11 files changed, 113 insertions, 32 deletions

diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 88d9af3d4581..5fb60ea7eee2 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -328,8 +328,8 @@ static inline void __btrfs_queue_work(struct __btrfs_workqueue *wq,
                 list_add_tail(&work->ordered_list, &wq->ordered_list);
                 spin_unlock_irqrestore(&wq->list_lock, flags);
         }
-        queue_work(wq->normal_wq, &work->normal_work);
         trace_btrfs_work_queued(work);
+        queue_work(wq->normal_wq, &work->normal_work);
 }
 
 void btrfs_queue_work(struct btrfs_workqueue *wq,
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index dd08e29f5117..4545e2e2ad45 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -182,6 +182,7 @@ static struct btrfs_lockdep_keyset {
         { .id = BTRFS_TREE_RELOC_OBJECTID,      .name_stem = "treloc"   },
         { .id = BTRFS_DATA_RELOC_TREE_OBJECTID, .name_stem = "dreloc"   },
         { .id = BTRFS_UUID_TREE_OBJECTID,       .name_stem = "uuid"     },
+        { .id = BTRFS_FREE_SPACE_TREE_OBJECTID, .name_stem = "free-space" },
         { .id = 0,                              .name_stem = "tree"     },
 };
 
@@ -1787,7 +1788,6 @@ static int cleaner_kthread(void *arg)
         int again;
         struct btrfs_trans_handle *trans;
 
-        set_freezable();
         do {
                 again = 0;
 
diff --git a/fs/btrfs/free-space-tree.c b/fs/btrfs/free-space-tree.c
index 393e36bd5845..53dbeaf6ce94 100644
--- a/fs/btrfs/free-space-tree.c
+++ b/fs/btrfs/free-space-tree.c
@@ -153,6 +153,20 @@ static inline u32 free_space_bitmap_size(u64 size, u32 sectorsize)
 
 static unsigned long *alloc_bitmap(u32 bitmap_size)
 {
+        void *mem;
+
+        /*
+         * The allocation size varies, observed numbers were < 4K up to 16K.
+         * Using vmalloc unconditionally would be too heavy, we'll try
+         * contiguous allocations first.
+         */
+        if  (bitmap_size <= PAGE_SIZE)
+                return kzalloc(bitmap_size, GFP_NOFS);
+
+        mem = kzalloc(bitmap_size, GFP_NOFS | __GFP_NOWARN);
+        if (mem)
+                return mem;
+
         return __vmalloc(bitmap_size, GFP_NOFS | __GFP_HIGHMEM | __GFP_ZERO,
                          PAGE_KERNEL);
 }
@@ -289,7 +303,7 @@ int convert_free_space_to_bitmaps(struct btrfs_trans_handle *trans,
 
         ret = 0;
 out:
-        vfree(bitmap);
+        kvfree(bitmap);
         if (ret)
                 btrfs_abort_transaction(trans, root, ret);
         return ret;
@@ -438,7 +452,7 @@ int convert_free_space_to_extents(struct btrfs_trans_handle *trans,
 
         ret = 0;
 out:
-        vfree(bitmap);
+        kvfree(bitmap);
         if (ret)
                 btrfs_abort_transaction(trans, root, ret);
         return ret;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index e28f3d4691af..5f06eb1f4384 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -7116,21 +7116,41 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
         if (ret)
                 return ERR_PTR(ret);
 
-        em = create_pinned_em(inode, start, ins.offset, start, ins.objectid,
-                              ins.offset, ins.offset, ins.offset, 0);
-        if (IS_ERR(em)) {
-                btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1);
-                return em;
-        }
-
+        /*
+         * Create the ordered extent before the extent map. This is to avoid
+         * races with the fast fsync path that would lead to it logging file
+         * extent items that point to disk extents that were not yet written to.
+         * The fast fsync path collects ordered extents into a local list and
+         * then collects all the new extent maps, so we must create the ordered
+         * extent first and make sure the fast fsync path collects any new
+         * ordered extents after collecting new extent maps as well.
+         * The fsync path simply can not rely on inode_dio_wait() because it
+         * causes deadlock with AIO.
+         */
         ret = btrfs_add_ordered_extent_dio(inode, start, ins.objectid,
                                            ins.offset, ins.offset, 0);
         if (ret) {
                 btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1);
-                free_extent_map(em);
                 return ERR_PTR(ret);
         }
 
+        em = create_pinned_em(inode, start, ins.offset, start, ins.objectid,
+                              ins.offset, ins.offset, ins.offset, 0);
+        if (IS_ERR(em)) {
+                struct btrfs_ordered_extent *oe;
+
+                btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1);
+                oe = btrfs_lookup_ordered_extent(inode, start);
+                ASSERT(oe);
+                if (WARN_ON(!oe))
+                        return em;
+                set_bit(BTRFS_ORDERED_IOERR, &oe->flags);
+                set_bit(BTRFS_ORDERED_IO_DONE, &oe->flags);
+                btrfs_remove_ordered_extent(inode, oe);
+                /* Once for our lookup and once for the ordered extents tree. */
+                btrfs_put_ordered_extent(oe);
+                btrfs_put_ordered_extent(oe);
+        }
         return em;
 }
 
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index fd1c4d982463..2bd0011450df 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -575,7 +575,8 @@ static int is_cowonly_root(u64 root_objectid)
             root_objectid == BTRFS_TREE_LOG_OBJECTID ||
             root_objectid == BTRFS_CSUM_TREE_OBJECTID ||
             root_objectid == BTRFS_UUID_TREE_OBJECTID ||
-            root_objectid == BTRFS_QUOTA_TREE_OBJECTID)
+            root_objectid == BTRFS_QUOTA_TREE_OBJECTID ||
+            root_objectid == BTRFS_FREE_SPACE_TREE_OBJECTID)
                 return 1;
         return 0;
 }
diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c
index e0ac85949067..539e7b5e3f86 100644
--- a/fs/btrfs/sysfs.c
+++ b/fs/btrfs/sysfs.c
@@ -202,6 +202,7 @@ BTRFS_FEAT_ATTR_INCOMPAT(extended_iref, EXTENDED_IREF);
 BTRFS_FEAT_ATTR_INCOMPAT(raid56, RAID56);
 BTRFS_FEAT_ATTR_INCOMPAT(skinny_metadata, SKINNY_METADATA);
 BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES);
+BTRFS_FEAT_ATTR_COMPAT_RO(free_space_tree, FREE_SPACE_TREE);
 
 static struct attribute *btrfs_supported_feature_attrs[] = {
         BTRFS_FEAT_ATTR_PTR(mixed_backref),
@@ -213,6 +214,7 @@ static struct attribute *btrfs_supported_feature_attrs[] = {
         BTRFS_FEAT_ATTR_PTR(raid56),
         BTRFS_FEAT_ATTR_PTR(skinny_metadata),
         BTRFS_FEAT_ATTR_PTR(no_holes),
+        BTRFS_FEAT_ATTR_PTR(free_space_tree),
         NULL
 };
 
@@ -780,6 +782,39 @@ failure:
         return error;
 }
 
+
+/*
+ * Change per-fs features in /sys/fs/btrfs/UUID/features to match current
+ * values in superblock. Call after any changes to incompat/compat_ro flags
+ */
+void btrfs_sysfs_feature_update(struct btrfs_fs_info *fs_info,
+                u64 bit, enum btrfs_feature_set set)
+{
+        struct btrfs_fs_devices *fs_devs;
+        struct kobject *fsid_kobj;
+        u64 features;
+        int ret;
+
+        if (!fs_info)
+                return;
+
+        features = get_features(fs_info, set);
+        ASSERT(bit & supported_feature_masks[set]);
+
+        fs_devs = fs_info->fs_devices;
+        fsid_kobj = &fs_devs->fsid_kobj;
+
+        if (!fsid_kobj->state_initialized)
+                return;
+
+        /*
+         * FIXME: this is too heavy to update just one value, ideally we'd like
+         * to use sysfs_update_group but some refactoring is needed first.
+         */
+        sysfs_remove_group(fsid_kobj, &btrfs_feature_attr_group);
+        ret = sysfs_create_group(fsid_kobj, &btrfs_feature_attr_group);
+}
+
 static int btrfs_init_debugfs(void)
 {
 #ifdef CONFIG_DEBUG_FS
diff --git a/fs/btrfs/sysfs.h b/fs/btrfs/sysfs.h
index 9c09522125a6..d7da1a4c2f6c 100644
--- a/fs/btrfs/sysfs.h
+++ b/fs/btrfs/sysfs.h
@@ -56,7 +56,7 @@ static struct btrfs_feature_attr btrfs_attr_##_name = {			     \
 #define BTRFS_FEAT_ATTR_COMPAT(name, feature) \
         BTRFS_FEAT_ATTR(name, FEAT_COMPAT, BTRFS_FEATURE_COMPAT, feature)
 #define BTRFS_FEAT_ATTR_COMPAT_RO(name, feature) \
-        BTRFS_FEAT_ATTR(name, FEAT_COMPAT_RO, BTRFS_FEATURE_COMPAT, feature)
+        BTRFS_FEAT_ATTR(name, FEAT_COMPAT_RO, BTRFS_FEATURE_COMPAT_RO, feature)
 #define BTRFS_FEAT_ATTR_INCOMPAT(name, feature) \
         BTRFS_FEAT_ATTR(name, FEAT_INCOMPAT, BTRFS_FEATURE_INCOMPAT, feature)
 
@@ -90,4 +90,7 @@ int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs,
                                 struct kobject *parent);
 int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs);
 void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs);
+void btrfs_sysfs_feature_update(struct btrfs_fs_info *fs_info,
+                u64 bit, enum btrfs_feature_set set);
+
 #endif /* _BTRFS_SYSFS_H_ */
diff --git a/fs/btrfs/tests/btrfs-tests.c b/fs/btrfs/tests/btrfs-tests.c
index b1d920b30070..0e1e61a7ec23 100644
--- a/fs/btrfs/tests/btrfs-tests.c
+++ b/fs/btrfs/tests/btrfs-tests.c
@@ -82,18 +82,18 @@ void btrfs_destroy_test_fs(void)
 struct btrfs_fs_info *btrfs_alloc_dummy_fs_info(void)
 {
         struct btrfs_fs_info *fs_info = kzalloc(sizeof(struct btrfs_fs_info),
-                                                GFP_NOFS);
+                                                GFP_KERNEL);
 
         if (!fs_info)
                 return fs_info;
         fs_info->fs_devices = kzalloc(sizeof(struct btrfs_fs_devices),
-                                      GFP_NOFS);
+                                      GFP_KERNEL);
         if (!fs_info->fs_devices) {
                 kfree(fs_info);
                 return NULL;
         }
         fs_info->super_copy = kzalloc(sizeof(struct btrfs_super_block),
-                                      GFP_NOFS);
+                                      GFP_KERNEL);
         if (!fs_info->super_copy) {
                 kfree(fs_info->fs_devices);
                 kfree(fs_info);
@@ -180,11 +180,11 @@ btrfs_alloc_dummy_block_group(unsigned long length)
 {
         struct btrfs_block_group_cache *cache;
 
-        cache = kzalloc(sizeof(*cache), GFP_NOFS);
+        cache = kzalloc(sizeof(*cache), GFP_KERNEL);
         if (!cache)
                 return NULL;
         cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
-                                        GFP_NOFS);
+                                        GFP_KERNEL);
         if (!cache->free_space_ctl) {
                 kfree(cache);
                 return NULL;
diff --git a/fs/btrfs/tests/extent-io-tests.c b/fs/btrfs/tests/extent-io-tests.c
index e29fa297e053..669b58201e36 100644
--- a/fs/btrfs/tests/extent-io-tests.c
+++ b/fs/btrfs/tests/extent-io-tests.c
@@ -94,7 +94,7 @@ static int test_find_delalloc(void)
          * test.
          */
         for (index = 0; index < (total_dirty >> PAGE_CACHE_SHIFT); index++) {
-                page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
+                page = find_or_create_page(inode->i_mapping, index, GFP_KERNEL);
                 if (!page) {
                         test_msg("Failed to allocate test page\n");
                         ret = -ENOMEM;
@@ -113,7 +113,7 @@ static int test_find_delalloc(void)
          * |--- delalloc ---|
          * |---  search  ---|
          */
-        set_extent_delalloc(&tmp, 0, 4095, NULL, GFP_NOFS);
+        set_extent_delalloc(&tmp, 0, 4095, NULL, GFP_KERNEL);
         start = 0;
         end = 0;
         found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
@@ -144,7 +144,7 @@ static int test_find_delalloc(void)
                 test_msg("Couldn't find the locked page\n");
                 goto out_bits;
         }
-        set_extent_delalloc(&tmp, 4096, max_bytes - 1, NULL, GFP_NOFS);
+        set_extent_delalloc(&tmp, 4096, max_bytes - 1, NULL, GFP_KERNEL);
         start = test_start;
         end = 0;
         found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
@@ -199,7 +199,7 @@ static int test_find_delalloc(void)
          *
          * We are re-using our test_start from above since it works out well.
          */
-        set_extent_delalloc(&tmp, max_bytes, total_dirty - 1, NULL, GFP_NOFS);
+        set_extent_delalloc(&tmp, max_bytes, total_dirty - 1, NULL, GFP_KERNEL);
         start = test_start;
         end = 0;
         found = find_lock_delalloc_range(inode, &tmp, locked_page, &start,
@@ -262,7 +262,7 @@ static int test_find_delalloc(void)
         }
         ret = 0;
 out_bits:
-        clear_extent_bits(&tmp, 0, total_dirty - 1, (unsigned)-1, GFP_NOFS);
+        clear_extent_bits(&tmp, 0, total_dirty - 1, (unsigned)-1, GFP_KERNEL);
 out:
         if (locked_page)
                 page_cache_release(locked_page);
@@ -360,7 +360,7 @@ static int test_eb_bitmaps(void)
 
         test_msg("Running extent buffer bitmap tests\n");
 
-        bitmap = kmalloc(len, GFP_NOFS);
+        bitmap = kmalloc(len, GFP_KERNEL);
         if (!bitmap) {
                 test_msg("Couldn't allocate test bitmap\n");
                 return -ENOMEM;
diff --git a/fs/btrfs/tests/inode-tests.c b/fs/btrfs/tests/inode-tests.c
index 5de55fdd28bc..e2d3da02deee 100644
--- a/fs/btrfs/tests/inode-tests.c
+++ b/fs/btrfs/tests/inode-tests.c
@@ -974,7 +974,7 @@ static int test_extent_accounting(void)
                                (BTRFS_MAX_EXTENT_SIZE >> 1) + 4095,
                                EXTENT_DELALLOC | EXTENT_DIRTY |
                                EXTENT_UPTODATE | EXTENT_DO_ACCOUNTING, 0, 0,
-                               NULL, GFP_NOFS);
+                               NULL, GFP_KERNEL);
         if (ret) {
                 test_msg("clear_extent_bit returned %d\n", ret);
                 goto out;
@@ -1045,7 +1045,7 @@ static int test_extent_accounting(void)
                                BTRFS_MAX_EXTENT_SIZE+8191,
                                EXTENT_DIRTY | EXTENT_DELALLOC |
                                EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0,
-                               NULL, GFP_NOFS);
+                               NULL, GFP_KERNEL);
         if (ret) {
                 test_msg("clear_extent_bit returned %d\n", ret);
                 goto out;
@@ -1079,7 +1079,7 @@ static int test_extent_accounting(void)
         ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1,
                                EXTENT_DIRTY | EXTENT_DELALLOC |
                                EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0,
-                               NULL, GFP_NOFS);
+                               NULL, GFP_KERNEL);
         if (ret) {
                 test_msg("clear_extent_bit returned %d\n", ret);
                 goto out;
@@ -1096,7 +1096,7 @@ out:
                 clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1,
                                  EXTENT_DIRTY | EXTENT_DELALLOC |
                                  EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0,
-                                 NULL, GFP_NOFS);
+                                 NULL, GFP_KERNEL);
         iput(inode);
         btrfs_free_dummy_root(root);
         return ret;
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 323e12cc9d2f..978c3a810893 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -4127,7 +4127,9 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
                                      struct inode *inode,
                                      struct btrfs_path *path,
                                      struct list_head *logged_list,
-                                     struct btrfs_log_ctx *ctx)
+                                     struct btrfs_log_ctx *ctx,
+                                     const u64 start,
+                                     const u64 end)
 {
         struct extent_map *em, *n;
         struct list_head extents;
@@ -4166,7 +4168,13 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
         }
 
         list_sort(NULL, &extents, extent_cmp);
-
+        /*
+         * Collect any new ordered extents within the range. This is to
+         * prevent logging file extent items without waiting for the disk
+         * location they point to being written. We do this only to deal
+         * with races against concurrent lockless direct IO writes.
+         */
+        btrfs_get_logged_extents(inode, logged_list, start, end);
 process:
         while (!list_empty(&extents)) {
                 em = list_entry(extents.next, struct extent_map, list);
@@ -4701,7 +4709,7 @@ log_extents:
                         goto out_unlock;
                 }
                 ret = btrfs_log_changed_extents(trans, root, inode, dst_path,
-                                                &logged_list, ctx);
+                                                &logged_list, ctx, start, end);
                 if (ret) {
                         err = ret;
                         goto out_unlock;