Merge tag 'for-f2fs-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk...

[cascardo/linux.git] / fs / f2fs / node.c

diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index ca1bb3c..b2fa4b6 100644 (file)
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -206,14 +206,14 @@ int need_dentry_mark(struct f2fs_sb_info *sbi, nid_t nid)
        struct nat_entry *e;
        bool need = false;
 
-       down_read(&nm_i->nat_tree_lock);
+       percpu_down_read(&nm_i->nat_tree_lock);
        e = __lookup_nat_cache(nm_i, nid);
        if (e) {
                if (!get_nat_flag(e, IS_CHECKPOINTED) &&
                                !get_nat_flag(e, HAS_FSYNCED_INODE))
                        need = true;
        }
-       up_read(&nm_i->nat_tree_lock);
+       percpu_up_read(&nm_i->nat_tree_lock);
        return need;
 }
 
@@ -223,11 +223,11 @@ bool is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid)
        struct nat_entry *e;
        bool is_cp = true;
 
-       down_read(&nm_i->nat_tree_lock);
+       percpu_down_read(&nm_i->nat_tree_lock);
        e = __lookup_nat_cache(nm_i, nid);
        if (e && !get_nat_flag(e, IS_CHECKPOINTED))
                is_cp = false;
-       up_read(&nm_i->nat_tree_lock);
+       percpu_up_read(&nm_i->nat_tree_lock);
        return is_cp;
 }
 
@@ -237,13 +237,13 @@ bool need_inode_block_update(struct f2fs_sb_info *sbi, nid_t ino)
        struct nat_entry *e;
        bool need_update = true;
 
-       down_read(&nm_i->nat_tree_lock);
+       percpu_down_read(&nm_i->nat_tree_lock);
        e = __lookup_nat_cache(nm_i, ino);
        if (e && get_nat_flag(e, HAS_LAST_FSYNC) &&
                        (get_nat_flag(e, IS_CHECKPOINTED) ||
                         get_nat_flag(e, HAS_FSYNCED_INODE)))
                need_update = false;
-       up_read(&nm_i->nat_tree_lock);
+       percpu_up_read(&nm_i->nat_tree_lock);
        return need_update;
 }
 
@@ -284,7 +284,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
        struct f2fs_nm_info *nm_i = NM_I(sbi);
        struct nat_entry *e;
 
-       down_write(&nm_i->nat_tree_lock);
+       percpu_down_write(&nm_i->nat_tree_lock);
        e = __lookup_nat_cache(nm_i, ni->nid);
        if (!e) {
                e = grab_nat_entry(nm_i, ni->nid);
@@ -334,7 +334,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
                        set_nat_flag(e, HAS_FSYNCED_INODE, true);
                set_nat_flag(e, HAS_LAST_FSYNC, fsync_done);
        }
-       up_write(&nm_i->nat_tree_lock);
+       percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
@@ -342,8 +342,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
        struct f2fs_nm_info *nm_i = NM_I(sbi);
        int nr = nr_shrink;
 
-       if (!down_write_trylock(&nm_i->nat_tree_lock))
-               return 0;
+       percpu_down_write(&nm_i->nat_tree_lock);
 
        while (nr_shrink && !list_empty(&nm_i->nat_entries)) {
                struct nat_entry *ne;
@@ -352,7 +351,7 @@ int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
                __del_from_nat_cache(nm_i, ne);
                nr_shrink--;
        }
-       up_write(&nm_i->nat_tree_lock);
+       percpu_up_write(&nm_i->nat_tree_lock);
        return nr - nr_shrink;
 }
 
@@ -374,13 +373,13 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
        ni->nid = nid;
 
        /* Check nat cache */
-       down_read(&nm_i->nat_tree_lock);
+       percpu_down_read(&nm_i->nat_tree_lock);
        e = __lookup_nat_cache(nm_i, nid);
        if (e) {
                ni->ino = nat_get_ino(e);
                ni->blk_addr = nat_get_blkaddr(e);
                ni->version = nat_get_version(e);
-               up_read(&nm_i->nat_tree_lock);
+               percpu_up_read(&nm_i->nat_tree_lock);
                return;
        }
 
@@ -404,11 +403,11 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)
        node_info_from_raw_nat(ni, &ne);
        f2fs_put_page(page, 1);
 cache:
-       up_read(&nm_i->nat_tree_lock);
+       percpu_up_read(&nm_i->nat_tree_lock);
        /* cache nat entry */
-       down_write(&nm_i->nat_tree_lock);
+       percpu_down_write(&nm_i->nat_tree_lock);
        cache_nat_entry(sbi, nid, &ne);
-       up_write(&nm_i->nat_tree_lock);
+       percpu_up_write(&nm_i->nat_tree_lock);
 }
 
 /*
@@ -650,6 +649,7 @@ release_out:
        if (err == -ENOENT) {
                dn->cur_level = i;
                dn->max_level = level;
+               dn->ofs_in_node = offset[level];
        }
        return err;
 }
@@ -1068,18 +1068,22 @@ fail:
  * 0: f2fs_put_page(page, 0)
  * LOCKED_PAGE or error: f2fs_put_page(page, 1)
  */
-static int read_node_page(struct page *page, int rw)
+static int read_node_page(struct page *page, int op_flags)
 {
        struct f2fs_sb_info *sbi = F2FS_P_SB(page);
        struct node_info ni;
        struct f2fs_io_info fio = {
                .sbi = sbi,
                .type = NODE,
-               .rw = rw,
+               .op = REQ_OP_READ,
+               .op_flags = op_flags,
                .page = page,
                .encrypted_page = NULL,
        };
 
+       if (PageUptodate(page))
+               return LOCKED_PAGE;
+
        get_node_info(sbi, page->index, &ni);
 
        if (unlikely(ni.blk_addr == NULL_ADDR)) {
@@ -1087,9 +1091,6 @@ static int read_node_page(struct page *page, int rw)
                return -ENOENT;
        }
 
-       if (PageUptodate(page))
-               return LOCKED_PAGE;
-
        fio.new_blkaddr = fio.old_blkaddr = ni.blk_addr;
        return f2fs_submit_page_bio(&fio);
 }
@@ -1116,7 +1117,7 @@ void ra_node_page(struct f2fs_sb_info *sbi, nid_t nid)
        if (!apage)
                return;
 
-       err = read_node_page(apage, READA);
+       err = read_node_page(apage, REQ_RAHEAD);
        f2fs_put_page(apage, err ? 1 : 0);
 }
 
@@ -1558,7 +1559,8 @@ static int f2fs_write_node_page(struct page *page,
        struct f2fs_io_info fio = {
                .sbi = sbi,
                .type = NODE,
-               .rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE,
+               .op = REQ_OP_WRITE,
+               .op_flags = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : 0,
                .page = page,
                .encrypted_page = NULL,
        };
@@ -1618,6 +1620,7 @@ static int f2fs_write_node_pages(struct address_space *mapping,
                            struct writeback_control *wbc)
 {
        struct f2fs_sb_info *sbi = F2FS_M_SB(mapping);
+       struct blk_plug plug;
        long diff;
 
        /* balancing f2fs's metadata in background */
@@ -1631,7 +1634,9 @@ static int f2fs_write_node_pages(struct address_space *mapping,
 
        diff = nr_pages_to_write(sbi, NODE, wbc);
        wbc->sync_mode = WB_SYNC_NONE;
+       blk_start_plug(&plug);
        sync_node_pages(sbi, wbc);
+       blk_finish_plug(&plug);
        wbc->nr_to_write = max((long)0, wbc->nr_to_write - diff);
        return 0;
 
@@ -1783,7 +1788,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
        ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nid), FREE_NID_PAGES,
                                                        META_NAT, true);
 
-       down_read(&nm_i->nat_tree_lock);
+       percpu_down_read(&nm_i->nat_tree_lock);
 
        while (1) {
                struct page *page = get_current_nat_page(sbi, nid);
@@ -1815,7 +1820,7 @@ void build_free_nids(struct f2fs_sb_info *sbi)
                        remove_free_nid(nm_i, nid);
        }
        up_read(&curseg->journal_rwsem);
-       up_read(&nm_i->nat_tree_lock);
+       percpu_up_read(&nm_i->nat_tree_lock);
 
        ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),
                                        nm_i->ra_nid_pages, META_NAT, false);
@@ -2204,7 +2209,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
        if (!nm_i->dirty_nat_cnt)
                return;
 
-       down_write(&nm_i->nat_tree_lock);
+       percpu_down_write(&nm_i->nat_tree_lock);
 
        /*
         * if there are no enough space in journal to store dirty nat
@@ -2227,7 +2232,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
        list_for_each_entry_safe(set, tmp, &sets, set_list)
                __flush_nat_entry_set(sbi, set);
 
-       up_write(&nm_i->nat_tree_lock);
+       percpu_up_write(&nm_i->nat_tree_lock);
 
        f2fs_bug_on(sbi, nm_i->dirty_nat_cnt);
 }
@@ -2263,7 +2268,8 @@ static int init_node_manager(struct f2fs_sb_info *sbi)
 
        mutex_init(&nm_i->build_lock);
        spin_lock_init(&nm_i->free_nid_list_lock);
-       init_rwsem(&nm_i->nat_tree_lock);
+       if (percpu_init_rwsem(&nm_i->nat_tree_lock))
+               return -ENOMEM;
 
        nm_i->next_scan_nid = le32_to_cpu(sbi->ckpt->next_free_nid);
        nm_i->bitmap_size = __bitmap_size(sbi, NAT_BITMAP);
@@ -2320,7 +2326,7 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
        spin_unlock(&nm_i->free_nid_list_lock);
 
        /* destroy nat cache */
-       down_write(&nm_i->nat_tree_lock);
+       percpu_down_write(&nm_i->nat_tree_lock);
        while ((found = __gang_lookup_nat_cache(nm_i,
                                        nid, NATVEC_SIZE, natvec))) {
                unsigned idx;
@@ -2345,8 +2351,9 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
                        kmem_cache_free(nat_entry_set_slab, setvec[idx]);
                }
        }
-       up_write(&nm_i->nat_tree_lock);
+       percpu_up_write(&nm_i->nat_tree_lock);
 
+       percpu_free_rwsem(&nm_i->nat_tree_lock);
        kfree(nm_i->nat_bitmap);
        sbi->nm_info = NULL;
        kfree(nm_i);