root, path, &new_key); - extent_offset += em->start + em->len - key.offset; - btrfs_set_token_file_extent_offset(leaf, fi, extent_offset, - &token); - btrfs_set_token_file_extent_num_bytes(leaf, fi, extent_end - - (em->start + em->len), - &token); - btrfs_mark_buffer_dirty(leaf); + offset = *last_extent; + len = key.offset - *last_extent; + ret = btrfs_insert_file_extent(trans, log, btrfs_ino(inode), + offset, 0, 0, len, 0, len, 0, + 0, 0); + if (ret) + break; + *last_extent = offset + len; } + /* + * Need to let the callers know we dropped the path so they should + * re-search. + */ + if (!ret && need_find_last_extent) + ret = 1; + return ret; +} - if (del_nr) - ret = btrfs_del_items(trans, root, path, del_slot, del_nr); +static int extent_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct extent_map *em1, *em2; - return ret; + em1 = list_entry(a, struct extent_map, list); + em2 = list_entry(b, struct extent_map, list); + + if (em1->start < em2->start) + return -1; + else if (em1->start > em2->start) + return 1; + return 0; } static int log_one_extent(struct btrfs_trans_handle *trans, struct inode *inode, struct btrfs_root *root, - struct extent_map *em, struct btrfs_path *path) + struct extent_map *em, struct btrfs_path *path, + struct list_head *logged_list) { struct btrfs_root *log = root->log_root; struct btrfs_file_extent_item *fi; @@ -3300,42 +3581,32 @@ static int log_one_extent(struct btrfs_trans_handle *trans, u64 extent_offset = em->start - em->orig_start; u64 block_len; int ret; - int index = log->log_transid % 2; bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; + int extent_inserted = 0; -insert: INIT_LIST_HEAD(&ordered_sums); btrfs_init_map_token(&token); - key.objectid = btrfs_ino(inode); - key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = em->start; - path->really_keep_locks = 1; - - ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*fi)); - if (ret && ret != -EEXIST) { - path->really_keep_locks = 0; + + ret = __btrfs_drop_extents(trans, log, inode, path, em->start, + em->start + em->len, NULL, 0, 1, + sizeof(*fi), &extent_inserted); + if (ret) return ret; + + if (!extent_inserted) { + key.objectid = btrfs_ino(inode); + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = em->start; + + ret = btrfs_insert_empty_item(trans, log, path, &key, + sizeof(*fi)); + if (ret) + return ret; } leaf = path->nodes[0]; fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); - /* - * If we are overwriting an inline extent with a real one then we need - * to just delete the inline extent as it may not be large enough to - * have the entire file_extent_item. - */ - if (ret && btrfs_token_file_extent_type(leaf, fi, &token) == - BTRFS_FILE_EXTENT_INLINE) { - ret = btrfs_del_item(trans, log, path); - btrfs_release_path(path); - if (ret) { - path->really_keep_locks = 0; - return ret; - } - goto insert; - } - btrfs_set_token_file_extent_generation(leaf, fi, em->generation, &token); if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) { @@ -3347,7 +3618,7 @@ insert: btrfs_set_token_file_extent_type(leaf, fi, BTRFS_FILE_EXTENT_REG, &token); - if (em->block_start == 0) + if (em->block_start == EXTENT_MAP_HOLE) skip_csum = true; } @@ -3374,22 +3645,14 @@ insert: em->start - em->orig_start, &token); btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token); - btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->len, &token); + btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token); btrfs_set_token_file_extent_compression(leaf, fi, em->compress_type, &token); btrfs_set_token_file_extent_encryption(leaf, fi, 0, &token); btrfs_set_token_file_extent_other_encoding(leaf, fi, 0, &token); btrfs_mark_buffer_dirty(leaf); - /* - * Have to check the extent to the right of us to make sure it doesn't - * fall in our current range. We're ok if the previous extent is in our - * range since the recovery stuff will run us in key order and thus just - * drop the part we overwrote. - */ - ret = drop_adjacent_extents(trans, log, inode, em, path); btrfs_release_path(path); - path->really_keep_locks = 0; if (ret) { return ret; } @@ -3397,26 +3660,16 @@ insert: if (skip_csum) return 0; - if (em->compress_type) { - csum_offset = 0; - csum_len = block_len; - } - /* * First check and see if our csums are on our outstanding ordered * extents. */ -again: - spin_lock_irq(&log->log_extents_lock[index]); - list_for_each_entry(ordered, &log->logged_list[index], log_list) { + list_for_each_entry(ordered, logged_list, log_list) { struct btrfs_ordered_sum *sum; if (!mod_len) break; - if (ordered->inode != inode) - continue; - if (ordered->file_offset + ordered->len <= mod_start || mod_start + mod_len <= ordered->file_offset) continue; @@ -3459,34 +3712,32 @@ again: if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM, &ordered->flags)) continue; - atomic_inc(&ordered->refs); - spin_unlock_irq(&log->log_extents_lock[index]); - /* - * we've dropped the lock, we must either break or - * start over after this. - */ - wait_event(ordered->wait, ordered->csum_bytes_left == 0); + if (ordered->csum_bytes_left) { + btrfs_start_ordered_extent(inode, ordered, 0); + wait_event(ordered->wait, + ordered->csum_bytes_left == 0); + } list_for_each_entry(sum, &ordered->list, list) { ret = btrfs_csum_file_blocks(trans, log, sum); - if (ret) { - btrfs_put_ordered_extent(ordered); + if (ret) goto unlocked; - } } - btrfs_put_ordered_extent(ordered); - goto again; } - spin_unlock_irq(&log->log_extents_lock[index]); unlocked: if (!mod_len || ret) return ret; - csum_offset = mod_start - em->start; - csum_len = mod_len; + if (em->compress_type) { + csum_offset = 0; + csum_len = block_len; + } else { + csum_offset = mod_start - em->start; + csum_len = mod_len; + } /* block start is already adjusted for the file extent offset. */ ret = btrfs_lookup_csums_range(log->fs_info->csum_root, @@ -3512,7 +3763,8 @@ unlocked: static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode, - struct btrfs_path *path) + struct btrfs_path *path, + struct list_head *logged_list) { struct extent_map *em, *n; struct list_head extents; @@ -3570,7 +3822,7 @@ process: write_unlock(&tree->lock); - ret = log_one_extent(trans, inode, root, em, path); + ret = log_one_extent(trans, inode, root, em, path, logged_list); write_lock(&tree->lock); clear_em_logging(tree, em); free_extent_map(em); @@ -3606,6 +3858,8 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, struct btrfs_key max_key; struct btrfs_root *log = root->log_root; struct extent_buffer *src = NULL; + LIST_HEAD(logged_list); + u64 last_extent = 0; int err = 0; int ret; int nritems; @@ -3614,8 +3868,6 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, bool fast_search = false; u64 ino = btrfs_ino(inode); - log = root->log_root; - path = btrfs_alloc_path(); if (!path) return -ENOMEM; @@ -3655,7 +3907,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, mutex_lock(&BTRFS_I(inode)->log_mutex); - btrfs_get_logged_extents(log, inode); + btrfs_get_logged_extents(inode, &logged_list); /* * a brute force approach to making sure we get the most uptodate @@ -3675,7 +3927,8 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, ret = btrfs_truncate_inode_items(trans, log, inode, 0, 0); } else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING, - &BTRFS_I(inode)->runtime_flags)) { + &BTRFS_I(inode)->runtime_flags) || + inode_only == LOG_INODE_EXISTS) { if (inode_only == LOG_INODE_ALL) fast_search = true; max_key.type = BTRFS_XATTR_ITEM_KEY; @@ -3701,7 +3954,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, while (1) { ins_nr = 0; - ret = btrfs_search_forward(root, &min_key, &max_key, + ret = btrfs_search_forward(root, &min_key, path, trans->transid); if (ret != 0) break; @@ -3722,11 +3975,15 @@ again: goto next_slot; } - ret = copy_items(trans, inode, dst_path, src, ins_start_slot, - ins_nr, inode_only); - if (ret) { + ret = copy_items(trans, inode, dst_path, path, &last_extent, + ins_start_slot, ins_nr, inode_only); + if (ret < 0) { err = ret; goto out_unlock; + } if (ret) { + ins_nr = 0; + btrfs_release_path(path); + continue; } ins_nr = 1; ins_start_slot = path->slots[0]; @@ -3740,45 +3997,49 @@ next_slot: goto again; } if (ins_nr) { - ret = copy_items(trans, inode, dst_path, src, - ins_start_slot, + ret = copy_items(trans, inode, dst_path, path, + &last_extent, ins_start_slot, ins_nr, inode_only); - if (ret) { + if (ret < 0) { err = ret; goto out_unlock; } + ret = 0; ins_nr = 0; } btrfs_release_path(path); - if (min_key.offset < (u64)-1) + if (min_key.offset < (u64)-1) { min_key.offset++; - else if (min_key.type < (u8)-1) + } else if (min_key.type < max_key.type) { min_key.type++; - else if (min_key.objectid < (u64)-1) - min_key.objectid++; - else + min_key.offset = 0; + } else { break; + } } if (ins_nr) { - ret = copy_items(trans, inode, dst_path, src, ins_start_slot, - ins_nr, inode_only); - if (ret) { + ret = copy_items(trans, inode, dst_path, path, &last_extent, + ins_start_slot, ins_nr, inode_only); + if (ret < 0) { err = ret; goto out_unlock; } + ret = 0; ins_nr = 0; } log_extents: + btrfs_release_path(path); + btrfs_release_path(dst_path); if (fast_search) { - btrfs_release_path(dst_path); - ret = btrfs_log_changed_extents(trans, root, inode, dst_path); + ret = btrfs_log_changed_extents(trans, root, inode, dst_path, + &logged_list); if (ret) { err = ret; goto out_unlock; } - } else { + } else if (inode_only == LOG_INODE_ALL) { struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree; struct extent_map *em, *n; @@ -3789,8 +4050,6 @@ log_extents: } if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) { - btrfs_release_path(path); - btrfs_release_path(dst_path); ret = log_directory_changes(trans, root, inode, path, dst_path); if (ret) { err = ret; @@ -3800,8 +4059,10 @@ log_extents: BTRFS_I(inode)->logged_trans = trans->transid; BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans; out_unlock: - if (err) - btrfs_free_logged_extents(log, log->log_transid); + if (unlikely(err)) + btrfs_put_logged_extents(&logged_list); + else + btrfs_submit_logged_extents(&logged_list, log); mutex_unlock(&BTRFS_I(inode)->log_mutex); btrfs_free_path(path); @@ -3824,6 +4085,7 @@ static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans, int ret = 0; struct btrfs_root *root; struct dentry *old_parent = NULL; + struct inode *orig_inode = inode; /* * for regular files, if its inode is already on disk, we don't @@ -3843,7 +4105,14 @@ static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans, } while (1) { - BTRFS_I(inode)->logged_trans = trans->transid; + /* + * If we are logging a directory then we start with our inode, + * not our parents inode, so we need to skipp setting the + * logged_trans so that further down in the log code we don't + * think this inode has already been logged. + */ + if (inode != orig_inode) + BTRFS_I(inode)->logged_trans = trans->transid; smp_mb(); if (BTRFS_I(inode)->last_unlink_trans > last_committed) { @@ -3853,8 +4122,7 @@ static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans, * make sure any commits to the log are forced * to be full commits */ - root->fs_info->last_trans_log_full_commit = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 1; break; } @@ -3882,9 +4150,10 @@ out: * only logging is done of any parent directories that are older than * the last committed transaction */ -int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct inode *inode, - struct dentry *parent, int exists_only) +static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct inode *inode, + struct dentry *parent, int exists_only, + struct btrfs_log_ctx *ctx) { int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL; struct super_block *sb; @@ -3899,6 +4168,10 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, goto end_no_trans; } + /* + * The prev transaction commit doesn't complete, we need do + * full commit by ourselves. + */ if (root->fs_info->last_trans_log_full_commit > root->fs_info->last_trans_committed) { ret = 1; @@ -3921,9 +4194,9 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, goto end_no_trans; } - ret = start_log_trans(trans, root); + ret = start_log_trans(trans, root, ctx); if (ret) - goto end_trans; + goto end_no_trans; ret = btrfs_log_inode(trans, root, inode, inode_only); if (ret) @@ -3968,9 +4241,12 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, end_trans: dput(old_parent); if (ret < 0) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 1; } + + if (ret) + btrfs_remove_log_ctx(root, ctx); btrfs_end_log_trans(root); end_no_trans: return ret; @@ -3983,12 +4259,14 @@ end_no_trans: * data on disk. */ int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct dentry *dentry) + struct btrfs_root *root, struct dentry *dentry, + struct btrfs_log_ctx *ctx) { struct dentry *parent = dget_parent(dentry); int ret; - ret = btrfs_log_inode_parent(trans, root, dentry->d_inode, parent, 0); + ret = btrfs_log_inode_parent(trans, root, dentry->d_inode, parent, + 0, ctx); dput(parent); return ret; @@ -4059,8 +4337,7 @@ again: if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID) break; - log = btrfs_read_fs_root_no_radix(log_root_tree, - &found_key); + log = btrfs_read_fs_root(log_root_tree, &found_key); if (IS_ERR(log)) { ret = PTR_ERR(log); btrfs_error(fs_info, ret, @@ -4075,6 +4352,9 @@ again: wc.replay_dest = btrfs_read_fs_root_no_name(fs_info, &tmp_key); if (IS_ERR(wc.replay_dest)) { ret = PTR_ERR(wc.replay_dest); + free_extent_buffer(log->node); + free_extent_buffer(log->commit_root); + kfree(log); btrfs_error(fs_info, ret, "Couldn't read target root " "for tree log recovery."); goto error; @@ -4083,12 +4363,10 @@ again: wc.replay_dest->log_root = log; btrfs_record_root_in_trans(trans, wc.replay_dest); ret = walk_log_tree(trans, log, &wc); - BUG_ON(ret); - if (wc.stage == LOG_WALK_REPLAY_ALL) { + if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) { ret = fixup_inode_link_counts(trans, wc.replay_dest, path); - BUG_ON(ret); } key.offset = found_key.offset - 1; @@ -4097,6 +4375,9 @@ again: free_extent_buffer(log->commit_root); kfree(log); + if (ret) + goto error; + if (found_key.offset == 0) break; } @@ -4117,17 +4398,20 @@ again: btrfs_free_path(path); + /* step 4: commit the transaction, which also unpins the blocks */ + ret = btrfs_commit_transaction(trans, fs_info->tree_root); + if (ret) + return ret; + free_extent_buffer(log_root_tree->node); log_root_tree->log_root = NULL; fs_info->log_root_recovering = 0; - - /* step 4: commit the transaction, which also unpins the blocks */ - btrfs_commit_transaction(trans, fs_info->tree_root); - kfree(log_root_tree); - return 0; + return 0; error: + if (wc.trans) + btrfs_end_transaction(wc.trans, fs_info->tree_root); btrfs_free_path(path); return ret; } @@ -4219,6 +4503,6 @@ int btrfs_log_new_name(struct btrfs_trans_handle *trans, root->fs_info->last_trans_committed)) return 0; - return btrfs_log_inode_parent(trans, root, inode, parent, 1); + return btrfs_log_inode_parent(trans, root, inode, parent, 1, NULL); } |