Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit a5131193 authored by Jaegeuk Kim's avatar Jaegeuk Kim
Browse files

f2fs: cover more area with nat_tree_lock 



There was a subtle bug on nat cache management which incurs wrong nid allocation
or wrong block addresses when try_to_free_nats is triggered heavily.
This patch enlarges the previous coverage of nat_tree_lock to avoid data race.

Signed-off-by: default avatarJaegeuk Kim <jaegeuk@kernel.org>
parent e0afc4d6

fs/f2fs/node.c

+12 −17
Original line number Original line Diff line number Diff line
@@ -262,13 +262,11 @@ static void cache_nat_entry(struct f2fs_nm_info *nm_i, nid_t nid, 
{

{

	struct nat_entry *e;

	struct nat_entry *e;





	down_write(&nm_i->nat_tree_lock);

	e = __lookup_nat_cache(nm_i, nid);

	e = __lookup_nat_cache(nm_i, nid);

	if (!e) {

	if (!e) {

		e = grab_nat_entry(nm_i, nid);

		e = grab_nat_entry(nm_i, nid);

		node_info_from_raw_nat(&e->ni, ne);

		node_info_from_raw_nat(&e->ni, ne);

	}

	}

	up_write(&nm_i->nat_tree_lock);

}

}





static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,

static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
@@ -380,6 +378,8 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni) 




	memset(&ne, 0, sizeof(struct f2fs_nat_entry));

	memset(&ne, 0, sizeof(struct f2fs_nat_entry));





	down_write(&nm_i->nat_tree_lock);



	/* Check current segment summary */

	/* Check current segment summary */

	mutex_lock(&curseg->curseg_mutex);

	mutex_lock(&curseg->curseg_mutex);

	i = lookup_journal_in_cursum(sum, NAT_JOURNAL, nid, 0);

	i = lookup_journal_in_cursum(sum, NAT_JOURNAL, nid, 0);
@@ -400,6 +400,7 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni) 
cache:

cache:

	/* cache nat entry */

	/* cache nat entry */

	cache_nat_entry(NM_I(sbi), nid, &ne);

	cache_nat_entry(NM_I(sbi), nid, &ne);

	up_write(&nm_i->nat_tree_lock);

}

}





/*

/*
@@ -1459,13 +1460,10 @@ static int add_free_nid(struct f2fs_sb_info *sbi, nid_t nid, bool build) 




	if (build) {

	if (build) {

		/* do not add allocated nids */

		/* do not add allocated nids */

		down_read(&nm_i->nat_tree_lock);

		ne = __lookup_nat_cache(nm_i, nid);

		ne = __lookup_nat_cache(nm_i, nid);

		if (ne &&

		if (ne && (!get_nat_flag(ne, IS_CHECKPOINTED) ||

			(!get_nat_flag(ne, IS_CHECKPOINTED) ||

				nat_get_blkaddr(ne) != NULL_ADDR))

				nat_get_blkaddr(ne) != NULL_ADDR))

			allocated = true;

			allocated = true;

		up_read(&nm_i->nat_tree_lock);

		if (allocated)

		if (allocated)

			return 0;

			return 0;

	}

	}
@@ -1551,6 +1549,8 @@ static void build_free_nids(struct f2fs_sb_info *sbi) 
	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nid), FREE_NID_PAGES,

	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nid), FREE_NID_PAGES,

							META_NAT, true);

							META_NAT, true);





	down_read(&nm_i->nat_tree_lock);



	while (1) {

	while (1) {

		struct page *page = get_current_nat_page(sbi, nid);

		struct page *page = get_current_nat_page(sbi, nid);
@@ -1579,6 +1579,7 @@ static void build_free_nids(struct f2fs_sb_info *sbi) 
			remove_free_nid(nm_i, nid);

			remove_free_nid(nm_i, nid);

	}

	}

	mutex_unlock(&curseg->curseg_mutex);

	mutex_unlock(&curseg->curseg_mutex);

	up_read(&nm_i->nat_tree_lock);





	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),

	ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),

					nm_i->ra_nid_pages, META_NAT, false);

					nm_i->ra_nid_pages, META_NAT, false);
@@ -1861,14 +1862,12 @@ static void remove_nats_in_journal(struct f2fs_sb_info *sbi) 




		raw_ne = nat_in_journal(sum, i);

		raw_ne = nat_in_journal(sum, i);





		down_write(&nm_i->nat_tree_lock);

		ne = __lookup_nat_cache(nm_i, nid);

		ne = __lookup_nat_cache(nm_i, nid);

		if (!ne) {

		if (!ne) {

			ne = grab_nat_entry(nm_i, nid);

			ne = grab_nat_entry(nm_i, nid);

			node_info_from_raw_nat(&ne->ni, &raw_ne);

			node_info_from_raw_nat(&ne->ni, &raw_ne);

		}

		}

		__set_nat_cache_dirty(nm_i, ne);

		__set_nat_cache_dirty(nm_i, ne);

		up_write(&nm_i->nat_tree_lock);

	}

	}

	update_nats_in_cursum(sum, -i);

	update_nats_in_cursum(sum, -i);

	mutex_unlock(&curseg->curseg_mutex);

	mutex_unlock(&curseg->curseg_mutex);
@@ -1902,7 +1901,6 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, 
	struct f2fs_nat_block *nat_blk;

	struct f2fs_nat_block *nat_blk;

	struct nat_entry *ne, *cur;

	struct nat_entry *ne, *cur;

	struct page *page = NULL;

	struct page *page = NULL;

	struct f2fs_nm_info *nm_i = NM_I(sbi);





	/*

	/*

	 * there are two steps to flush nat entries:

	 * there are two steps to flush nat entries:
@@ -1939,12 +1937,8 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, 
			raw_ne = &nat_blk->entries[nid - start_nid];

			raw_ne = &nat_blk->entries[nid - start_nid];

		}

		}

		raw_nat_from_node_info(raw_ne, &ne->ni);

		raw_nat_from_node_info(raw_ne, &ne->ni);



		down_write(&NM_I(sbi)->nat_tree_lock);

		nat_reset_flag(ne);

		nat_reset_flag(ne);

		__clear_nat_cache_dirty(NM_I(sbi), ne);

		__clear_nat_cache_dirty(NM_I(sbi), ne);

		up_write(&NM_I(sbi)->nat_tree_lock);



		if (nat_get_blkaddr(ne) == NULL_ADDR)

		if (nat_get_blkaddr(ne) == NULL_ADDR)

			add_free_nid(sbi, nid, false);

			add_free_nid(sbi, nid, false);

	}

	}
@@ -1956,9 +1950,7 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, 




	f2fs_bug_on(sbi, set->entry_cnt);

	f2fs_bug_on(sbi, set->entry_cnt);





	down_write(&nm_i->nat_tree_lock);

	radix_tree_delete(&NM_I(sbi)->nat_set_root, set->set);

	radix_tree_delete(&NM_I(sbi)->nat_set_root, set->set);

	up_write(&nm_i->nat_tree_lock);

	kmem_cache_free(nat_entry_set_slab, set);

	kmem_cache_free(nat_entry_set_slab, set);

}

}
@@ -1978,6 +1970,9 @@ void flush_nat_entries(struct f2fs_sb_info *sbi) 




	if (!nm_i->dirty_nat_cnt)

	if (!nm_i->dirty_nat_cnt)

		return;

		return;



	down_write(&nm_i->nat_tree_lock);



	/*

	/*

	 * if there are no enough space in journal to store dirty nat

	 * if there are no enough space in journal to store dirty nat

	 * entries, remove all entries from journal and merge them

	 * entries, remove all entries from journal and merge them
@@ -1986,7 +1981,6 @@ void flush_nat_entries(struct f2fs_sb_info *sbi) 
	if (!__has_cursum_space(sum, nm_i->dirty_nat_cnt, NAT_JOURNAL))

	if (!__has_cursum_space(sum, nm_i->dirty_nat_cnt, NAT_JOURNAL))

		remove_nats_in_journal(sbi);

		remove_nats_in_journal(sbi);





	down_write(&nm_i->nat_tree_lock);

	while ((found = __gang_lookup_nat_set(nm_i,

	while ((found = __gang_lookup_nat_set(nm_i,

					set_idx, SETVEC_SIZE, setvec))) {

					set_idx, SETVEC_SIZE, setvec))) {

		unsigned idx;

		unsigned idx;
@@ -1995,12 +1989,13 @@ void flush_nat_entries(struct f2fs_sb_info *sbi) 
			__adjust_nat_entry_set(setvec[idx], &sets,

			__adjust_nat_entry_set(setvec[idx], &sets,

							MAX_NAT_JENTRIES(sum));

							MAX_NAT_JENTRIES(sum));

	}

	}

	up_write(&nm_i->nat_tree_lock);





	/* flush dirty nats in nat entry set */

	/* flush dirty nats in nat entry set */

	list_for_each_entry_safe(set, tmp, &sets, set_list)

	list_for_each_entry_safe(set, tmp, &sets, set_list)

		__flush_nat_entry_set(sbi, set);

		__flush_nat_entry_set(sbi, set);





	up_write(&nm_i->nat_tree_lock);



	f2fs_bug_on(sbi, nm_i->dirty_nat_cnt);

	f2fs_bug_on(sbi, nm_i->dirty_nat_cnt);

}

}