struct kmem_cache *btrfs_delayed_ref_head_cachep; struct kmem_cache *btrfs_delayed_ref_node_cachep; struct kmem_cache *btrfs_delayed_extent_op_cachep; /* * delayed back reference update tracking. For subvolume trees * we queue up extent allocations and backref maintenance for * delayed processing. This avoids deep call chains where we * add extents in the middle of btrfs_search_slot, and it allows * us to buffer up frequently modified backrefs in an rb tree instead * of hammering updates on the extent allocation tree.
*/
/* * Since the global reserve is just kind of magic we don't really want * to rely on it to save our bacon, so if our size is more than the * delayed_refs_rsv and the global rsv then it's time to think about * bailing.
*/
spin_lock(&delayed_refs_rsv->lock);
reserved += delayed_refs_rsv->reserved; if (delayed_refs_rsv->size >= reserved)
ret = true;
spin_unlock(&delayed_refs_rsv->lock); return ret;
}
/* * Release a ref head's reservation. * * @fs_info: the filesystem * @nr_refs: number of delayed refs to drop * @nr_csums: number of csum items to drop * * Drops the delayed ref head's count from the delayed refs rsv and free any * excess reservation we had.
*/ void btrfs_delayed_refs_rsv_release(struct btrfs_fs_info *fs_info, int nr_refs, int nr_csums)
{ struct btrfs_block_rsv *block_rsv = &fs_info->delayed_refs_rsv;
u64 num_bytes;
u64 released;
released = btrfs_block_rsv_release(fs_info, block_rsv, num_bytes, NULL); if (released)
trace_btrfs_space_reservation(fs_info, "delayed_refs_rsv",
0, released, 0);
}
/* * Adjust the size of the delayed refs rsv. * * This is to be called anytime we may have adjusted trans->delayed_ref_updates * or trans->delayed_ref_csum_deletions, it'll calculate the additional size and * add it to the delayed_refs_rsv.
*/ void btrfs_update_delayed_refs_rsv(struct btrfs_trans_handle *trans)
{ struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_refs_rsv; struct btrfs_block_rsv *local_rsv = &trans->delayed_rsv;
u64 num_bytes;
u64 reserved_bytes;
/* * Try to take num_bytes from the transaction's local delayed reserve. * If not possible, try to take as much as it's available. If the local * reserve doesn't have enough reserved space, the delayed refs reserve * will be refilled next time btrfs_delayed_refs_rsv_refill() is called * by someone or if a transaction commit is triggered before that, the * global block reserve will be used. We want to minimize using the * global block reserve for cases we can account for in advance, to * avoid exhausting it and reach -ENOSPC during a transaction commit.
*/
spin_lock(&local_rsv->lock);
reserved_bytes = min(num_bytes, local_rsv->reserved);
local_rsv->reserved -= reserved_bytes;
local_rsv->full = (local_rsv->reserved >= local_rsv->size);
spin_unlock(&local_rsv->lock);
/* * Adjust the size of the delayed refs block reserve for 1 block group item * insertion, used after allocating a block group.
*/ void btrfs_inc_delayed_refs_rsv_bg_inserts(struct btrfs_fs_info *fs_info)
{ struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_refs_rsv;
spin_lock(&delayed_rsv->lock); /* * Inserting a block group item does not require changing the free space * tree, only the extent tree or the block group tree, so this is all we * need.
*/
delayed_rsv->size += btrfs_calc_insert_metadata_size(fs_info, 1);
delayed_rsv->full = false;
spin_unlock(&delayed_rsv->lock);
}
/* * Adjust the size of the delayed refs block reserve to release space for 1 * block group item insertion.
*/ void btrfs_dec_delayed_refs_rsv_bg_inserts(struct btrfs_fs_info *fs_info)
{ struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_refs_rsv; const u64 num_bytes = btrfs_calc_insert_metadata_size(fs_info, 1);
u64 released;
released = btrfs_block_rsv_release(fs_info, delayed_rsv, num_bytes, NULL); if (released > 0)
trace_btrfs_space_reservation(fs_info, "delayed_refs_rsv",
0, released, 0);
}
/* * Adjust the size of the delayed refs block reserve for 1 block group item * update.
*/ void btrfs_inc_delayed_refs_rsv_bg_updates(struct btrfs_fs_info *fs_info)
{ struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_refs_rsv;
spin_lock(&delayed_rsv->lock); /* * Updating a block group item does not result in new nodes/leaves and * does not require changing the free space tree, only the extent tree * or the block group tree, so this is all we need.
*/
delayed_rsv->size += btrfs_calc_metadata_size(fs_info, 1);
delayed_rsv->full = false;
spin_unlock(&delayed_rsv->lock);
}
/* * Adjust the size of the delayed refs block reserve to release space for 1 * block group item update.
*/ void btrfs_dec_delayed_refs_rsv_bg_updates(struct btrfs_fs_info *fs_info)
{ struct btrfs_block_rsv *delayed_rsv = &fs_info->delayed_refs_rsv; const u64 num_bytes = btrfs_calc_metadata_size(fs_info, 1);
u64 released;
released = btrfs_block_rsv_release(fs_info, delayed_rsv, num_bytes, NULL); if (released > 0)
trace_btrfs_space_reservation(fs_info, "delayed_refs_rsv",
0, released, 0);
}
/* * Refill based on our delayed refs usage. * * @fs_info: the filesystem * @flush: control how we can flush for this reservation. * * This will refill the delayed block_rsv up to 1 items size worth of space and * will return -ENOSPC if we can't make the reservation.
*/ int btrfs_delayed_refs_rsv_refill(struct btrfs_fs_info *fs_info, enum btrfs_reserve_flush_enum flush)
{ struct btrfs_block_rsv *block_rsv = &fs_info->delayed_refs_rsv; struct btrfs_space_info *space_info = block_rsv->space_info;
u64 limit = btrfs_calc_delayed_ref_bytes(fs_info, 1);
u64 num_bytes = 0;
u64 refilled_bytes;
u64 to_free; int ret = -ENOSPC;
ret = btrfs_reserve_metadata_bytes(fs_info, space_info, num_bytes, flush); if (ret) return ret;
/* * We may have raced with someone else, so check again if we the block * reserve is still not full and release any excess space.
*/
spin_lock(&block_rsv->lock); if (block_rsv->reserved < block_rsv->size) {
u64 needed = block_rsv->size - block_rsv->reserved;
/* * compare two delayed data backrefs with same bytenr and type
*/ staticint comp_data_refs(conststruct btrfs_delayed_ref_node *ref1, conststruct btrfs_delayed_ref_node *ref2)
{ if (ref1->data_ref.objectid < ref2->data_ref.objectid) return -1; if (ref1->data_ref.objectid > ref2->data_ref.objectid) return 1; if (ref1->data_ref.offset < ref2->data_ref.offset) return -1; if (ref1->data_ref.offset > ref2->data_ref.offset) return 1; return 0;
}
staticint comp_refs(conststruct btrfs_delayed_ref_node *ref1, conststruct btrfs_delayed_ref_node *ref2, bool check_seq)
{ int ret = 0;
if (ref1->type < ref2->type) return -1; if (ref1->type > ref2->type) return 1; if (ref1->type == BTRFS_SHARED_BLOCK_REF_KEY ||
ref1->type == BTRFS_SHARED_DATA_REF_KEY) { if (ref1->parent < ref2->parent) return -1; if (ref1->parent > ref2->parent) return 1;
} else { if (ref1->ref_root < ref2->ref_root) return -1; if (ref1->ref_root > ref2->ref_root) return 1; if (ref1->type == BTRFS_EXTENT_DATA_REF_KEY)
ret = comp_data_refs(ref1, ref2);
} if (ret) return ret; if (check_seq) { if (ref1->seq < ref2->seq) return -1; if (ref1->seq > ref2->seq) return 1;
} return 0;
}
/* * We may have dropped the spin lock to get the head mutex lock, and * that might have given someone else time to free the head. If that's * true, it has been removed from our list and we can move on.
*/ if (!locked) return ERR_PTR(-EAGAIN);
if (RB_EMPTY_ROOT(&head->ref_tree.rb_root)) return NULL;
/* * Select a delayed ref of type BTRFS_ADD_DELAYED_REF first. * This is to prevent a ref count from going down to zero, which deletes * the extent item from the extent tree, when there still are references * to add, which would fail because they would not find the extent item.
*/ if (!list_empty(&head->ref_add_list)) return list_first_entry(&head->ref_add_list, struct btrfs_delayed_ref_node, add_list);
/* * Helper to insert the ref_node to the tail or merge with tail. * * Return false if the ref was inserted. * Return true if the ref was merged into an existing one (and therefore can be * freed by the caller).
*/ staticbool insert_delayed_ref(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *href, struct btrfs_delayed_ref_node *ref)
{ struct btrfs_delayed_ref_root *root = &trans->transaction->delayed_refs; struct btrfs_delayed_ref_node *exist; int mod;
spin_lock(&href->lock);
exist = tree_insert(&href->ref_tree, ref); if (!exist) { if (ref->action == BTRFS_ADD_DELAYED_REF)
list_add_tail(&ref->add_list, &href->ref_add_list);
spin_unlock(&href->lock);
trans->delayed_ref_updates++; returnfalse;
}
/* Now we are sure we can merge */ if (exist->action == ref->action) {
mod = ref->ref_mod;
} else { /* Need to change action */ if (exist->ref_mod < ref->ref_mod) {
exist->action = ref->action;
mod = -exist->ref_mod;
exist->ref_mod = ref->ref_mod; if (ref->action == BTRFS_ADD_DELAYED_REF)
list_add_tail(&exist->add_list,
&href->ref_add_list); elseif (ref->action == BTRFS_DROP_DELAYED_REF) {
ASSERT(!list_empty(&exist->add_list));
list_del_init(&exist->add_list);
} else {
ASSERT(0);
}
} else
mod = -ref->ref_mod;
}
exist->ref_mod += mod;
/* remove existing tail if its ref_mod is zero */ if (exist->ref_mod == 0)
drop_delayed_ref(trans->fs_info, root, href, exist);
spin_unlock(&href->lock); returntrue;
}
/* * helper function to update the accounting in the head ref * existing and update must have the same bytenr
*/ static noinline void update_existing_head_ref(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *existing, struct btrfs_delayed_ref_head *update)
{ struct btrfs_delayed_ref_root *delayed_refs =
&trans->transaction->delayed_refs; struct btrfs_fs_info *fs_info = trans->fs_info; int old_ref_mod;
BUG_ON(existing->is_data != update->is_data);
spin_lock(&existing->lock);
/* * When freeing an extent, we may not know the owning root when we * first create the head_ref. However, some deref before the last deref * will know it, so we just need to update the head_ref accordingly.
*/ if (!existing->owning_root)
existing->owning_root = update->owning_root;
if (update->must_insert_reserved) { /* if the extent was freed and then * reallocated before the delayed ref * entries were processed, we can end up * with an existing head ref without * the must_insert_reserved flag set. * Set it again here
*/
existing->must_insert_reserved = update->must_insert_reserved;
existing->owning_root = update->owning_root;
/* * update the num_bytes so we make sure the accounting * is done correctly
*/
existing->num_bytes = update->num_bytes;
}
if (update->extent_op) { if (!existing->extent_op) {
existing->extent_op = update->extent_op;
} else { if (update->extent_op->update_key) {
memcpy(&existing->extent_op->key,
&update->extent_op->key, sizeof(update->extent_op->key));
existing->extent_op->update_key = true;
} if (update->extent_op->update_flags) {
existing->extent_op->flags_to_set |=
update->extent_op->flags_to_set;
existing->extent_op->update_flags = true;
}
btrfs_free_delayed_extent_op(update->extent_op);
}
} /* * update the reference mod on the head to reflect this new operation, * only need the lock for this case cause we could be processing it * currently, for refs we just added we know we're a-ok.
*/
old_ref_mod = existing->total_ref_mod;
existing->ref_mod += update->ref_mod;
existing->total_ref_mod += update->ref_mod;
/* * If we are going to from a positive ref mod to a negative or vice * versa we need to make sure to adjust pending_csums accordingly. * We reserve bytes for csum deletion when adding or updating a ref head * see add_delayed_ref_head() for more details.
*/ if (existing->is_data) {
u64 csum_leaves =
btrfs_csum_bytes_to_leaves(fs_info,
existing->num_bytes);
/* If reserved is provided, it must be a data extent. */
BUG_ON(generic_ref->type != BTRFS_REF_DATA && reserved);
switch (generic_ref->action) { case BTRFS_ADD_DELAYED_REF: /* count_mod is already set to 1. */ break; case BTRFS_UPDATE_DELAYED_HEAD:
count_mod = 0; break; case BTRFS_DROP_DELAYED_REF: /* * The head node stores the sum of all the mods, so dropping a ref * should drop the sum in the head node by one.
*/
count_mod = -1; break; case BTRFS_ADD_DELAYED_EXTENT: /* * BTRFS_ADD_DELAYED_EXTENT means that we need to update the * reserved accounting when the extent is finally added, or if a * later modification deletes the delayed ref without ever * inserting the extent into the extent allocation tree. * ref->must_insert_reserved is the flag used to record that * accounting mods are required. * * Once we record must_insert_reserved, switch the action to * BTRFS_ADD_DELAYED_REF because other special casing is not * required.
*/
must_insert_reserved = true; break;
}
/* If not metadata set an impossible level to help debugging. */ if (generic_ref->type == BTRFS_REF_METADATA)
head_ref->level = generic_ref->tree_ref.level; else
head_ref->level = U8_MAX;
/* * helper function to actually insert a head node into the rbtree. * this does all the dirty work in terms of maintaining the correct * overall modification count. * * Returns an error pointer in case of an error.
*/ static noinline struct btrfs_delayed_ref_head *
add_delayed_ref_head(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *head_ref, struct btrfs_qgroup_extent_record *qrecord, int action, bool *qrecord_inserted_ret)
{ struct btrfs_fs_info *fs_info = trans->fs_info; struct btrfs_delayed_ref_head *existing; struct btrfs_delayed_ref_root *delayed_refs; constunsignedlong index = (head_ref->bytenr >> fs_info->sectorsize_bits); bool qrecord_inserted = false;
existing = xa_load(&delayed_refs->head_refs, index); if (existing) {
update_existing_head_ref(trans, existing, head_ref); /* * we've updated the existing ref, free the newly * allocated ref
*/
kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref);
head_ref = existing;
} else {
existing = xa_store(&delayed_refs->head_refs, index, head_ref, GFP_ATOMIC); if (xa_is_err(existing)) { /* Memory was preallocated by the caller. */
ASSERT(xa_err(existing) != -ENOMEM); return ERR_PTR(xa_err(existing));
} elseif (WARN_ON(existing)) { /* * Shouldn't happen we just did a lookup before under * delayed_refs->lock.
*/ return ERR_PTR(-EEXIST);
}
head_ref->tracked = true; /* * We reserve the amount of bytes needed to delete csums when * adding the ref head and not when adding individual drop refs * since the csum items are deleted only after running the last * delayed drop ref (the data extent's ref count drops to 0).
*/ if (head_ref->is_data && head_ref->ref_mod < 0) {
delayed_refs->pending_csums += head_ref->num_bytes;
trans->delayed_ref_csum_deletions +=
btrfs_csum_bytes_to_leaves(fs_info, head_ref->num_bytes);
}
delayed_refs->num_heads++;
delayed_refs->num_heads_ready++;
} if (qrecord_inserted_ret)
*qrecord_inserted_ret = qrecord_inserted;
return head_ref;
}
/* * Initialize the structure which represents a modification to a an extent. * * @fs_info: Internal to the mounted filesystem mount structure. * * @ref: The structure which is going to be initialized. * * @bytenr: The logical address of the extent for which a modification is * going to be recorded. * * @num_bytes: Size of the extent whose modification is being recorded. * * @ref_root: The id of the root where this modification has originated, this * can be either one of the well-known metadata trees or the * subvolume id which references this extent. * * @action: Can be one of BTRFS_ADD_DELAYED_REF/BTRFS_DROP_DELAYED_REF or * BTRFS_ADD_DELAYED_EXTENT * * @ref_type: Holds the type of the extent which is being recorded, can be * one of BTRFS_SHARED_BLOCK_REF_KEY/BTRFS_TREE_BLOCK_REF_KEY * when recording a metadata extent or BTRFS_SHARED_DATA_REF_KEY/ * BTRFS_EXTENT_DATA_REF_KEY when recording data extent
*/ staticvoid init_delayed_ref_common(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_node *ref, struct btrfs_ref *generic_ref)
{ int action = generic_ref->action;
u64 seq = 0;
if (action == BTRFS_ADD_DELAYED_EXTENT)
action = BTRFS_ADD_DELAYED_REF;
if (btrfs_is_fstree(generic_ref->ref_root))
seq = atomic64_read(&fs_info->tree_mod_seq);
node = kmem_cache_alloc(btrfs_delayed_ref_node_cachep, GFP_NOFS); if (!node) return -ENOMEM;
head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS); if (!head_ref) {
ret = -ENOMEM; goto free_node;
}
delayed_refs = &trans->transaction->delayed_refs;
if (btrfs_qgroup_full_accounting(fs_info) && !generic_ref->skip_qgroup) {
record = kzalloc(sizeof(*record), GFP_NOFS); if (!record) {
ret = -ENOMEM; goto free_head_ref;
} if (xa_reserve(&delayed_refs->dirty_extents, index, GFP_NOFS)) {
ret = -ENOMEM; goto free_record;
}
qrecord_reserved = true;
}
ret = xa_reserve(&delayed_refs->head_refs, index, GFP_NOFS); if (ret) { if (qrecord_reserved)
xa_release(&delayed_refs->dirty_extents, index); goto free_record;
}
/* * insert both the head node and the new ref without dropping * the spin lock
*/
new_head_ref = add_delayed_ref_head(trans, head_ref, record,
action, &qrecord_inserted); if (IS_ERR(new_head_ref)) {
xa_release(&delayed_refs->head_refs, index);
spin_unlock(&delayed_refs->lock);
ret = PTR_ERR(new_head_ref); goto free_record;
}
head_ref = new_head_ref;
/* * Add a delayed tree ref. This does all of the accounting required to make sure * the delayed ref is eventually processed before this transaction commits.
*/ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans, struct btrfs_ref *generic_ref, struct btrfs_delayed_extent_op *extent_op)
{
ASSERT(generic_ref->type == BTRFS_REF_METADATA && generic_ref->action); return add_delayed_ref(trans, generic_ref, extent_op, 0);
}
/* * add a delayed data ref. it's similar to btrfs_add_delayed_tree_ref.
*/ int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans, struct btrfs_ref *generic_ref,
u64 reserved)
{
ASSERT(generic_ref->type == BTRFS_REF_DATA && generic_ref->action); return add_delayed_ref(trans, generic_ref, NULL, reserved);
}
/* * This does a simple search for the head node for a given extent. Returns the * head node if found, or NULL if not.
*/ struct btrfs_delayed_ref_head *
btrfs_find_delayed_ref_head(conststruct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_root *delayed_refs,
u64 bytenr)
{ constunsignedlong index = (bytenr >> fs_info->sectorsize_bits);
staticint find_comp(struct btrfs_delayed_ref_node *entry, u64 root, u64 parent)
{ int type = parent ? BTRFS_SHARED_BLOCK_REF_KEY : BTRFS_TREE_BLOCK_REF_KEY;
if (type < entry->type) return -1; if (type > entry->type) return 1;
if (type == BTRFS_TREE_BLOCK_REF_KEY) { if (root < entry->ref_root) return -1; if (root > entry->ref_root) return 1;
} else { if (parent < entry->parent) return -1; if (parent > entry->parent) return 1;
} return 0;
}
/* * Check to see if a given root/parent reference is attached to the head. This * only checks for BTRFS_ADD_DELAYED_REF references that match, as that * indicates the reference exists for the given root or parent. This is for * tree blocks only. * * @head: the head of the bytenr we're searching. * @root: the root objectid of the reference if it is a normal reference. * @parent: the parent if this is a shared backref.
*/ bool btrfs_find_delayed_tree_ref(struct btrfs_delayed_ref_head *head,
u64 root, u64 parent)
{ struct rb_node *node; bool found = false;
lockdep_assert_held(&head->mutex);
spin_lock(&head->lock);
node = head->ref_tree.rb_root.rb_node; while (node) { struct btrfs_delayed_ref_node *entry; int ret;
entry = rb_entry(node, struct btrfs_delayed_ref_node, ref_node);
ret = find_comp(entry, root, parent); if (ret < 0) {
node = node->rb_left;
} elseif (ret > 0) {
node = node->rb_right;
} else { /* * We only want to count ADD actions, as drops mean the * ref doesn't exist.
*/ if (entry->action == BTRFS_ADD_DELAYED_REF)
found = true; break;
}
}
spin_unlock(&head->lock); return found;
}
if (!testing && pin_bytes) { struct btrfs_block_group *bg;
bg = btrfs_lookup_block_group(fs_info, head->bytenr); if (WARN_ON_ONCE(bg == NULL)) { /* * Unexpected and there's nothing we can do here * because we are in a transaction abort path, * so any errors can only be ignored or reported * while attempting to cleanup all resources.
*/
btrfs_err(fs_info, "block group for delayed ref at %llu was not found while destroying ref head",
head->bytenr);
} else {
spin_lock(&bg->space_info->lock);
spin_lock(&bg->lock);
bg->pinned += head->num_bytes;
btrfs_space_info_update_bytes_pinned(bg->space_info,
head->num_bytes);
bg->reserved -= head->num_bytes;
bg->space_info->bytes_reserved -= head->num_bytes;
spin_unlock(&bg->lock);
spin_unlock(&bg->space_info->lock);
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.