Merge branch 'for-4.13-part1' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "The core updates improve error handling (mostly related to bios), with
  the usual incremental work on the GFP_NOFS (mis)use removal,
  refactoring or cleanups. Except the two top patches, all have been in
  for-next for an extensive amount of time.

  User visible changes:

   - statx support

   - quota override tunable

   - improved compression thresholds

   - obsoleted mount option alloc_start

  Core updates:

   - bio-related updates:
       - faster bio cloning
       - no allocation failures
       - preallocated flush bios

   - more kvzalloc use, memalloc_nofs protections, GFP_NOFS updates

   - prep work for btree_inode removal

   - dir-item validation

   - qgoup fixes and updates

   - cleanups:
       - removed unused struct members, unused code, refactoring
       - argument refactoring (fs_info/root, caller -> callee sink)
       - SEARCH_TREE ioctl docs"

* 'for-4.13-part1' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (115 commits)
  btrfs: Remove false alert when fiemap range is smaller than on-disk extent
  btrfs: Don't clear SGID when inheriting ACLs
  btrfs: fix integer overflow in calc_reclaim_items_nr
  btrfs: scrub: fix target device intialization while setting up scrub context
  btrfs: qgroup: Fix qgroup reserved space underflow by only freeing reserved ranges
  btrfs: qgroup: Introduce extent changeset for qgroup reserve functions
  btrfs: qgroup: Fix qgroup reserved space underflow caused by buffered write and quotas being enabled
  btrfs: qgroup: Return actually freed bytes for qgroup release or free data
  btrfs: qgroup: Cleanup btrfs_qgroup_prepare_account_extents function
  btrfs: qgroup: Add quick exit for non-fs extents
  Btrfs: rework delayed ref total_bytes_pinned accounting
  Btrfs: return old and new total ref mods when adding delayed refs
  Btrfs: always account pinned bytes when dropping a tree block ref
  Btrfs: update total_bytes_pinned when pinning down extents
  Btrfs: make BUG_ON() in add_pinned_bytes() an ASSERT()
  Btrfs: make add_pinned_bytes() take an s64 num_bytes instead of u64
  btrfs: fix validation of XATTR_ITEM dir items
  btrfs: Verify dir_item in iterate_object_props
  btrfs: Check name_len before in btrfs_del_root_ref
  btrfs: Check name_len before reading btrfs_get_name
  ...

fs/btrfs/acl.c

@@ -78,12 +78,6 @@ static int __btrfs_set_acl(struct btrfs_trans_handle *trans,
 	switch (type) {
 	case ACL_TYPE_ACCESS:
 		name = XATTR_NAME_POSIX_ACL_ACCESS;
-		if (acl) {
-			ret = posix_acl_update_mode(inode, &inode->i_mode, &acl);
-			if (ret)
-				return ret;
-		}
-		ret = 0;
 		break;
 	case ACL_TYPE_DEFAULT:
 		if (!S_ISDIR(inode->i_mode))
@@ -119,6 +113,13 @@ out:
 
 int btrfs_set_acl(struct inode *inode, struct posix_acl *acl, int type)
 {
+	int ret;
+
+	if (type == ACL_TYPE_ACCESS && acl) {
+		ret = posix_acl_update_mode(inode, &inode->i_mode, &acl);
+		if (ret)
+			return ret;
+	}
 	return __btrfs_set_acl(NULL, inode, acl, type);
 }
 

fs/btrfs/backref.c

@@ -16,7 +16,7 @@
  * Boston, MA 021110-1307, USA.
  */
 
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <linux/rbtree.h>
 #include "ctree.h"
 #include "disk-io.h"
@@ -2305,7 +2305,7 @@ struct btrfs_data_container *init_data_container(u32 total_bytes)
 	size_t alloc_bytes;
 
 	alloc_bytes = max_t(size_t, total_bytes, sizeof(*data));
-	data = vmalloc(alloc_bytes);
+	data = kvmalloc(alloc_bytes, GFP_KERNEL);
 	if (!data)
 		return ERR_PTR(-ENOMEM);
 
@@ -2339,9 +2339,9 @@ struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root,
 	if (IS_ERR(fspath))
 		return (void *)fspath;
 
-	ifp = kmalloc(sizeof(*ifp), GFP_NOFS);
+	ifp = kmalloc(sizeof(*ifp), GFP_KERNEL);
 	if (!ifp) {
-		vfree(fspath);
+		kvfree(fspath);
 		return ERR_PTR(-ENOMEM);
 	}
 
@@ -2356,6 +2356,6 @@ void free_ipath(struct inode_fs_paths *ipath)
 {
 	if (!ipath)
 		return;
-	vfree(ipath->fspath);
+	kvfree(ipath->fspath);
 	kfree(ipath);
 }

fs/btrfs/check-integrity.c

@@ -94,7 +94,7 @@
 #include <linux/mutex.h>
 #include <linux/genhd.h>
 #include <linux/blkdev.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <linux/string.h>
 #include "ctree.h"
 #include "disk-io.h"
@@ -1638,12 +1638,7 @@ static int btrfsic_read_block(struct btrfsic_state *state,
 		struct bio *bio;
 		unsigned int j;
 
-		bio = btrfs_io_bio_alloc(GFP_NOFS, num_pages - i);
-		if (!bio) {
-			pr_info("btrfsic: bio_alloc() for %u pages failed!\n",
-			       num_pages - i);
-			return -1;
-		}
+		bio = btrfs_io_bio_alloc(num_pages - i);
 		bio->bi_bdev = block_ctx->dev->bdev;
 		bio->bi_iter.bi_sector = dev_bytenr >> 9;
 		bio_set_op_attrs(bio, REQ_OP_READ, 0);
@@ -1668,14 +1663,8 @@ static int btrfsic_read_block(struct btrfsic_state *state,
 		dev_bytenr += (j - i) * PAGE_SIZE;
 		i = j;
 	}
-	for (i = 0; i < num_pages; i++) {
+	for (i = 0; i < num_pages; i++)
 		block_ctx->datav[i] = kmap(block_ctx->pagev[i]);
-		if (!block_ctx->datav[i]) {
-			pr_info("btrfsic: kmap() failed (dev %s)!\n",
-			       block_ctx->dev->name);
-			return -1;
-		}
-	}
 
 	return block_ctx->len;
 }
@@ -2822,44 +2811,47 @@ static void __btrfsic_submit_bio(struct bio *bio)
 	dev_state = btrfsic_dev_state_lookup(bio->bi_bdev);
 	if (NULL != dev_state &&
 	    (bio_op(bio) == REQ_OP_WRITE) && bio_has_data(bio)) {
-		unsigned int i;
+		unsigned int i = 0;
 		u64 dev_bytenr;
 		u64 cur_bytenr;
-		struct bio_vec *bvec;
+		struct bio_vec bvec;
+		struct bvec_iter iter;
 		int bio_is_patched;
 		char **mapped_datav;
+		unsigned int segs = bio_segments(bio);
 
 		dev_bytenr = 512 * bio->bi_iter.bi_sector;
 		bio_is_patched = 0;
 		if (dev_state->state->print_mask &
 		    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
 			pr_info("submit_bio(rw=%d,0x%x, bi_vcnt=%u, bi_sector=%llu (bytenr %llu), bi_bdev=%p)\n",
-			       bio_op(bio), bio->bi_opf, bio->bi_vcnt,
+			       bio_op(bio), bio->bi_opf, segs,
 			       (unsigned long long)bio->bi_iter.bi_sector,
 			       dev_bytenr, bio->bi_bdev);
 
-		mapped_datav = kmalloc_array(bio->bi_vcnt,
+		mapped_datav = kmalloc_array(segs,
 					     sizeof(*mapped_datav), GFP_NOFS);
 		if (!mapped_datav)
 			goto leave;
 		cur_bytenr = dev_bytenr;
 
-		bio_for_each_segment_all(bvec, bio, i) {
-			BUG_ON(bvec->bv_len != PAGE_SIZE);
-			mapped_datav[i] = kmap(bvec->bv_page);
+		bio_for_each_segment(bvec, bio, iter) {
+			BUG_ON(bvec.bv_len != PAGE_SIZE);
+			mapped_datav[i] = kmap(bvec.bv_page);
+			i++;
 
 			if (dev_state->state->print_mask &
 			    BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH_VERBOSE)
 				pr_info("#%u: bytenr=%llu, len=%u, offset=%u\n",
-				       i, cur_bytenr, bvec->bv_len, bvec->bv_offset);
-			cur_bytenr += bvec->bv_len;
+				       i, cur_bytenr, bvec.bv_len, bvec.bv_offset);
+			cur_bytenr += bvec.bv_len;
 		}
 		btrfsic_process_written_block(dev_state, dev_bytenr,
-					      mapped_datav, bio->bi_vcnt,
+					      mapped_datav, segs,
 					      bio, &bio_is_patched,
 					      NULL, bio->bi_opf);
-		bio_for_each_segment_all(bvec, bio, i)
-			kunmap(bvec->bv_page);
+		bio_for_each_segment(bvec, bio, iter)
+			kunmap(bvec.bv_page);
 		kfree(mapped_datav);
 	} else if (NULL != dev_state && (bio->bi_opf & REQ_PREFLUSH)) {
 		if (dev_state->state->print_mask &
@@ -2923,13 +2915,10 @@ int btrfsic_mount(struct btrfs_fs_info *fs_info,
 		       fs_info->sectorsize, PAGE_SIZE);
 		return -1;
 	}
-	state = kzalloc(sizeof(*state), GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
+	state = kvzalloc(sizeof(*state), GFP_KERNEL);
 	if (!state) {
-		state = vzalloc(sizeof(*state));
-		if (!state) {
-			pr_info("btrfs check-integrity: vzalloc() failed!\n");
-			return -1;
-		}
+		pr_info("btrfs check-integrity: allocation failed!\n");
+		return -1;
 	}
 
 	if (!btrfsic_is_initialized) {

fs/btrfs/compression.c

@@ -32,6 +32,7 @@
 #include <linux/writeback.h>
 #include <linux/bit_spinlock.h>
 #include <linux/slab.h>
+#include <linux/sched/mm.h>
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -42,48 +43,7 @@
 #include "extent_io.h"
 #include "extent_map.h"
 
-struct compressed_bio {
-	/* number of bios pending for this compressed extent */
-	refcount_t pending_bios;
-
-	/* the pages with the compressed data on them */
-	struct page **compressed_pages;
-
-	/* inode that owns this data */
-	struct inode *inode;
-
-	/* starting offset in the inode for our pages */
-	u64 start;
-
-	/* number of bytes in the inode we're working on */
-	unsigned long len;
-
-	/* number of bytes on disk */
-	unsigned long compressed_len;
-
-	/* the compression algorithm for this bio */
-	int compress_type;
-
-	/* number of compressed pages in the array */
-	unsigned long nr_pages;
-
-	/* IO errors */
-	int errors;
-	int mirror_num;
-
-	/* for reads, this is the bio we are copying the data into */
-	struct bio *orig_bio;
-
-	/*
-	 * the start of a variable length array of checksums only
-	 * used by reads
-	 */
-	u32 sums;
-};
-
-static int btrfs_decompress_bio(int type, struct page **pages_in,
-				   u64 disk_start, struct bio *orig_bio,
-				   size_t srclen);
+static int btrfs_decompress_bio(struct compressed_bio *cb);
 
 static inline int compressed_bio_size(struct btrfs_fs_info *fs_info,
 				      unsigned long disk_size)
@@ -94,12 +54,6 @@ static inline int compressed_bio_size(struct btrfs_fs_info *fs_info,
 		(DIV_ROUND_UP(disk_size, fs_info->sectorsize)) * csum_size;
 }
 
-static struct bio *compressed_bio_alloc(struct block_device *bdev,
-					u64 first_byte, gfp_t gfp_flags)
-{
-	return btrfs_bio_alloc(bdev, first_byte >> 9, BIO_MAX_PAGES, gfp_flags);
-}
-
 static int check_compressed_csum(struct btrfs_inode *inode,
 				 struct compressed_bio *cb,
 				 u64 disk_start)
@@ -173,11 +127,8 @@ static void end_compressed_bio_read(struct bio *bio)
 	/* ok, we're the last bio for this extent, lets start
 	 * the decompression.
 	 */
-	ret = btrfs_decompress_bio(cb->compress_type,
-				      cb->compressed_pages,
-				      cb->start,
-				      cb->orig_bio,
-				      cb->compressed_len);
+	ret = btrfs_decompress_bio(cb);
+
 csum_failed:
 	if (ret)
 		cb->errors = 1;
@@ -355,11 +306,7 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start,
 
 	bdev = fs_info->fs_devices->latest_bdev;
 
-	bio = compressed_bio_alloc(bdev, first_byte, GFP_NOFS);
-	if (!bio) {
-		kfree(cb);
-		return BLK_STS_RESOURCE;
-	}
+	bio = btrfs_bio_alloc(bdev, first_byte);
 	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
 	bio->bi_private = cb;
 	bio->bi_end_io = end_compressed_bio_write;
@@ -406,8 +353,7 @@ blk_status_t btrfs_submit_compressed_write(struct inode *inode, u64 start,
 
 			bio_put(bio);
 
-			bio = compressed_bio_alloc(bdev, first_byte, GFP_NOFS);
-			BUG_ON(!bio);
+			bio = btrfs_bio_alloc(bdev, first_byte);
 			bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
 			bio->bi_private = cb;
 			bio->bi_end_io = end_compressed_bio_write;
@@ -650,9 +596,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	/* include any pages we added in add_ra-bio_pages */
 	cb->len = bio->bi_iter.bi_size;
 
-	comp_bio = compressed_bio_alloc(bdev, cur_disk_byte, GFP_NOFS);
-	if (!comp_bio)
-		goto fail2;
+	comp_bio = btrfs_bio_alloc(bdev, cur_disk_byte);
 	bio_set_op_attrs (comp_bio, REQ_OP_READ, 0);
 	comp_bio->bi_private = cb;
 	comp_bio->bi_end_io = end_compressed_bio_read;
@@ -703,9 +647,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 
 			bio_put(comp_bio);
 
-			comp_bio = compressed_bio_alloc(bdev, cur_disk_byte,
-							GFP_NOFS);
-			BUG_ON(!comp_bio);
+			comp_bio = btrfs_bio_alloc(bdev, cur_disk_byte);
 			bio_set_op_attrs(comp_bio, REQ_OP_READ, 0);
 			comp_bio->bi_private = cb;
 			comp_bio->bi_end_io = end_compressed_bio_read;
@@ -801,6 +743,7 @@ static struct list_head *find_workspace(int type)
 	struct list_head *workspace;
 	int cpus = num_online_cpus();
 	int idx = type - 1;
+	unsigned nofs_flag;
 
 	struct list_head *idle_ws	= &btrfs_comp_ws[idx].idle_ws;
 	spinlock_t *ws_lock		= &btrfs_comp_ws[idx].ws_lock;
@@ -830,7 +773,15 @@ again:
 	atomic_inc(total_ws);
 	spin_unlock(ws_lock);
 
+	/*
+	 * Allocation helpers call vmalloc that can't use GFP_NOFS, so we have
+	 * to turn it off here because we might get called from the restricted
+	 * context of btrfs_compress_bio/btrfs_compress_pages
+	 */
+	nofs_flag = memalloc_nofs_save();
 	workspace = btrfs_compress_op[idx]->alloc_workspace();
+	memalloc_nofs_restore(nofs_flag);
+
 	if (IS_ERR(workspace)) {
 		atomic_dec(total_ws);
 		wake_up(ws_wait);
@@ -961,19 +912,16 @@ int btrfs_compress_pages(int type, struct address_space *mapping,
  * be contiguous.  They all correspond to the range of bytes covered by
  * the compressed extent.
  */
-static int btrfs_decompress_bio(int type, struct page **pages_in,
-				   u64 disk_start, struct bio *orig_bio,
-				   size_t srclen)
+static int btrfs_decompress_bio(struct compressed_bio *cb)
 {
 	struct list_head *workspace;
 	int ret;
+	int type = cb->compress_type;
 
 	workspace = find_workspace(type);
-
-	ret = btrfs_compress_op[type-1]->decompress_bio(workspace, pages_in,
-							 disk_start, orig_bio,
-							 srclen);
+	ret = btrfs_compress_op[type - 1]->decompress_bio(workspace, cb);
 	free_workspace(type, workspace);
+
 	return ret;
 }
 

fs/btrfs/compression.h

@@ -34,6 +34,45 @@
 /* Maximum size of data before compression */
 #define BTRFS_MAX_UNCOMPRESSED		(SZ_128K)
 
+struct compressed_bio {
+	/* number of bios pending for this compressed extent */
+	refcount_t pending_bios;
+
+	/* the pages with the compressed data on them */
+	struct page **compressed_pages;
+
+	/* inode that owns this data */
+	struct inode *inode;
+
+	/* starting offset in the inode for our pages */
+	u64 start;
+
+	/* number of bytes in the inode we're working on */
+	unsigned long len;
+
+	/* number of bytes on disk */
+	unsigned long compressed_len;
+
+	/* the compression algorithm for this bio */
+	int compress_type;
+
+	/* number of compressed pages in the array */
+	unsigned long nr_pages;
+
+	/* IO errors */
+	int errors;
+	int mirror_num;
+
+	/* for reads, this is the bio we are copying the data into */
+	struct bio *orig_bio;
+
+	/*
+	 * the start of a variable length array of checksums only
+	 * used by reads
+	 */
+	u32 sums;
+};
+
 void btrfs_init_compress(void);
 void btrfs_exit_compress(void);
 
@@ -78,10 +117,7 @@ struct btrfs_compress_op {
 			      unsigned long *total_out);
 
 	int (*decompress_bio)(struct list_head *workspace,
-				 struct page **pages_in,
-				 u64 disk_start,
-				 struct bio *orig_bio,
-				 size_t srclen);
+				struct compressed_bio *cb);
 
 	int (*decompress)(struct list_head *workspace,
 			  unsigned char *data_in,

fs/btrfs/ctree.c

@@ -19,7 +19,7 @@
 #include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/rbtree.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include "ctree.h"
 #include "disk-io.h"
 #include "transaction.h"
@@ -3667,14 +3667,14 @@ static noinline int __push_leaf_right(struct btrfs_fs_info *fs_info,
 	/* make room in the right data area */
 	data_end = leaf_data_end(fs_info, right);
 	memmove_extent_buffer(right,
-			      btrfs_leaf_data(right) + data_end - push_space,
-			      btrfs_leaf_data(right) + data_end,
+			      BTRFS_LEAF_DATA_OFFSET + data_end - push_space,
+			      BTRFS_LEAF_DATA_OFFSET + data_end,
 			      BTRFS_LEAF_DATA_SIZE(fs_info) - data_end);
 
 	/* copy from the left data area */
-	copy_extent_buffer(right, left, btrfs_leaf_data(right) +
+	copy_extent_buffer(right, left, BTRFS_LEAF_DATA_OFFSET +
 		     BTRFS_LEAF_DATA_SIZE(fs_info) - push_space,
-		     btrfs_leaf_data(left) + leaf_data_end(fs_info, left),
+		     BTRFS_LEAF_DATA_OFFSET + leaf_data_end(fs_info, left),
 		     push_space);
 
 	memmove_extent_buffer(right, btrfs_item_nr_offset(push_items),
@@ -3888,9 +3888,9 @@ static noinline int __push_leaf_left(struct btrfs_fs_info *fs_info,
 	push_space = BTRFS_LEAF_DATA_SIZE(fs_info) -
 		     btrfs_item_offset_nr(right, push_items - 1);
 
-	copy_extent_buffer(left, right, btrfs_leaf_data(left) +
+	copy_extent_buffer(left, right, BTRFS_LEAF_DATA_OFFSET +
 		     leaf_data_end(fs_info, left) - push_space,
-		     btrfs_leaf_data(right) +
+		     BTRFS_LEAF_DATA_OFFSET +
 		     btrfs_item_offset_nr(right, push_items - 1),
 		     push_space);
 	old_left_nritems = btrfs_header_nritems(left);
@@ -3917,9 +3917,9 @@ static noinline int __push_leaf_left(struct btrfs_fs_info *fs_info,
 	if (push_items < right_nritems) {
 		push_space = btrfs_item_offset_nr(right, push_items - 1) -
 						  leaf_data_end(fs_info, right);
-		memmove_extent_buffer(right, btrfs_leaf_data(right) +
+		memmove_extent_buffer(right, BTRFS_LEAF_DATA_OFFSET +
 				      BTRFS_LEAF_DATA_SIZE(fs_info) - push_space,
-				      btrfs_leaf_data(right) +
+				      BTRFS_LEAF_DATA_OFFSET +
 				      leaf_data_end(fs_info, right), push_space);
 
 		memmove_extent_buffer(right, btrfs_item_nr_offset(0),
@@ -4069,8 +4069,8 @@ static noinline void copy_for_split(struct btrfs_trans_handle *trans,
 			   nritems * sizeof(struct btrfs_item));
 
 	copy_extent_buffer(right, l,
-		     btrfs_leaf_data(right) + BTRFS_LEAF_DATA_SIZE(fs_info) -
-		     data_copy_size, btrfs_leaf_data(l) +
+		     BTRFS_LEAF_DATA_OFFSET + BTRFS_LEAF_DATA_SIZE(fs_info) -
+		     data_copy_size, BTRFS_LEAF_DATA_OFFSET +
 		     leaf_data_end(fs_info, l), data_copy_size);
 
 	rt_data_off = BTRFS_LEAF_DATA_SIZE(fs_info) - btrfs_item_end_nr(l, mid);
@@ -4607,8 +4607,8 @@ void btrfs_truncate_item(struct btrfs_fs_info *fs_info,
 
 	/* shift the data */
 	if (from_end) {
-		memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
-			      data_end + size_diff, btrfs_leaf_data(leaf) +
+		memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
+			      data_end + size_diff, BTRFS_LEAF_DATA_OFFSET +
 			      data_end, old_data_start + new_size - data_end);
 	} else {
 		struct btrfs_disk_key disk_key;
@@ -4634,8 +4634,8 @@ void btrfs_truncate_item(struct btrfs_fs_info *fs_info,
 			}
 		}
 
-		memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
-			      data_end + size_diff, btrfs_leaf_data(leaf) +
+		memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
+			      data_end + size_diff, BTRFS_LEAF_DATA_OFFSET +
 			      data_end, old_data_start - data_end);
 
 		offset = btrfs_disk_key_offset(&disk_key);
@@ -4707,8 +4707,8 @@ void btrfs_extend_item(struct btrfs_fs_info *fs_info, struct btrfs_path *path,
 	}
 
 	/* shift the data */
-	memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
-		      data_end - data_size, btrfs_leaf_data(leaf) +
+	memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
+		      data_end - data_size, BTRFS_LEAF_DATA_OFFSET +
 		      data_end, old_data - data_end);
 
 	data_end = old_data;
@@ -4790,8 +4790,8 @@ void setup_items_for_insert(struct btrfs_root *root, struct btrfs_path *path,
 			      (nritems - slot) * sizeof(struct btrfs_item));
 
 		/* shift the data */
-		memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
-			      data_end - total_data, btrfs_leaf_data(leaf) +
+		memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
+			      data_end - total_data, BTRFS_LEAF_DATA_OFFSET +
 			      data_end, old_data - data_end);
 		data_end = old_data;
 	}
@@ -4983,9 +4983,9 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	if (slot + nr != nritems) {
 		int data_end = leaf_data_end(fs_info, leaf);
 
-		memmove_extent_buffer(leaf, btrfs_leaf_data(leaf) +
+		memmove_extent_buffer(leaf, BTRFS_LEAF_DATA_OFFSET +
 			      data_end + dsize,
-			      btrfs_leaf_data(leaf) + data_end,
+			      BTRFS_LEAF_DATA_OFFSET + data_end,
 			      last_off - data_end);
 
 		for (i = slot + nr; i < nritems; i++) {

fs/btrfs/ctree.h

@@ -48,7 +48,6 @@ struct btrfs_trans_handle;
 struct btrfs_transaction;
 struct btrfs_pending_snapshot;
 extern struct kmem_cache *btrfs_trans_handle_cachep;
-extern struct kmem_cache *btrfs_transaction_cachep;
 extern struct kmem_cache *btrfs_bit_radix_cachep;
 extern struct kmem_cache *btrfs_path_cachep;
 extern struct kmem_cache *btrfs_free_space_cachep;
@@ -716,6 +715,10 @@ struct btrfs_delayed_root;
 #define BTRFS_FS_BTREE_ERR			11
 #define BTRFS_FS_LOG1_ERR			12
 #define BTRFS_FS_LOG2_ERR			13
+#define BTRFS_FS_QUOTA_OVERRIDE			14
+/* Used to record internally whether fs has been frozen */
+#define BTRFS_FS_FROZEN				15
+
 /*
  * Indicate that a whole-filesystem exclusive operation is running
  * (device replace, resize, device add/delete, balance)
@@ -748,8 +751,7 @@ struct btrfs_fs_info {
 	struct rb_root block_group_cache_tree;
 
 	/* keep track of unallocated space */
-	spinlock_t free_chunk_lock;
-	u64 free_chunk_space;
+	atomic64_t free_chunk_space;
 
 	struct extent_io_tree freed_extents[2];
 	struct extent_io_tree *pinned_extents;
@@ -797,17 +799,7 @@ struct btrfs_fs_info {
 	 * so it is also safe.
 	 */
 	u64 max_inline;
-	/*
-	 * Protected by ->chunk_mutex and sb->s_umount.
-	 *
-	 * The reason that we use two lock to protect it is because only
-	 * remount and mount operations can change it and these two operations
-	 * are under sb->s_umount, but the read side (chunk allocation) can not
-	 * acquire sb->s_umount or the deadlock would happen. So we use two
-	 * locks to protect it. On the write side, we must acquire two locks,
-	 * and on the read side, we just need acquire one of them.
-	 */
-	u64 alloc_start;
+
 	struct btrfs_transaction *running_transaction;
 	wait_queue_head_t transaction_throttle;
 	wait_queue_head_t transaction_wait;
@@ -1107,9 +1099,6 @@ struct btrfs_fs_info {
 	 */
 	struct list_head pinned_chunks;
 
-	/* Used to record internally whether fs has been frozen */
-	int fs_frozen;
-
 	/* Cached block sizes */
 	u32 nodesize;
 	u32 sectorsize;
@@ -1277,21 +1266,20 @@ struct btrfs_root {
 	/* For qgroup metadata space reserve */
 	atomic64_t qgroup_meta_rsv;
 };
+
 static inline u32 btrfs_inode_sectorsize(const struct inode *inode)
 {
 	return btrfs_sb(inode->i_sb)->sectorsize;
 }
 
-static inline u32 __BTRFS_LEAF_DATA_SIZE(u32 blocksize)
-{
-	return blocksize - sizeof(struct btrfs_header);
-}
-
 static inline u32 BTRFS_LEAF_DATA_SIZE(const struct btrfs_fs_info *info)
 {
-	return __BTRFS_LEAF_DATA_SIZE(info->nodesize);
+
+	return info->nodesize - sizeof(struct btrfs_header);
 }
 
+#define BTRFS_LEAF_DATA_OFFSET		offsetof(struct btrfs_leaf, items)
+
 static inline u32 BTRFS_MAX_ITEM_SIZE(const struct btrfs_fs_info *info)
 {
 	return BTRFS_LEAF_DATA_SIZE(info) - sizeof(struct btrfs_item);
@@ -1553,8 +1541,27 @@ static inline void btrfs_set_##name(type *s, u##bits val)		\
 	s->member = cpu_to_le##bits(val);				\
 }
 
+
+static inline u64 btrfs_device_total_bytes(struct extent_buffer *eb,
+					   struct btrfs_dev_item *s)
+{
+	BUILD_BUG_ON(sizeof(u64) !=
+		     sizeof(((struct btrfs_dev_item *)0))->total_bytes);
+	return btrfs_get_64(eb, s, offsetof(struct btrfs_dev_item,
+					    total_bytes));
+}
+static inline void btrfs_set_device_total_bytes(struct extent_buffer *eb,
+						struct btrfs_dev_item *s,
+						u64 val)
+{
+	BUILD_BUG_ON(sizeof(u64) !=
+		     sizeof(((struct btrfs_dev_item *)0))->total_bytes);
+	WARN_ON(!IS_ALIGNED(val, eb->fs_info->sectorsize));
+	btrfs_set_64(eb, s, offsetof(struct btrfs_dev_item, total_bytes), val);
+}
+
+
 BTRFS_SETGET_FUNCS(device_type, struct btrfs_dev_item, type, 64);
-BTRFS_SETGET_FUNCS(device_total_bytes, struct btrfs_dev_item, total_bytes, 64);
 BTRFS_SETGET_FUNCS(device_bytes_used, struct btrfs_dev_item, bytes_used, 64);
 BTRFS_SETGET_FUNCS(device_io_align, struct btrfs_dev_item, io_align, 32);
 BTRFS_SETGET_FUNCS(device_io_width, struct btrfs_dev_item, io_width, 32);
@@ -2324,10 +2331,6 @@ static inline int btrfs_super_csum_size(struct btrfs_super_block *s)
 	return btrfs_csum_sizes[t];
 }
 
-static inline unsigned long btrfs_leaf_data(struct extent_buffer *l)
-{
-	return offsetof(struct btrfs_leaf, items);
-}
 
 /*
  * The leaf data grows from end-to-front in the node.
@@ -2538,11 +2541,11 @@ BTRFS_SETGET_STACK_FUNCS(stack_dev_replace_cursor_right,
 
 /* helper function to cast into the data area of the leaf. */
 #define btrfs_item_ptr(leaf, slot, type) \
-	((type *)(btrfs_leaf_data(leaf) + \
+	((type *)(BTRFS_LEAF_DATA_OFFSET + \
 	btrfs_item_offset_nr(leaf, slot)))
 
 #define btrfs_item_ptr_offset(leaf, slot) \
-	((unsigned long)(btrfs_leaf_data(leaf) + \
+	((unsigned long)(BTRFS_LEAF_DATA_OFFSET + \
 	btrfs_item_offset_nr(leaf, slot)))
 
 static inline bool btrfs_mixed_space_info(struct btrfs_space_info *space_info)
@@ -2680,7 +2683,9 @@ void btrfs_get_block_group_trimming(struct btrfs_block_group_cache *cache);
 void btrfs_put_block_group_trimming(struct btrfs_block_group_cache *cache);
 void btrfs_create_pending_block_groups(struct btrfs_trans_handle *trans,
 				       struct btrfs_fs_info *fs_info);
-u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data);
+u64 btrfs_data_alloc_profile(struct btrfs_fs_info *fs_info);
+u64 btrfs_metadata_alloc_profile(struct btrfs_fs_info *fs_info);
+u64 btrfs_system_alloc_profile(struct btrfs_fs_info *fs_info);
 void btrfs_clear_space_info_full(struct btrfs_fs_info *info);
 
 enum btrfs_reserve_flush_enum {
@@ -2703,9 +2708,13 @@ enum btrfs_flush_state {
 	COMMIT_TRANS		=	6,
 };
 
-int btrfs_check_data_free_space(struct inode *inode, u64 start, u64 len);
 int btrfs_alloc_data_chunk_ondemand(struct btrfs_inode *inode, u64 bytes);
-void btrfs_free_reserved_data_space(struct inode *inode, u64 start, u64 len);
+int btrfs_check_data_free_space(struct inode *inode,
+			struct extent_changeset **reserved, u64 start, u64 len);
+void btrfs_free_reserved_data_space(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len);
+void btrfs_delalloc_release_space(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len);
 void btrfs_free_reserved_data_space_noquota(struct inode *inode, u64 start,
 					    u64 len);
 void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans,
@@ -2722,8 +2731,8 @@ void btrfs_subvolume_release_metadata(struct btrfs_fs_info *fs_info,
 				      struct btrfs_block_rsv *rsv);
 int btrfs_delalloc_reserve_metadata(struct btrfs_inode *inode, u64 num_bytes);
 void btrfs_delalloc_release_metadata(struct btrfs_inode *inode, u64 num_bytes);
-int btrfs_delalloc_reserve_space(struct inode *inode, u64 start, u64 len);
-void btrfs_delalloc_release_space(struct inode *inode, u64 start, u64 len);
+int btrfs_delalloc_reserve_space(struct inode *inode,
+			struct extent_changeset **reserved, u64 start, u64 len);
 void btrfs_init_block_rsv(struct btrfs_block_rsv *rsv, unsigned short type);
 struct btrfs_block_rsv *btrfs_alloc_block_rsv(struct btrfs_fs_info *fs_info,
 					      unsigned short type);
@@ -3031,12 +3040,14 @@ struct btrfs_dir_item *btrfs_lookup_xattr(struct btrfs_trans_handle *trans,
 					  const char *name, u16 name_len,
 					  int mod);
 int verify_dir_item(struct btrfs_fs_info *fs_info,
-		    struct extent_buffer *leaf,
+		    struct extent_buffer *leaf, int slot,
 		    struct btrfs_dir_item *dir_item);
 struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_fs_info *fs_info,
 						 struct btrfs_path *path,
 						 const char *name,
 						 int name_len);
+bool btrfs_is_name_len_valid(struct extent_buffer *leaf, int slot,
+			     unsigned long start, u16 name_len);
 
 /* orphan.c */
 int btrfs_insert_orphan_item(struct btrfs_trans_handle *trans,
@@ -3171,6 +3182,7 @@ int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
 int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
 			 size_t size, struct bio *bio,
 			 unsigned long bio_flags);
+void btrfs_set_range_writeback(void *private_data, u64 start, u64 end);
 int btrfs_page_mkwrite(struct vm_fault *vmf);
 int btrfs_readpage(struct file *file, struct page *page);
 void btrfs_evict_inode(struct inode *inode);

fs/btrfs/delayed-ref.c

@@ -470,7 +470,8 @@ add_tail:
 static noinline void
 update_existing_head_ref(struct btrfs_delayed_ref_root *delayed_refs,
 			 struct btrfs_delayed_ref_node *existing,
-			 struct btrfs_delayed_ref_node *update)
+			 struct btrfs_delayed_ref_node *update,
+			 int *old_ref_mod_ret)
 {
 	struct btrfs_delayed_ref_head *existing_ref;
 	struct btrfs_delayed_ref_head *ref;
@@ -523,6 +524,8 @@ update_existing_head_ref(struct btrfs_delayed_ref_root *delayed_refs,
 	 * currently, for refs we just added we know we're a-ok.
 	 */
 	old_ref_mod = existing_ref->total_ref_mod;
+	if (old_ref_mod_ret)
+		*old_ref_mod_ret = old_ref_mod;
 	existing->ref_mod += update->ref_mod;
 	existing_ref->total_ref_mod += update->ref_mod;
 
@@ -550,7 +553,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		     struct btrfs_delayed_ref_node *ref,
 		     struct btrfs_qgroup_extent_record *qrecord,
 		     u64 bytenr, u64 num_bytes, u64 ref_root, u64 reserved,
-		     int action, int is_data, int *qrecord_inserted_ret)
+		     int action, int is_data, int *qrecord_inserted_ret,
+		     int *old_ref_mod, int *new_ref_mod)
 {
 	struct btrfs_delayed_ref_head *existing;
 	struct btrfs_delayed_ref_head *head_ref = NULL;
@@ -638,7 +642,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	if (existing) {
 		WARN_ON(ref_root && reserved && existing->qgroup_ref_root
 			&& existing->qgroup_reserved);
-		update_existing_head_ref(delayed_refs, &existing->node, ref);
+		update_existing_head_ref(delayed_refs, &existing->node, ref,
+					 old_ref_mod);
 		/*
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
@@ -646,6 +651,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref);
 		head_ref = existing;
 	} else {
+		if (old_ref_mod)
+			*old_ref_mod = 0;
 		if (is_data && count_mod < 0)
 			delayed_refs->pending_csums += num_bytes;
 		delayed_refs->num_heads++;
@@ -655,6 +662,8 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 	}
 	if (qrecord_inserted_ret)
 		*qrecord_inserted_ret = qrecord_inserted;
+	if (new_ref_mod)
+		*new_ref_mod = head_ref->total_ref_mod;
 	return head_ref;
 }
 
@@ -778,7 +787,8 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes, u64 parent,
 			       u64 ref_root,  int level, int action,
-			       struct btrfs_delayed_extent_op *extent_op)
+			       struct btrfs_delayed_extent_op *extent_op,
+			       int *old_ref_mod, int *new_ref_mod)
 {
 	struct btrfs_delayed_tree_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
@@ -813,7 +823,8 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record,
 					bytenr, num_bytes, 0, 0, action, 0,
-					&qrecord_inserted);
+					&qrecord_inserted, old_ref_mod,
+					new_ref_mod);
 
 	add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 			     num_bytes, parent, ref_root, level, action);
@@ -838,7 +849,8 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes,
 			       u64 parent, u64 ref_root,
-			       u64 owner, u64 offset, u64 reserved, int action)
+			       u64 owner, u64 offset, u64 reserved, int action,
+			       int *old_ref_mod, int *new_ref_mod)
 {
 	struct btrfs_delayed_data_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
@@ -878,7 +890,8 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	 */
 	head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record,
 					bytenr, num_bytes, ref_root, reserved,
-					action, 1, &qrecord_inserted);
+					action, 1, &qrecord_inserted,
+					old_ref_mod, new_ref_mod);
 
 	add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, owner, offset,
@@ -909,7 +922,7 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 
 	add_delayed_ref_head(fs_info, trans, &head_ref->node, NULL, bytenr,
 			     num_bytes, 0, 0, BTRFS_UPDATE_DELAYED_HEAD,
-			     extent_op->is_data, NULL);
+			     extent_op->is_data, NULL, NULL, NULL);
 
 	spin_unlock(&delayed_refs->lock);
 	return 0;

fs/btrfs/delayed-ref.h

@@ -247,12 +247,14 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes, u64 parent,
 			       u64 ref_root, int level, int action,
-			       struct btrfs_delayed_extent_op *extent_op);
+			       struct btrfs_delayed_extent_op *extent_op,
+			       int *old_ref_mod, int *new_ref_mod);
 int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 			       struct btrfs_trans_handle *trans,
 			       u64 bytenr, u64 num_bytes,
 			       u64 parent, u64 ref_root,
-			       u64 owner, u64 offset, u64 reserved, int action);
+			       u64 owner, u64 offset, u64 reserved, int action,
+			       int *old_ref_mod, int *new_ref_mod);
 int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 				struct btrfs_trans_handle *trans,
 				u64 bytenr, u64 num_bytes,

fs/btrfs/dev-replace.c

@@ -388,7 +388,7 @@ int btrfs_dev_replace_start(struct btrfs_fs_info *fs_info,
 	if (ret)
 		btrfs_err(fs_info, "kobj add dev failed %d", ret);
 
-	btrfs_wait_ordered_roots(fs_info, -1, 0, (u64)-1);
+	btrfs_wait_ordered_roots(fs_info, U64_MAX, 0, (u64)-1);
 
 	/* force writing the updated state information to disk */
 	trans = btrfs_start_transaction(root, 0);
@@ -507,7 +507,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info,
 		mutex_unlock(&dev_replace->lock_finishing_cancel_unmount);
 		return ret;
 	}
-	btrfs_wait_ordered_roots(fs_info, -1, 0, (u64)-1);
+	btrfs_wait_ordered_roots(fs_info, U64_MAX, 0, (u64)-1);
 
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {

fs/btrfs/dir-item.c

@@ -395,8 +395,6 @@ struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_fs_info *fs_info,
 
 	leaf = path->nodes[0];
 	dir_item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dir_item);
-	if (verify_dir_item(fs_info, leaf, dir_item))
-		return NULL;
 
 	total_len = btrfs_item_size_nr(leaf, path->slots[0]);
 	while (cur < total_len) {
@@ -405,6 +403,8 @@ struct btrfs_dir_item *btrfs_match_dir_item_name(struct btrfs_fs_info *fs_info,
 			btrfs_dir_data_len(leaf, dir_item);
 		name_ptr = (unsigned long)(dir_item + 1);
 
+		if (verify_dir_item(fs_info, leaf, path->slots[0], dir_item))
+			return NULL;
 		if (btrfs_dir_name_len(leaf, dir_item) == name_len &&
 		    memcmp_extent_buffer(leaf, name, name_ptr, name_len) == 0)
 			return dir_item;
@@ -453,9 +453,11 @@ int btrfs_delete_one_dir_name(struct btrfs_trans_handle *trans,
 
 int verify_dir_item(struct btrfs_fs_info *fs_info,
 		    struct extent_buffer *leaf,
+		    int slot,
 		    struct btrfs_dir_item *dir_item)
 {
 	u16 namelen = BTRFS_NAME_LEN;
+	int ret;
 	u8 type = btrfs_dir_type(leaf, dir_item);
 
 	if (type >= BTRFS_FT_MAX) {
@@ -472,6 +474,12 @@ int verify_dir_item(struct btrfs_fs_info *fs_info,
 		return 1;
 	}
 
+	namelen = btrfs_dir_name_len(leaf, dir_item);
+	ret = btrfs_is_name_len_valid(leaf, slot,
+				      (unsigned long)(dir_item + 1), namelen);
+	if (!ret)
+		return 1;
+
 	/* BTRFS_MAX_XATTR_SIZE is the same for all dir items */
 	if ((btrfs_dir_data_len(leaf, dir_item) +
 	     btrfs_dir_name_len(leaf, dir_item)) >
@@ -484,3 +492,67 @@ int verify_dir_item(struct btrfs_fs_info *fs_info,
 
 	return 0;
 }
+
+bool btrfs_is_name_len_valid(struct extent_buffer *leaf, int slot,
+			     unsigned long start, u16 name_len)
+{
+	struct btrfs_fs_info *fs_info = leaf->fs_info;
+	struct btrfs_key key;
+	u32 read_start;
+	u32 read_end;
+	u32 item_start;
+	u32 item_end;
+	u32 size;
+	bool ret = true;
+
+	ASSERT(start > BTRFS_LEAF_DATA_OFFSET);
+
+	read_start = start - BTRFS_LEAF_DATA_OFFSET;
+	read_end = read_start + name_len;
+	item_start = btrfs_item_offset_nr(leaf, slot);
+	item_end = btrfs_item_end_nr(leaf, slot);
+
+	btrfs_item_key_to_cpu(leaf, &key, slot);
+
+	switch (key.type) {
+	case BTRFS_DIR_ITEM_KEY:
+	case BTRFS_XATTR_ITEM_KEY:
+	case BTRFS_DIR_INDEX_KEY:
+		size = sizeof(struct btrfs_dir_item);
+		break;
+	case BTRFS_INODE_REF_KEY:
+		size = sizeof(struct btrfs_inode_ref);
+		break;
+	case BTRFS_INODE_EXTREF_KEY:
+		size = sizeof(struct btrfs_inode_extref);
+		break;
+	case BTRFS_ROOT_REF_KEY:
+	case BTRFS_ROOT_BACKREF_KEY:
+		size = sizeof(struct btrfs_root_ref);
+		break;
+	default:
+		ret = false;
+		goto out;
+	}
+
+	if (read_start < item_start) {
+		ret = false;
+		goto out;
+	}
+	if (read_end > item_end) {
+		ret = false;
+		goto out;
+	}
+
+	/* there shall be item(s) before name */
+	if (read_start - item_start < size) {
+		ret = false;
+		goto out;
+	}
+
+out:
+	if (!ret)
+		btrfs_crit(fs_info, "invalid dir item name len: %u",
+			   (unsigned int)name_len);
+	return ret;
+}

fs/btrfs/disk-io.c

@@ -89,7 +89,6 @@ struct btrfs_end_io_wq {
 	struct btrfs_fs_info *info;
 	blk_status_t status;
 	enum btrfs_wq_endio_type metadata;
-	struct list_head list;
 	struct btrfs_work work;
 };
 
@@ -118,9 +117,9 @@ void btrfs_end_io_wq_exit(void)
  * just before they are sent down the IO stack.
  */
 struct async_submit_bio {
-	struct inode *inode;
+	void *private_data;
+	struct btrfs_fs_info *fs_info;
 	struct bio *bio;
-	struct list_head list;
 	extent_submit_bio_hook_t *submit_bio_start;
 	extent_submit_bio_hook_t *submit_bio_done;
 	int mirror_num;
@@ -871,7 +870,7 @@ static void run_one_async_start(struct btrfs_work *work)
 	blk_status_t ret;
 
 	async = container_of(work, struct  async_submit_bio, work);
-	ret = async->submit_bio_start(async->inode, async->bio,
+	ret = async->submit_bio_start(async->private_data, async->bio,
 				      async->mirror_num, async->bio_flags,
 				      async->bio_offset);
 	if (ret)
@@ -885,7 +884,7 @@ static void run_one_async_done(struct btrfs_work *work)
 	int limit;
 
 	async = container_of(work, struct  async_submit_bio, work);
-	fs_info = BTRFS_I(async->inode)->root->fs_info;
+	fs_info = async->fs_info;
 
 	limit = btrfs_async_submit_limit(fs_info);
 	limit = limit * 2 / 3;
@@ -904,7 +903,7 @@ static void run_one_async_done(struct btrfs_work *work)
 		return;
 	}
 
-	async->submit_bio_done(async->inode, async->bio, async->mirror_num,
+	async->submit_bio_done(async->private_data, async->bio, async->mirror_num,
 			       async->bio_flags, async->bio_offset);
 }
 
@@ -916,11 +915,11 @@ static void run_one_async_free(struct btrfs_work *work)
 	kfree(async);
 }
 
-blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info,
-		struct inode *inode, struct bio *bio, int mirror_num,
-		unsigned long bio_flags, u64 bio_offset,
-		extent_submit_bio_hook_t *submit_bio_start,
-		extent_submit_bio_hook_t *submit_bio_done)
+blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
+				 int mirror_num, unsigned long bio_flags,
+				 u64 bio_offset, void *private_data,
+				 extent_submit_bio_hook_t *submit_bio_start,
+				 extent_submit_bio_hook_t *submit_bio_done)
 {
 	struct async_submit_bio *async;
 
@@ -928,7 +927,8 @@ blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info,
 	if (!async)
 		return BLK_STS_RESOURCE;
 
-	async->inode = inode;
+	async->private_data = private_data;
+	async->fs_info = fs_info;
 	async->bio = bio;
 	async->mirror_num = mirror_num;
 	async->submit_bio_start = submit_bio_start;
@@ -974,9 +974,9 @@ static blk_status_t btree_csum_one_bio(struct bio *bio)
 	return errno_to_blk_status(ret);
 }
 
-static blk_status_t __btree_submit_bio_start(struct inode *inode,
-		struct bio *bio, int mirror_num, unsigned long bio_flags,
-		u64 bio_offset)
+static blk_status_t __btree_submit_bio_start(void *private_data, struct bio *bio,
+					     int mirror_num, unsigned long bio_flags,
+					     u64 bio_offset)
 {
 	/*
 	 * when we're called for a write, we're already in the async
@@ -985,10 +985,11 @@ static blk_status_t __btree_submit_bio_start(struct inode *inode,
 	return btree_csum_one_bio(bio);
 }
 
-static blk_status_t __btree_submit_bio_done(struct inode *inode,
-		struct bio *bio, int mirror_num, unsigned long bio_flags,
-		u64 bio_offset)
+static blk_status_t __btree_submit_bio_done(void *private_data, struct bio *bio,
+					    int mirror_num, unsigned long bio_flags,
+					    u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	blk_status_t ret;
 
 	/*
@@ -1014,10 +1015,11 @@ static int check_async_write(unsigned long bio_flags)
 	return 1;
 }
 
-static blk_status_t btree_submit_bio_hook(struct inode *inode, struct bio *bio,
-				 int mirror_num, unsigned long bio_flags,
-				 u64 bio_offset)
+static blk_status_t btree_submit_bio_hook(void *private_data, struct bio *bio,
+					  int mirror_num, unsigned long bio_flags,
+					  u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	int async = check_async_write(bio_flags);
 	blk_status_t ret;
@@ -1042,8 +1044,8 @@ static blk_status_t btree_submit_bio_hook(struct inode *inode, struct bio *bio,
 		 * kthread helpers are used to submit writes so that
 		 * checksumming can happen in parallel across all CPUs
 		 */
-		ret = btrfs_wq_submit_bio(fs_info, inode, bio, mirror_num, 0,
-					  bio_offset,
+		ret = btrfs_wq_submit_bio(fs_info, bio, mirror_num, 0,
+					  bio_offset, private_data,
 					  __btree_submit_bio_start,
 					  __btree_submit_bio_done);
 	}
@@ -1221,10 +1223,10 @@ int btrfs_write_tree_block(struct extent_buffer *buf)
 					buf->start + buf->len - 1);
 }
 
-int btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
+void btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
 {
-	return filemap_fdatawait_range(buf->pages[0]->mapping,
-				       buf->start, buf->start + buf->len - 1);
+	filemap_fdatawait_range(buf->pages[0]->mapping,
+			        buf->start, buf->start + buf->len - 1);
 }
 
 struct extent_buffer *read_tree_block(struct btrfs_fs_info *fs_info, u64 bytenr,
@@ -1346,8 +1348,7 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 	root->log_transid_committed = -1;
 	root->last_log_commit = 0;
 	if (!dummy)
-		extent_io_tree_init(&root->dirty_log_pages,
-				     fs_info->btree_inode->i_mapping);
+		extent_io_tree_init(&root->dirty_log_pages, NULL);
 
 	memset(&root->root_key, 0, sizeof(root->root_key));
 	memset(&root->root_item, 0, sizeof(root->root_item));
@@ -2308,7 +2309,7 @@ static void btrfs_init_btree_inode(struct btrfs_fs_info *fs_info)
 	inode->i_mapping->a_ops = &btree_aops;
 
 	RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
-	extent_io_tree_init(&BTRFS_I(inode)->io_tree, inode->i_mapping);
+	extent_io_tree_init(&BTRFS_I(inode)->io_tree, inode);
 	BTRFS_I(inode)->io_tree.track_uptodate = 0;
 	extent_map_tree_init(&BTRFS_I(inode)->extent_tree);
 
@@ -2625,7 +2626,6 @@ int open_ctree(struct super_block *sb,
 	spin_lock_init(&fs_info->fs_roots_radix_lock);
 	spin_lock_init(&fs_info->delayed_iput_lock);
 	spin_lock_init(&fs_info->defrag_inodes_lock);
-	spin_lock_init(&fs_info->free_chunk_lock);
 	spin_lock_init(&fs_info->tree_mod_seq_lock);
 	spin_lock_init(&fs_info->super_lock);
 	spin_lock_init(&fs_info->qgroup_op_lock);
@@ -2661,12 +2661,11 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->qgroup_op_seq, 0);
 	atomic_set(&fs_info->reada_works_cnt, 0);
 	atomic64_set(&fs_info->tree_mod_seq, 0);
-	fs_info->fs_frozen = 0;
 	fs_info->sb = sb;
 	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
 	fs_info->metadata_ratio = 0;
 	fs_info->defrag_inodes = RB_ROOT;
-	fs_info->free_chunk_space = 0;
+	atomic64_set(&fs_info->free_chunk_space, 0);
 	fs_info->tree_mod_log = RB_ROOT;
 	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 	fs_info->avg_delayed_ref_runtime = NSEC_PER_SEC >> 6; /* div by 64 */
@@ -2703,10 +2702,8 @@ int open_ctree(struct super_block *sb,
 	fs_info->block_group_cache_tree = RB_ROOT;
 	fs_info->first_logical_byte = (u64)-1;
 
-	extent_io_tree_init(&fs_info->freed_extents[0],
-			     fs_info->btree_inode->i_mapping);
-	extent_io_tree_init(&fs_info->freed_extents[1],
-			     fs_info->btree_inode->i_mapping);
+	extent_io_tree_init(&fs_info->freed_extents[0], NULL);
+	extent_io_tree_init(&fs_info->freed_extents[1], NULL);
 	fs_info->pinned_extents = &fs_info->freed_extents[0];
 	set_bit(BTRFS_FS_BARRIER, &fs_info->flags);
 
@@ -3484,65 +3481,61 @@ static int write_dev_supers(struct btrfs_device *device,
  */
 static void btrfs_end_empty_barrier(struct bio *bio)
 {
-	if (bio->bi_private)
-		complete(bio->bi_private);
-	bio_put(bio);
+	complete(bio->bi_private);
 }
 
 /*
- * trigger flushes for one the devices.  If you pass wait == 0, the flushes are
- * sent down.  With wait == 1, it waits for the previous flush.
- *
- * any device where the flush fails with eopnotsupp are flagged as not-barrier
- * capable
+ * Submit a flush request to the device if it supports it. Error handling is
+ * done in the waiting counterpart.
  */
-static blk_status_t write_dev_flush(struct btrfs_device *device, int wait)
+static void write_dev_flush(struct btrfs_device *device)
 {
 	struct request_queue *q = bdev_get_queue(device->bdev);
-	struct bio *bio;
-	blk_status_t ret = 0;
+	struct bio *bio = device->flush_bio;
 
 	if (!test_bit(QUEUE_FLAG_WC, &q->queue_flags))
-		return 0;
+		return;
 
-	if (wait) {
-		bio = device->flush_bio;
-		if (!bio)
-			return 0;
+	bio_reset(bio);
+	bio->bi_end_io = btrfs_end_empty_barrier;
+	bio->bi_bdev = device->bdev;
+	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_PREFLUSH;
+	init_completion(&device->flush_wait);
+	bio->bi_private = &device->flush_wait;
 
-		wait_for_completion(&device->flush_wait);
+	submit_bio(bio);
+	device->flush_bio_sent = 1;
+}
 
-		if (bio->bi_status) {
-			ret = bio->bi_status;
-			btrfs_dev_stat_inc_and_print(device,
-				BTRFS_DEV_STAT_FLUSH_ERRS);
-		}
+/*
+ * If the flush bio has been submitted by write_dev_flush, wait for it.
+ */
+static blk_status_t wait_dev_flush(struct btrfs_device *device)
+{
+	struct bio *bio = device->flush_bio;
 
-		/* drop the reference from the wait == 0 run */
-		bio_put(bio);
-		device->flush_bio = NULL;
+	if (!device->flush_bio_sent)
+		return 0;
 
-		return ret;
-	}
+	device->flush_bio_sent = 0;
+	wait_for_completion_io(&device->flush_wait);
 
-	/*
-	 * one reference for us, and we leave it for the
-	 * caller
-	 */
-	device->flush_bio = NULL;
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 0);
-	if (!bio)
-		return BLK_STS_RESOURCE;
+	return bio->bi_status;
+}
 
-	bio->bi_end_io = btrfs_end_empty_barrier;
-	bio->bi_bdev = device->bdev;
-	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_PREFLUSH;
-	init_completion(&device->flush_wait);
-	bio->bi_private = &device->flush_wait;
-	device->flush_bio = bio;
+static int check_barrier_error(struct btrfs_fs_devices *fsdevs)
+{
+	int dev_flush_error = 0;
+	struct btrfs_device *dev;
 
-	bio_get(bio);
-	btrfsic_submit_bio(bio);
+	list_for_each_entry_rcu(dev, &fsdevs->devices, dev_list) {
+		if (!dev->bdev || dev->last_flush_error)
+			dev_flush_error++;
+	}
+
+	if (dev_flush_error >
+	    fsdevs->fs_info->num_tolerated_disk_barrier_failures)
+		return -EIO;
 
 	return 0;
 }
@@ -3555,7 +3548,6 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 {
 	struct list_head *head;
 	struct btrfs_device *dev;
-	int errors_send = 0;
 	int errors_wait = 0;
 	blk_status_t ret;
 
@@ -3564,16 +3556,13 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 	list_for_each_entry_rcu(dev, head, dev_list) {
 		if (dev->missing)
 			continue;
-		if (!dev->bdev) {
-			errors_send++;
+		if (!dev->bdev)
 			continue;
-		}
 		if (!dev->in_fs_metadata || !dev->writeable)
 			continue;
 
-		ret = write_dev_flush(dev, 0);
-		if (ret)
-			errors_send++;
+		write_dev_flush(dev);
+		dev->last_flush_error = 0;
 	}
 
 	/* wait for all the barriers */
@@ -3587,13 +3576,23 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 		if (!dev->in_fs_metadata || !dev->writeable)
 			continue;
 
-		ret = write_dev_flush(dev, 1);
-		if (ret)
+		ret = wait_dev_flush(dev);
+		if (ret) {
+			dev->last_flush_error = ret;
+			btrfs_dev_stat_inc_and_print(dev,
+					BTRFS_DEV_STAT_FLUSH_ERRS);
 			errors_wait++;
+		}
+	}
+
+	if (errors_wait) {
+		/*
+		 * At some point we need the status of all disks
+		 * to arrive at the volume status. So error checking
+		 * is being pushed to a separate loop.
+		 */
+		return check_barrier_error(info->fs_devices);
 	}
-	if (errors_send > info->num_tolerated_disk_barrier_failures ||
-	    errors_wait > info->num_tolerated_disk_barrier_failures)
-		return -EIO;
 	return 0;
 }
 
@@ -4577,11 +4576,6 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 
 	cur_trans->state =TRANS_STATE_COMPLETED;
 	wake_up(&cur_trans->commit_wait);
-
-	/*
-	memset(cur_trans, 0, sizeof(*cur_trans));
-	kmem_cache_free(btrfs_transaction_cachep, cur_trans);
-	*/
 }
 
 static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info)
@@ -4637,6 +4631,12 @@ static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
+static struct btrfs_fs_info *btree_fs_info(void *private_data)
+{
+	struct inode *inode = private_data;
+	return btrfs_sb(inode->i_sb);
+}
+
 static const struct extent_io_ops btree_extent_io_ops = {
 	/* mandatory callbacks */
 	.submit_bio_hook = btree_submit_bio_hook,
@@ -4644,6 +4644,8 @@ static const struct extent_io_ops btree_extent_io_ops = {
 	/* note we're sharing with inode.c for the merge bio hook */
 	.merge_bio_hook = btrfs_merge_bio_hook,
 	.readpage_io_failed_hook = btree_io_failed_hook,
+	.set_range_writeback = btrfs_set_range_writeback,
+	.tree_fs_info = btree_fs_info,
 
 	/* optional callbacks */
 };

fs/btrfs/disk-io.h

@@ -120,14 +120,14 @@ u32 btrfs_csum_data(const char *data, u32 seed, size_t len);
 void btrfs_csum_final(u32 crc, u8 *result);
 blk_status_t btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
 			enum btrfs_wq_endio_type metadata);
-blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info,
-		struct inode *inode, struct bio *bio, int mirror_num,
-		unsigned long bio_flags, u64 bio_offset,
-		extent_submit_bio_hook_t *submit_bio_start,
-		extent_submit_bio_hook_t *submit_bio_done);
+blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
+			int mirror_num, unsigned long bio_flags,
+			u64 bio_offset, void *private_data,
+			extent_submit_bio_hook_t *submit_bio_start,
+			extent_submit_bio_hook_t *submit_bio_done);
 unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info);
 int btrfs_write_tree_block(struct extent_buffer *buf);
-int btrfs_wait_tree_block_writeback(struct extent_buffer *buf);
+void btrfs_wait_tree_block_writeback(struct extent_buffer *buf);
 int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
 int btrfs_add_log_tree(struct btrfs_trans_handle *trans,

fs/btrfs/export.c

@@ -282,6 +282,11 @@ static int btrfs_get_name(struct dentry *parent, char *name,
 		name_len = btrfs_inode_ref_name_len(leaf, iref);
 	}
 
+	ret = btrfs_is_name_len_valid(leaf, path->slots[0], name_ptr, name_len);
+	if (!ret) {
+		btrfs_free_path(path);
+		return -EIO;
+	}
 	read_extent_buffer(leaf, name, name_ptr, name_len);
 	btrfs_free_path(path);
 

fs/btrfs/extent-tree.c

@@ -97,10 +97,11 @@ static int btrfs_free_reserved_bytes(struct btrfs_block_group_cache *cache,
 				     u64 num_bytes, int delalloc);
 static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv,
 			       u64 num_bytes);
-static int __reserve_metadata_bytes(struct btrfs_root *root,
+static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info,
 				    struct btrfs_space_info *space_info,
 				    u64 orig_bytes,
-				    enum btrfs_reserve_flush_enum flush);
+				    enum btrfs_reserve_flush_enum flush,
+				    bool system_chunk);
 static void space_info_add_new_bytes(struct btrfs_fs_info *fs_info,
 				     struct btrfs_space_info *space_info,
 				     u64 num_bytes);
@@ -766,6 +767,26 @@ static struct btrfs_space_info *__find_space_info(struct btrfs_fs_info *info,
 	return NULL;
 }
 
+static void add_pinned_bytes(struct btrfs_fs_info *fs_info, s64 num_bytes,
+			     u64 owner, u64 root_objectid)
+{
+	struct btrfs_space_info *space_info;
+	u64 flags;
+
+	if (owner < BTRFS_FIRST_FREE_OBJECTID) {
+		if (root_objectid == BTRFS_CHUNK_TREE_OBJECTID)
+			flags = BTRFS_BLOCK_GROUP_SYSTEM;
+		else
+			flags = BTRFS_BLOCK_GROUP_METADATA;
+	} else {
+		flags = BTRFS_BLOCK_GROUP_DATA;
+	}
+
+	space_info = __find_space_info(fs_info, flags);
+	ASSERT(space_info);
+	percpu_counter_add(&space_info->total_bytes_pinned, num_bytes);
+}
+
 /*
  * after adding space to the filesystem, we need to clear the full flags
  * on all the space infos.
@@ -2092,6 +2113,7 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 			 u64 bytenr, u64 num_bytes, u64 parent,
 			 u64 root_objectid, u64 owner, u64 offset)
 {
+	int old_ref_mod, new_ref_mod;
 	int ret;
 
 	BUG_ON(owner < BTRFS_FIRST_FREE_OBJECTID &&
@@ -2099,15 +2121,21 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
 
 	if (owner < BTRFS_FIRST_FREE_OBJECTID) {
 		ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr,
-					num_bytes,
-					parent, root_objectid, (int)owner,
-					BTRFS_ADD_DELAYED_REF, NULL);
+						 num_bytes, parent,
+						 root_objectid, (int)owner,
+						 BTRFS_ADD_DELAYED_REF, NULL,
+						 &old_ref_mod, &new_ref_mod);
 	} else {
 		ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr,
-					num_bytes, parent, root_objectid,
-					owner, offset, 0,
-					BTRFS_ADD_DELAYED_REF);
+						 num_bytes, parent,
+						 root_objectid, owner, offset,
+						 0, BTRFS_ADD_DELAYED_REF,
+						 &old_ref_mod, &new_ref_mod);
 	}
+
+	if (ret == 0 && old_ref_mod < 0 && new_ref_mod >= 0)
+		add_pinned_bytes(fs_info, -num_bytes, owner, root_objectid);
+
 	return ret;
 }
 
@@ -2411,6 +2439,16 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans,
 		head = btrfs_delayed_node_to_head(node);
 		trace_run_delayed_ref_head(fs_info, node, head, node->action);
 
+		if (head->total_ref_mod < 0) {
+			struct btrfs_block_group_cache *cache;
+
+			cache = btrfs_lookup_block_group(fs_info, node->bytenr);
+			ASSERT(cache);
+			percpu_counter_add(&cache->space_info->total_bytes_pinned,
+					   -node->num_bytes);
+			btrfs_put_block_group(cache);
+		}
+
 		if (insert_reserved) {
 			btrfs_pin_extent(fs_info, node->bytenr,
 					 node->num_bytes, 1);
@@ -3364,6 +3402,7 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 	struct btrfs_fs_info *fs_info = block_group->fs_info;
 	struct btrfs_root *root = fs_info->tree_root;
 	struct inode *inode = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	u64 alloc_hint = 0;
 	int dcs = BTRFS_DC_ERROR;
 	u64 num_pages = 0;
@@ -3483,7 +3522,7 @@ again:
 	num_pages *= 16;
 	num_pages *= PAGE_SIZE;
 
-	ret = btrfs_check_data_free_space(inode, 0, num_pages);
+	ret = btrfs_check_data_free_space(inode, &data_reserved, 0, num_pages);
 	if (ret)
 		goto out_put;
 
@@ -3514,6 +3553,7 @@ out:
 	block_group->disk_cache_state = dcs;
 	spin_unlock(&block_group->lock);
 
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 
@@ -3924,88 +3964,83 @@ static const char *alloc_name(u64 flags)
 	};
 }
 
-static int update_space_info(struct btrfs_fs_info *info, u64 flags,
-			     u64 total_bytes, u64 bytes_used,
-			     u64 bytes_readonly,
-			     struct btrfs_space_info **space_info)
+static int create_space_info(struct btrfs_fs_info *info, u64 flags,
+			     struct btrfs_space_info **new)
 {
-	struct btrfs_space_info *found;
+
+	struct btrfs_space_info *space_info;
 	int i;
-	int factor;
 	int ret;
 
-	if (flags & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 |
-		     BTRFS_BLOCK_GROUP_RAID10))
-		factor = 2;
-	else
-		factor = 1;
-
-	found = __find_space_info(info, flags);
-	if (found) {
-		spin_lock(&found->lock);
-		found->total_bytes += total_bytes;
-		found->disk_total += total_bytes * factor;
-		found->bytes_used += bytes_used;
-		found->disk_used += bytes_used * factor;
-		found->bytes_readonly += bytes_readonly;
-		if (total_bytes > 0)
-			found->full = 0;
-		space_info_add_new_bytes(info, found, total_bytes -
-					 bytes_used - bytes_readonly);
-		spin_unlock(&found->lock);
-		*space_info = found;
-		return 0;
-	}
-	found = kzalloc(sizeof(*found), GFP_NOFS);
-	if (!found)
+	space_info = kzalloc(sizeof(*space_info), GFP_NOFS);
+	if (!space_info)
 		return -ENOMEM;
 
-	ret = percpu_counter_init(&found->total_bytes_pinned, 0, GFP_KERNEL);
+	ret = percpu_counter_init(&space_info->total_bytes_pinned, 0,
+				 GFP_KERNEL);
 	if (ret) {
-		kfree(found);
+		kfree(space_info);
 		return ret;
 	}
 
 	for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
-		INIT_LIST_HEAD(&found->block_groups[i]);
-	init_rwsem(&found->groups_sem);
-	spin_lock_init(&found->lock);
-	found->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
-	found->total_bytes = total_bytes;
-	found->disk_total = total_bytes * factor;
-	found->bytes_used = bytes_used;
-	found->disk_used = bytes_used * factor;
-	found->bytes_pinned = 0;
-	found->bytes_reserved = 0;
-	found->bytes_readonly = bytes_readonly;
-	found->bytes_may_use = 0;
-	found->full = 0;
-	found->max_extent_size = 0;
-	found->force_alloc = CHUNK_ALLOC_NO_FORCE;
-	found->chunk_alloc = 0;
-	found->flush = 0;
-	init_waitqueue_head(&found->wait);
-	INIT_LIST_HEAD(&found->ro_bgs);
-	INIT_LIST_HEAD(&found->tickets);
-	INIT_LIST_HEAD(&found->priority_tickets);
-
-	ret = kobject_init_and_add(&found->kobj, &space_info_ktype,
+		INIT_LIST_HEAD(&space_info->block_groups[i]);
+	init_rwsem(&space_info->groups_sem);
+	spin_lock_init(&space_info->lock);
+	space_info->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
+	space_info->force_alloc = CHUNK_ALLOC_NO_FORCE;
+	init_waitqueue_head(&space_info->wait);
+	INIT_LIST_HEAD(&space_info->ro_bgs);
+	INIT_LIST_HEAD(&space_info->tickets);
+	INIT_LIST_HEAD(&space_info->priority_tickets);
+
+	ret = kobject_init_and_add(&space_info->kobj, &space_info_ktype,
 				    info->space_info_kobj, "%s",
-				    alloc_name(found->flags));
+				    alloc_name(space_info->flags));
 	if (ret) {
-		percpu_counter_destroy(&found->total_bytes_pinned);
-		kfree(found);
+		percpu_counter_destroy(&space_info->total_bytes_pinned);
+		kfree(space_info);
 		return ret;
 	}
 
-	*space_info = found;
-	list_add_rcu(&found->list, &info->space_info);
+	*new = space_info;
+	list_add_rcu(&space_info->list, &info->space_info);
 	if (flags & BTRFS_BLOCK_GROUP_DATA)
-		info->data_sinfo = found;
+		info->data_sinfo = space_info;
 
 	return ret;
 }
 
+static void update_space_info(struct btrfs_fs_info *info, u64 flags,
+			     u64 total_bytes, u64 bytes_used,
+			     u64 bytes_readonly,
+			     struct btrfs_space_info **space_info)
+{
+	struct btrfs_space_info *found;
+	int factor;
+
+	if (flags & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 |
+		     BTRFS_BLOCK_GROUP_RAID10))
+		factor = 2;
+	else
+		factor = 1;
+
+	found = __find_space_info(info, flags);
+	ASSERT(found);
+	spin_lock(&found->lock);
+	found->total_bytes += total_bytes;
+	found->disk_total += total_bytes * factor;
+	found->bytes_used += bytes_used;
+	found->disk_used += bytes_used * factor;
+	found->bytes_readonly += bytes_readonly;
+	if (total_bytes > 0)
+		found->full = 0;
+	space_info_add_new_bytes(info, found, total_bytes -
+				 bytes_used - bytes_readonly);
+	spin_unlock(&found->lock);
+	*space_info = found;
+}
+
 static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
 {
 	u64 extra_flags = chunk_to_extended(flags) &
@@ -4121,7 +4156,7 @@ static u64 get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags)
 	return btrfs_reduce_alloc_profile(fs_info, flags);
 }
 
-u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data)
+static u64 get_alloc_profile_by_root(struct btrfs_root *root, int data)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	u64 flags;
@@ -4138,6 +4173,21 @@ u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data)
 	return ret;
 }
 
+u64 btrfs_data_alloc_profile(struct btrfs_fs_info *fs_info)
+{
+	return get_alloc_profile(fs_info, BTRFS_BLOCK_GROUP_DATA);
+}
+
+u64 btrfs_metadata_alloc_profile(struct btrfs_fs_info *fs_info)
+{
+	return get_alloc_profile(fs_info, BTRFS_BLOCK_GROUP_METADATA);
+}
+
+u64 btrfs_system_alloc_profile(struct btrfs_fs_info *fs_info)
+{
+	return get_alloc_profile(fs_info, BTRFS_BLOCK_GROUP_SYSTEM);
+}
+
 static u64 btrfs_space_info_used(struct btrfs_space_info *s_info,
 				 bool may_use_included)
 {
@@ -4187,7 +4237,7 @@ again:
 			data_sinfo->force_alloc = CHUNK_ALLOC_FORCE;
 			spin_unlock(&data_sinfo->lock);
 alloc:
-			alloc_target = btrfs_get_alloc_profile(root, 1);
+			alloc_target = btrfs_data_alloc_profile(fs_info);
 			/*
 			 * It is ugly that we don't call nolock join
 			 * transaction for the free space inode case here.
@@ -4238,7 +4288,7 @@ commit_trans:
 
 			if (need_commit > 0) {
 				btrfs_start_delalloc_roots(fs_info, 0, -1);
-				btrfs_wait_ordered_roots(fs_info, -1, 0,
+				btrfs_wait_ordered_roots(fs_info, U64_MAX, 0,
 							 (u64)-1);
 			}
 
@@ -4278,12 +4328,8 @@ commit_trans:
 	return ret;
 }
 
-/*
- * New check_data_free_space() with ability for precious data reservation
- * Will replace old btrfs_check_data_free_space(), but for patch split,
- * add a new function first and then replace it.
- */
-int btrfs_check_data_free_space(struct inode *inode, u64 start, u64 len)
+int btrfs_check_data_free_space(struct inode *inode,
+			struct extent_changeset **reserved, u64 start, u64 len)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	int ret;
@@ -4298,9 +4344,11 @@ int btrfs_check_data_free_space(struct inode *inode, u64 start, u64 len)
 		return ret;
 
 	/* Use new btrfs_qgroup_reserve_data to reserve precious data space. */
-	ret = btrfs_qgroup_reserve_data(inode, start, len);
-	if (ret)
+	ret = btrfs_qgroup_reserve_data(inode, reserved, start, len);
+	if (ret < 0)
 		btrfs_free_reserved_data_space_noquota(inode, start, len);
+	else
+		ret = 0;
 	return ret;
 }
 
@@ -4341,7 +4389,8 @@ void btrfs_free_reserved_data_space_noquota(struct inode *inode, u64 start,
  * This one will handle the per-inode data rsv map for accurate reserved
  * space framework.
  */
-void btrfs_free_reserved_data_space(struct inode *inode, u64 start, u64 len)
+void btrfs_free_reserved_data_space(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 
@@ -4351,7 +4400,7 @@ void btrfs_free_reserved_data_space(struct inode *inode, u64 start, u64 len)
 	start = round_down(start, root->fs_info->sectorsize);
 
 	btrfs_free_reserved_data_space_noquota(inode, start, len);
-	btrfs_qgroup_free_data(inode, start, len);
+	btrfs_qgroup_free_data(inode, reserved, start, len);
 }
 
 static void force_metadata_allocation(struct btrfs_fs_info *info)
@@ -4463,9 +4512,8 @@ void check_system_chunk(struct btrfs_trans_handle *trans,
 	}
 
 	if (left < thresh) {
-		u64 flags;
+		u64 flags = btrfs_system_alloc_profile(fs_info);
 
-		flags = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
 		/*
 		 * Ignore failure to create system chunk. We might end up not
 		 * needing it, as we might not need to COW all nodes/leafs from
@@ -4506,10 +4554,10 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 
 	space_info = __find_space_info(fs_info, flags);
 	if (!space_info) {
-		ret = update_space_info(fs_info, flags, 0, 0, 0, &space_info);
-		BUG_ON(ret); /* -ENOMEM */
+		ret = create_space_info(fs_info, flags, &space_info);
+		if (ret)
+			return ret;
 	}
-	BUG_ON(!space_info); /* Logic error */
 
 again:
 	spin_lock(&space_info->lock);
@@ -4614,11 +4662,11 @@ out:
 	return ret;
 }
 
-static int can_overcommit(struct btrfs_root *root,
+static int can_overcommit(struct btrfs_fs_info *fs_info,
 			  struct btrfs_space_info *space_info, u64 bytes,
-			  enum btrfs_reserve_flush_enum flush)
+			  enum btrfs_reserve_flush_enum flush,
+			  bool system_chunk)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
 	u64 profile;
 	u64 space_size;
@@ -4629,7 +4677,11 @@ static int can_overcommit(struct btrfs_root *root,
 	if (space_info->flags & BTRFS_BLOCK_GROUP_DATA)
 		return 0;
 
-	profile = btrfs_get_alloc_profile(root, 0);
+	if (system_chunk)
+		profile = btrfs_system_alloc_profile(fs_info);
+	else
+		profile = btrfs_metadata_alloc_profile(fs_info);
+
 	used = btrfs_space_info_used(space_info, false);
 
 	/*
@@ -4646,9 +4698,7 @@ static int can_overcommit(struct btrfs_root *root,
 
 	used += space_info->bytes_may_use;
 
-	spin_lock(&fs_info->free_chunk_lock);
-	avail = fs_info->free_chunk_space;
-	spin_unlock(&fs_info->free_chunk_lock);
+	avail = atomic64_read(&fs_info->free_chunk_space);
 
 	/*
 	 * If we have dup, raid1 or raid10 then only half of the free
@@ -4698,14 +4748,14 @@ static void btrfs_writeback_inodes_sb_nr(struct btrfs_fs_info *fs_info,
 	}
 }
 
-static inline int calc_reclaim_items_nr(struct btrfs_fs_info *fs_info,
+static inline u64 calc_reclaim_items_nr(struct btrfs_fs_info *fs_info,
 					u64 to_reclaim)
 {
 	u64 bytes;
-	int nr;
+	u64 nr;
 
 	bytes = btrfs_calc_trans_metadata_size(fs_info, 1);
-	nr = (int)div64_u64(to_reclaim, bytes);
+	nr = div64_u64(to_reclaim, bytes);
 	if (!nr)
 		nr = 1;
 	return nr;
@@ -4716,24 +4766,23 @@ static inline int calc_reclaim_items_nr(struct btrfs_fs_info *fs_info,
 /*
  * shrink metadata reservation for delalloc
  */
-static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig,
-			    bool wait_ordered)
+static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim,
+			    u64 orig, bool wait_ordered)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_block_rsv *block_rsv;
 	struct btrfs_space_info *space_info;
 	struct btrfs_trans_handle *trans;
 	u64 delalloc_bytes;
 	u64 max_reclaim;
+	u64 items;
 	long time_left;
 	unsigned long nr_pages;
 	int loops;
-	int items;
 	enum btrfs_reserve_flush_enum flush;
 
 	/* Calc the number of the pages we need flush for space reservation */
 	items = calc_reclaim_items_nr(fs_info, to_reclaim);
-	to_reclaim = (u64)items * EXTENT_SIZE_PER_ITEM;
+	to_reclaim = items * EXTENT_SIZE_PER_ITEM;
 
 	trans = (struct btrfs_trans_handle *)current->journal_info;
 	block_rsv = &fs_info->delalloc_block_rsv;
@@ -4776,7 +4825,7 @@ skip_async:
 		else
 			flush = BTRFS_RESERVE_NO_FLUSH;
 		spin_lock(&space_info->lock);
-		if (can_overcommit(root, space_info, orig, flush)) {
+		if (can_overcommit(fs_info, space_info, orig, flush, false)) {
 			spin_unlock(&space_info->lock);
 			break;
 		}
@@ -4838,7 +4887,7 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
 
 	spin_lock(&delayed_rsv->lock);
 	if (percpu_counter_compare(&space_info->total_bytes_pinned,
-				   bytes - delayed_rsv->size) >= 0) {
+				   bytes - delayed_rsv->size) < 0) {
 		spin_unlock(&delayed_rsv->lock);
 		return -ENOSPC;
 	}
@@ -4886,7 +4935,7 @@ static int flush_space(struct btrfs_fs_info *fs_info,
 		break;
 	case FLUSH_DELALLOC:
 	case FLUSH_DELALLOC_WAIT:
-		shrink_delalloc(root, num_bytes * 2, orig_bytes,
+		shrink_delalloc(fs_info, num_bytes * 2, orig_bytes,
 				state == FLUSH_DELALLOC_WAIT);
 		break;
 	case ALLOC_CHUNK:
@@ -4896,7 +4945,7 @@ static int flush_space(struct btrfs_fs_info *fs_info,
 			break;
 		}
 		ret = do_chunk_alloc(trans, fs_info,
-				     btrfs_get_alloc_profile(root, 0),
+				     btrfs_metadata_alloc_profile(fs_info),
 				     CHUNK_ALLOC_NO_FORCE);
 		btrfs_end_transaction(trans);
 		if (ret > 0 || ret == -ENOSPC)
@@ -4917,8 +4966,9 @@ static int flush_space(struct btrfs_fs_info *fs_info,
 }
 
 static inline u64
-btrfs_calc_reclaim_metadata_size(struct btrfs_root *root,
-				 struct btrfs_space_info *space_info)
+btrfs_calc_reclaim_metadata_size(struct btrfs_fs_info *fs_info,
+				 struct btrfs_space_info *space_info,
+				 bool system_chunk)
 {
 	struct reserve_ticket *ticket;
 	u64 used;
@@ -4933,14 +4983,14 @@ btrfs_calc_reclaim_metadata_size(struct btrfs_root *root,
 		return to_reclaim;
 
 	to_reclaim = min_t(u64, num_online_cpus() * SZ_1M, SZ_16M);
-	if (can_overcommit(root, space_info, to_reclaim,
-			   BTRFS_RESERVE_FLUSH_ALL))
+	if (can_overcommit(fs_info, space_info, to_reclaim,
+			   BTRFS_RESERVE_FLUSH_ALL, system_chunk))
 		return 0;
 
-	used = space_info->bytes_used + space_info->bytes_reserved +
-	       space_info->bytes_pinned + space_info->bytes_readonly +
-	       space_info->bytes_may_use;
-	if (can_overcommit(root, space_info, SZ_1M, BTRFS_RESERVE_FLUSH_ALL))
+	used = btrfs_space_info_used(space_info, true);
+
+	if (can_overcommit(fs_info, space_info, SZ_1M,
+			   BTRFS_RESERVE_FLUSH_ALL, system_chunk))
 		expected = div_factor_fine(space_info->total_bytes, 95);
 	else
 		expected = div_factor_fine(space_info->total_bytes, 90);
@@ -4954,17 +5004,18 @@ btrfs_calc_reclaim_metadata_size(struct btrfs_root *root,
 	return to_reclaim;
 }
 
-static inline int need_do_async_reclaim(struct btrfs_space_info *space_info,
-					struct btrfs_root *root, u64 used)
+static inline int need_do_async_reclaim(struct btrfs_fs_info *fs_info,
+					struct btrfs_space_info *space_info,
+					u64 used, bool system_chunk)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	u64 thresh = div_factor_fine(space_info->total_bytes, 98);
 
 	/* If we're just plain full then async reclaim just slows us down. */
 	if ((space_info->bytes_used + space_info->bytes_reserved) >= thresh)
 		return 0;
 
-	if (!btrfs_calc_reclaim_metadata_size(root, space_info))
+	if (!btrfs_calc_reclaim_metadata_size(fs_info, space_info,
+					      system_chunk))
 		return 0;
 
 	return (used >= thresh && !btrfs_fs_closing(fs_info) &&
@@ -5001,8 +5052,8 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work)
 	space_info = __find_space_info(fs_info, BTRFS_BLOCK_GROUP_METADATA);
 
 	spin_lock(&space_info->lock);
-	to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info->fs_root,
-						      space_info);
+	to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info,
+						      false);
 	if (!to_reclaim) {
 		space_info->flush = 0;
 		spin_unlock(&space_info->lock);
@@ -5024,8 +5075,9 @@ static void btrfs_async_reclaim_metadata_space(struct work_struct *work)
 			spin_unlock(&space_info->lock);
 			return;
 		}
-		to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info->fs_root,
-							      space_info);
+		to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info,
+							      space_info,
+							      false);
 		ticket = list_first_entry(&space_info->tickets,
 					  struct reserve_ticket, list);
 		if (last_tickets_id == space_info->tickets_id) {
@@ -5063,8 +5115,8 @@ static void priority_reclaim_metadata_space(struct btrfs_fs_info *fs_info,
 	int flush_state = FLUSH_DELAYED_ITEMS_NR;
 
 	spin_lock(&space_info->lock);
-	to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info->extent_root,
-						      space_info);
+	to_reclaim = btrfs_calc_reclaim_metadata_size(fs_info, space_info,
+						      false);
 	if (!to_reclaim) {
 		spin_unlock(&space_info->lock);
 		return;
@@ -5143,12 +5195,12 @@ static int wait_reserve_ticket(struct btrfs_fs_info *fs_info,
  * regain reservations will be made and this will fail if there is not enough
  * space already.
  */
-static int __reserve_metadata_bytes(struct btrfs_root *root,
+static int __reserve_metadata_bytes(struct btrfs_fs_info *fs_info,
 				    struct btrfs_space_info *space_info,
 				    u64 orig_bytes,
-				    enum btrfs_reserve_flush_enum flush)
+				    enum btrfs_reserve_flush_enum flush,
+				    bool system_chunk)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct reserve_ticket ticket;
 	u64 used;
 	int ret = 0;
@@ -5170,7 +5222,8 @@ static int __reserve_metadata_bytes(struct btrfs_root *root,
 		trace_btrfs_space_reservation(fs_info, "space_info",
 					      space_info->flags, orig_bytes, 1);
 		ret = 0;
-	} else if (can_overcommit(root, space_info, orig_bytes, flush)) {
+	} else if (can_overcommit(fs_info, space_info, orig_bytes, flush,
+				  system_chunk)) {
 		space_info->bytes_may_use += orig_bytes;
 		trace_btrfs_space_reservation(fs_info, "space_info",
 					      space_info->flags, orig_bytes, 1);
@@ -5197,7 +5250,7 @@ static int __reserve_metadata_bytes(struct btrfs_root *root,
 							  orig_bytes, flush,
 							  "enospc");
 				queue_work(system_unbound_wq,
-					   &root->fs_info->async_reclaim_work);
+					   &fs_info->async_reclaim_work);
 			}
 		} else {
 			list_add_tail(&ticket.list,
@@ -5211,7 +5264,8 @@ static int __reserve_metadata_bytes(struct btrfs_root *root,
 		 * the async reclaim as we will panic.
 		 */
 		if (!test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags) &&
-		    need_do_async_reclaim(space_info, root, used) &&
+		    need_do_async_reclaim(fs_info, space_info,
+					  used, system_chunk) &&
 		    !work_busy(&fs_info->async_reclaim_work)) {
 			trace_btrfs_trigger_flush(fs_info, space_info->flags,
 						  orig_bytes, flush, "preempt");
@@ -5269,9 +5323,10 @@ static int reserve_metadata_bytes(struct btrfs_root *root,
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_block_rsv *global_rsv = &fs_info->global_block_rsv;
 	int ret;
+	bool system_chunk = (root == fs_info->chunk_root);
 
-	ret = __reserve_metadata_bytes(root, block_rsv->space_info, orig_bytes,
-				       flush);
+	ret = __reserve_metadata_bytes(fs_info, block_rsv->space_info,
+				       orig_bytes, flush, system_chunk);
 	if (ret == -ENOSPC &&
 	    unlikely(root->orphan_cleanup_state == ORPHAN_CLEANUP_STARTED)) {
 		if (block_rsv != global_rsv &&
@@ -5380,9 +5435,7 @@ static void space_info_add_old_bytes(struct btrfs_fs_info *fs_info,
 	 * overcommit, and if we can't then we just need to free up our space
 	 * and not satisfy any requests.
 	 */
-	used = space_info->bytes_used + space_info->bytes_reserved +
-		space_info->bytes_pinned + space_info->bytes_readonly +
-		space_info->bytes_may_use;
+	used = btrfs_space_info_used(space_info, true);
 	if (used - num_bytes >= space_info->total_bytes)
 		check_overcommit = true;
 again:
@@ -5394,8 +5447,7 @@ again:
 		 * adding the ticket space would be a double count.
 		 */
 		if (check_overcommit &&
-		    !can_overcommit(fs_info->extent_root, space_info, 0,
-				    flush))
+		    !can_overcommit(fs_info, space_info, 0, flush, false))
 			break;
 		if (num_bytes >= ticket->bytes) {
 			list_del_init(&ticket->list);
@@ -6124,6 +6176,8 @@ void btrfs_delalloc_release_metadata(struct btrfs_inode *inode, u64 num_bytes)
  * @inode: inode we're writing to
  * @start: start range we are writing to
  * @len: how long the range we are writing to
+ * @reserved: mandatory parameter, record actually reserved qgroup ranges of
+ * 	      current reservation.
  *
  * This will do the following things
  *
@@ -6141,16 +6195,17 @@ void btrfs_delalloc_release_metadata(struct btrfs_inode *inode, u64 num_bytes)
  * Return 0 for success
  * Return <0 for error(-ENOSPC or -EQUOT)
  */
-int btrfs_delalloc_reserve_space(struct inode *inode, u64 start, u64 len)
+int btrfs_delalloc_reserve_space(struct inode *inode,
+			struct extent_changeset **reserved, u64 start, u64 len)
 {
 	int ret;
 
-	ret = btrfs_check_data_free_space(inode, start, len);
+	ret = btrfs_check_data_free_space(inode, reserved, start, len);
 	if (ret < 0)
 		return ret;
 	ret = btrfs_delalloc_reserve_metadata(BTRFS_I(inode), len);
 	if (ret < 0)
-		btrfs_free_reserved_data_space(inode, start, len);
+		btrfs_free_reserved_data_space(inode, *reserved, start, len);
 	return ret;
 }
 
@@ -6169,10 +6224,11 @@ int btrfs_delalloc_reserve_space(struct inode *inode, u64 start, u64 len)
  * list if there are no delalloc bytes left.
  * Also it will handle the qgroup reserved space.
  */
-void btrfs_delalloc_release_space(struct inode *inode, u64 start, u64 len)
+void btrfs_delalloc_release_space(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len)
 {
 	btrfs_delalloc_release_metadata(BTRFS_I(inode), len);
-	btrfs_free_reserved_data_space(inode, start, len);
+	btrfs_free_reserved_data_space(inode, reserved, start, len);
 }
 
 static int update_block_group(struct btrfs_trans_handle *trans,
@@ -6248,6 +6304,8 @@ static int update_block_group(struct btrfs_trans_handle *trans,
 			trace_btrfs_space_reservation(info, "pinned",
 						      cache->space_info->flags,
 						      num_bytes, 1);
+			percpu_counter_add(&cache->space_info->total_bytes_pinned,
+					   num_bytes);
 			set_extent_dirty(info->pinned_extents,
 					 bytenr, bytenr + num_bytes - 1,
 					 GFP_NOFS | __GFP_NOFAIL);
@@ -6324,6 +6382,7 @@ static int pin_down_extent(struct btrfs_fs_info *fs_info,
 
 	trace_btrfs_space_reservation(fs_info, "pinned",
 				      cache->space_info->flags, num_bytes, 1);
+	percpu_counter_add(&cache->space_info->total_bytes_pinned, num_bytes);
 	set_extent_dirty(fs_info->pinned_extents, bytenr,
 			 bytenr + num_bytes - 1, GFP_NOFS | __GFP_NOFAIL);
 	return 0;
@@ -6794,27 +6853,6 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-static void add_pinned_bytes(struct btrfs_fs_info *fs_info, u64 num_bytes,
-			     u64 owner, u64 root_objectid)
-{
-	struct btrfs_space_info *space_info;
-	u64 flags;
-
-	if (owner < BTRFS_FIRST_FREE_OBJECTID) {
-		if (root_objectid == BTRFS_CHUNK_TREE_OBJECTID)
-			flags = BTRFS_BLOCK_GROUP_SYSTEM;
-		else
-			flags = BTRFS_BLOCK_GROUP_METADATA;
-	} else {
-		flags = BTRFS_BLOCK_GROUP_DATA;
-	}
-
-	space_info = __find_space_info(fs_info, flags);
-	BUG_ON(!space_info); /* Logic bug */
-	percpu_counter_add(&space_info->total_bytes_pinned, num_bytes);
-}
-
-
 static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 				struct btrfs_fs_info *info,
 				struct btrfs_delayed_ref_node *node, u64 parent,
@@ -7037,8 +7075,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans,
 				goto out;
 			}
 		}
-		add_pinned_bytes(info, -num_bytes, owner_objectid,
-				 root_objectid);
 	} else {
 		if (found_extent) {
 			BUG_ON(is_data && refs_to_drop !=
@@ -7170,19 +7206,19 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 	int ret;
 
 	if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
-		ret = btrfs_add_delayed_tree_ref(fs_info, trans,
-						 buf->start, buf->len,
-						 parent,
+		int old_ref_mod, new_ref_mod;
+
+		ret = btrfs_add_delayed_tree_ref(fs_info, trans, buf->start,
+						 buf->len, parent,
 						 root->root_key.objectid,
 						 btrfs_header_level(buf),
-						 BTRFS_DROP_DELAYED_REF, NULL);
+						 BTRFS_DROP_DELAYED_REF, NULL,
+						 &old_ref_mod, &new_ref_mod);
 		BUG_ON(ret); /* -ENOMEM */
+		pin = old_ref_mod >= 0 && new_ref_mod < 0;
 	}
 
-	if (!last_ref)
-		return;
-
-	if (btrfs_header_generation(buf) == trans->transid) {
+	if (last_ref && btrfs_header_generation(buf) == trans->transid) {
 		struct btrfs_block_group_cache *cache;
 
 		if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
@@ -7191,6 +7227,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 				goto out;
 		}
 
+		pin = 0;
 		cache = btrfs_lookup_block_group(fs_info, buf->start);
 
 		if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) {
@@ -7206,18 +7243,19 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 		btrfs_free_reserved_bytes(cache, buf->len, 0);
 		btrfs_put_block_group(cache);
 		trace_btrfs_reserved_extent_free(fs_info, buf->start, buf->len);
-		pin = 0;
 	}
 out:
 	if (pin)
 		add_pinned_bytes(fs_info, buf->len, btrfs_header_level(buf),
 				 root->root_key.objectid);
 
-	/*
-	 * Deleting the buffer, clear the corrupt flag since it doesn't matter
-	 * anymore.
-	 */
-	clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags);
+	if (last_ref) {
+		/*
+		 * Deleting the buffer, clear the corrupt flag since it doesn't
+		 * matter anymore.
+		 */
+		clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags);
+	}
 }
 
 /* Can return -ENOMEM */
@@ -7226,12 +7264,12 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans,
 		      u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid,
 		      u64 owner, u64 offset)
 {
+	int old_ref_mod, new_ref_mod;
 	int ret;
 
 	if (btrfs_is_testing(fs_info))
 		return 0;
 
-	add_pinned_bytes(fs_info, num_bytes, owner, root_objectid);
 
 	/*
 	 * tree log blocks never actually go into the extent allocation
@@ -7241,19 +7279,25 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans,
 		WARN_ON(owner >= BTRFS_FIRST_FREE_OBJECTID);
 		/* unlocks the pinned mutex */
 		btrfs_pin_extent(fs_info, bytenr, num_bytes, 1);
+		old_ref_mod = new_ref_mod = 0;
 		ret = 0;
 	} else if (owner < BTRFS_FIRST_FREE_OBJECTID) {
 		ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr,
-					num_bytes,
-					parent, root_objectid, (int)owner,
-					BTRFS_DROP_DELAYED_REF, NULL);
+						 num_bytes, parent,
+						 root_objectid, (int)owner,
+						 BTRFS_DROP_DELAYED_REF, NULL,
+						 &old_ref_mod, &new_ref_mod);
 	} else {
 		ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr,
-						num_bytes,
-						parent, root_objectid, owner,
-						offset, 0,
-						BTRFS_DROP_DELAYED_REF);
+						 num_bytes, parent,
+						 root_objectid, owner, offset,
+						 0, BTRFS_DROP_DELAYED_REF,
+						 &old_ref_mod, &new_ref_mod);
 	}
+
+	if (ret == 0 && old_ref_mod >= 0 && new_ref_mod < 0)
+		add_pinned_bytes(fs_info, num_bytes, owner, root_objectid);
+
 	return ret;
 }
 
@@ -7956,7 +8000,7 @@ int btrfs_reserve_extent(struct btrfs_root *root, u64 ram_bytes,
 	u64 flags;
 	int ret;
 
-	flags = btrfs_get_alloc_profile(root, is_data);
+	flags = get_alloc_profile_by_root(root, is_data);
 again:
 	WARN_ON(num_bytes < fs_info->sectorsize);
 	ret = find_free_extent(fs_info, ram_bytes, num_bytes, empty_size,
@@ -8200,9 +8244,9 @@ int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans,
 	BUG_ON(root_objectid == BTRFS_TREE_LOG_OBJECTID);
 
 	ret = btrfs_add_delayed_data_ref(fs_info, trans, ins->objectid,
-					 ins->offset, 0,
-					 root_objectid, owner, offset,
-					 ram_bytes, BTRFS_ADD_DELAYED_EXTENT);
+					 ins->offset, 0, root_objectid, owner,
+					 offset, ram_bytes,
+					 BTRFS_ADD_DELAYED_EXTENT, NULL, NULL);
 	return ret;
 }
 
@@ -8422,11 +8466,11 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans,
 		extent_op->is_data = false;
 		extent_op->level = level;
 
-		ret = btrfs_add_delayed_tree_ref(fs_info, trans,
-						 ins.objectid, ins.offset,
-						 parent, root_objectid, level,
+		ret = btrfs_add_delayed_tree_ref(fs_info, trans, ins.objectid,
+						 ins.offset, parent,
+						 root_objectid, level,
 						 BTRFS_ADD_DELAYED_EXTENT,
-						 extent_op);
+						 extent_op, NULL, NULL);
 		if (ret)
 			goto out_free_delayed;
 	}
@@ -10059,19 +10103,9 @@ int btrfs_read_block_groups(struct btrfs_fs_info *info)
 		}
 
 		trace_btrfs_add_block_group(info, cache, 0);
-		ret = update_space_info(info, cache->flags, found_key.offset,
-					btrfs_block_group_used(&cache->item),
-					cache->bytes_super, &space_info);
-		if (ret) {
-			btrfs_remove_free_space_cache(cache);
-			spin_lock(&info->block_group_cache_lock);
-			rb_erase(&cache->cache_node,
-				 &info->block_group_cache_tree);
-			RB_CLEAR_NODE(&cache->cache_node);
-			spin_unlock(&info->block_group_cache_lock);
-			btrfs_put_block_group(cache);
-			goto error;
-		}
+		update_space_info(info, cache->flags, found_key.offset,
+				  btrfs_block_group_used(&cache->item),
+				  cache->bytes_super, &space_info);
 
 		cache->space_info = space_info;
 
@@ -10203,16 +10237,19 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 	}
 #endif
 	/*
-	 * Call to ensure the corresponding space_info object is created and
-	 * assigned to our block group, but don't update its counters just yet.
-	 * We want our bg to be added to the rbtree with its ->space_info set.
+	 * Ensure the corresponding space_info object is created and
+	 * assigned to our block group. We want our bg to be added to the rbtree
+	 * with its ->space_info set.
 	 */
-	ret = update_space_info(fs_info, cache->flags, 0, 0, 0,
-				&cache->space_info);
-	if (ret) {
-		btrfs_remove_free_space_cache(cache);
-		btrfs_put_block_group(cache);
-		return ret;
+	cache->space_info = __find_space_info(fs_info, cache->flags);
+	if (!cache->space_info) {
+		ret = create_space_info(fs_info, cache->flags,
+				       &cache->space_info);
+		if (ret) {
+			btrfs_remove_free_space_cache(cache);
+			btrfs_put_block_group(cache);
+			return ret;
+		}
 	}
 
 	ret = btrfs_add_block_group_cache(fs_info, cache);
@@ -10227,18 +10264,8 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans,
 	 * the rbtree, update the space info's counters.
 	 */
 	trace_btrfs_add_block_group(fs_info, cache, 1);
-	ret = update_space_info(fs_info, cache->flags, size, bytes_used,
+	update_space_info(fs_info, cache->flags, size, bytes_used,
 				cache->bytes_super, &cache->space_info);
-	if (ret) {
-		btrfs_remove_free_space_cache(cache);
-		spin_lock(&fs_info->block_group_cache_lock);
-		rb_erase(&cache->cache_node,
-			 &fs_info->block_group_cache_tree);
-		RB_CLEAR_NODE(&cache->cache_node);
-		spin_unlock(&fs_info->block_group_cache_lock);
-		btrfs_put_block_group(cache);
-		return ret;
-	}
 	update_global_block_rsv(fs_info);
 
 	__link_block_group(cache->space_info, cache);
@@ -10786,21 +10813,21 @@ int btrfs_init_space_info(struct btrfs_fs_info *fs_info)
 		mixed = 1;
 
 	flags = BTRFS_BLOCK_GROUP_SYSTEM;
-	ret = update_space_info(fs_info, flags, 0, 0, 0, &space_info);
+	ret = create_space_info(fs_info, flags, &space_info);
 	if (ret)
 		goto out;
 
 	if (mixed) {
 		flags = BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA;
-		ret = update_space_info(fs_info, flags, 0, 0, 0, &space_info);
+		ret = create_space_info(fs_info, flags, &space_info);
 	} else {
 		flags = BTRFS_BLOCK_GROUP_METADATA;
-		ret = update_space_info(fs_info, flags, 0, 0, 0, &space_info);
+		ret = create_space_info(fs_info, flags, &space_info);
 		if (ret)
 			goto out;
 
 		flags = BTRFS_BLOCK_GROUP_DATA;
-		ret = update_space_info(fs_info, flags, 0, 0, 0, &space_info);
+		ret = create_space_info(fs_info, flags, &space_info);
 	}
 out:
 	return ret;

fs/btrfs/extent_io.c

@@ -87,19 +87,9 @@ void btrfs_leak_debug_check(void)
 static inline void __btrfs_debug_check_extent_io_range(const char *caller,
 		struct extent_io_tree *tree, u64 start, u64 end)
 {
-	struct inode *inode;
-	u64 isize;
-
-	if (!tree->mapping)
-		return;
-
-	inode = tree->mapping->host;
-	isize = i_size_read(inode);
-	if (end >= PAGE_SIZE && (end % 2) == 0 && end != isize - 1) {
-		btrfs_debug_rl(BTRFS_I(inode)->root->fs_info,
-		    "%s: ino %llu isize %llu odd range [%llu,%llu]",
-			caller, btrfs_ino(BTRFS_I(inode)), isize, start, end);
-	}
+	if (tree->ops && tree->ops->check_extent_io_range)
+		tree->ops->check_extent_io_range(tree->private_data, caller,
+						 start, end);
 }
 #else
 #define btrfs_leak_debug_add(new, head)	do {} while (0)
@@ -154,9 +144,9 @@ static noinline void flush_write_bio(void *data);
 static inline struct btrfs_fs_info *
 tree_fs_info(struct extent_io_tree *tree)
 {
-	if (!tree->mapping)
-		return NULL;
-	return btrfs_sb(tree->mapping->host->i_sb);
+	if (tree->ops)
+		return tree->ops->tree_fs_info(tree->private_data);
+	return NULL;
 }
 
 int __init extent_io_init(void)
@@ -214,13 +204,13 @@ void extent_io_exit(void)
 }
 
 void extent_io_tree_init(struct extent_io_tree *tree,
-			 struct address_space *mapping)
+			 void *private_data)
 {
 	tree->state = RB_ROOT;
 	tree->ops = NULL;
 	tree->dirty_bytes = 0;
 	spin_lock_init(&tree->lock);
-	tree->mapping = mapping;
+	tree->private_data = private_data;
 }
 
 static struct extent_state *alloc_extent_state(gfp_t mask)
@@ -370,8 +360,7 @@ static void merge_cb(struct extent_io_tree *tree, struct extent_state *new,
 		     struct extent_state *other)
 {
 	if (tree->ops && tree->ops->merge_extent_hook)
-		tree->ops->merge_extent_hook(tree->mapping->host, new,
-					     other);
+		tree->ops->merge_extent_hook(tree->private_data, new, other);
 }
 
 /*
@@ -422,15 +411,14 @@ static void set_state_cb(struct extent_io_tree *tree,
 			 struct extent_state *state, unsigned *bits)
 {
 	if (tree->ops && tree->ops->set_bit_hook)
-		tree->ops->set_bit_hook(tree->mapping->host, state, bits);
+		tree->ops->set_bit_hook(tree->private_data, state, bits);
 }
 
 static void clear_state_cb(struct extent_io_tree *tree,
 			   struct extent_state *state, unsigned *bits)
 {
 	if (tree->ops && tree->ops->clear_bit_hook)
-		tree->ops->clear_bit_hook(BTRFS_I(tree->mapping->host),
-				state, bits);
+		tree->ops->clear_bit_hook(tree->private_data, state, bits);
 }
 
 static void set_state_bits(struct extent_io_tree *tree,
@@ -479,7 +467,7 @@ static void split_cb(struct extent_io_tree *tree, struct extent_state *orig,
 		     u64 split)
 {
 	if (tree->ops && tree->ops->split_extent_hook)
-		tree->ops->split_extent_hook(tree->mapping->host, orig, split);
+		tree->ops->split_extent_hook(tree->private_data, orig, split);
 }
 
 /*
@@ -1403,17 +1391,7 @@ void extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end)
  */
 static void set_range_writeback(struct extent_io_tree *tree, u64 start, u64 end)
 {
-	unsigned long index = start >> PAGE_SHIFT;
-	unsigned long end_index = end >> PAGE_SHIFT;
-	struct page *page;
-
-	while (index <= end_index) {
-		page = find_get_page(tree->mapping, index);
-		BUG_ON(!page); /* Pages should be in the extent_io_tree */
-		set_page_writeback(page);
-		put_page(page);
-		index++;
-	}
+	tree->ops->set_range_writeback(tree->private_data, start, end);
 }
 
 /* find the first state struct with 'bits' set after 'start', and
@@ -1962,11 +1940,12 @@ static void check_page_uptodate(struct extent_io_tree *tree, struct page *page)
 		SetPageUptodate(page);
 }
 
-int free_io_failure(struct btrfs_inode *inode, struct io_failure_record *rec)
+int free_io_failure(struct extent_io_tree *failure_tree,
+		    struct extent_io_tree *io_tree,
+		    struct io_failure_record *rec)
 {
 	int ret;
 	int err = 0;
-	struct extent_io_tree *failure_tree = &inode->io_failure_tree;
 
 	set_state_failrec(failure_tree, rec->start, NULL);
 	ret = clear_extent_bits(failure_tree, rec->start,
@@ -1975,7 +1954,7 @@ int free_io_failure(struct btrfs_inode *inode, struct io_failure_record *rec)
 	if (ret)
 		err = ret;
 
-	ret = clear_extent_bits(&inode->io_tree, rec->start,
+	ret = clear_extent_bits(io_tree, rec->start,
 				rec->start + rec->len - 1,
 				EXTENT_DAMAGED);
 	if (ret && !err)
@@ -1995,11 +1974,10 @@ int free_io_failure(struct btrfs_inode *inode, struct io_failure_record *rec)
  * currently, there can be no more than two copies of every data bit. thus,
  * exactly one rewrite is required.
  */
-int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length,
-		u64 logical, struct page *page,
-		unsigned int pg_offset, int mirror_num)
+int repair_io_failure(struct btrfs_fs_info *fs_info, u64 ino, u64 start,
+		      u64 length, u64 logical, struct page *page,
+		      unsigned int pg_offset, int mirror_num)
 {
-	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	struct bio *bio;
 	struct btrfs_device *dev;
 	u64 map_length = 0;
@@ -2010,9 +1988,7 @@ int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length,
 	ASSERT(!(fs_info->sb->s_flags & MS_RDONLY));
 	BUG_ON(!mirror_num);
 
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-	if (!bio)
-		return -EIO;
+	bio = btrfs_io_bio_alloc(1);
 	bio->bi_iter.bi_size = 0;
 	map_length = length;
 
@@ -2071,7 +2047,7 @@ int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length,
 
 	btrfs_info_rl_in_rcu(fs_info,
 		"read error corrected: ino %llu off %llu (dev %s sector %llu)",
-				  btrfs_ino(inode), start,
+				  ino, start,
 				  rcu_str_deref(dev->name), sector);
 	btrfs_bio_counter_dec(fs_info);
 	bio_put(bio);
@@ -2091,8 +2067,7 @@ int repair_eb_io_failure(struct btrfs_fs_info *fs_info,
 	for (i = 0; i < num_pages; i++) {
 		struct page *p = eb->pages[i];
 
-		ret = repair_io_failure(BTRFS_I(fs_info->btree_inode), start,
-					PAGE_SIZE, start, p,
+		ret = repair_io_failure(fs_info, 0, start, PAGE_SIZE, start, p,
 					start - page_offset(p), mirror_num);
 		if (ret)
 			break;
@@ -2106,24 +2081,24 @@ int repair_eb_io_failure(struct btrfs_fs_info *fs_info,
  * each time an IO finishes, we do a fast check in the IO failure tree
  * to see if we need to process or clean up an io_failure_record
  */
-int clean_io_failure(struct btrfs_inode *inode, u64 start, struct page *page,
-		     unsigned int pg_offset)
+int clean_io_failure(struct btrfs_fs_info *fs_info,
+		     struct extent_io_tree *failure_tree,
+		     struct extent_io_tree *io_tree, u64 start,
+		     struct page *page, u64 ino, unsigned int pg_offset)
 {
 	u64 private;
 	struct io_failure_record *failrec;
-	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	struct extent_state *state;
 	int num_copies;
 	int ret;
 
 	private = 0;
-	ret = count_range_bits(&inode->io_failure_tree, &private,
-				(u64)-1, 1, EXTENT_DIRTY, 0);
+	ret = count_range_bits(failure_tree, &private, (u64)-1, 1,
+			       EXTENT_DIRTY, 0);
 	if (!ret)
 		return 0;
 
-	ret = get_state_failrec(&inode->io_failure_tree, start,
-			&failrec);
+	ret = get_state_failrec(failure_tree, start, &failrec);
 	if (ret)
 		return 0;
 
@@ -2139,25 +2114,25 @@ int clean_io_failure(struct btrfs_inode *inode, u64 start, struct page *page,
 	if (fs_info->sb->s_flags & MS_RDONLY)
 		goto out;
 
-	spin_lock(&inode->io_tree.lock);
-	state = find_first_extent_bit_state(&inode->io_tree,
+	spin_lock(&io_tree->lock);
+	state = find_first_extent_bit_state(io_tree,
 					    failrec->start,
 					    EXTENT_LOCKED);
-	spin_unlock(&inode->io_tree.lock);
+	spin_unlock(&io_tree->lock);
 
 	if (state && state->start <= failrec->start &&
 	    state->end >= failrec->start + failrec->len - 1) {
 		num_copies = btrfs_num_copies(fs_info, failrec->logical,
 					      failrec->len);
 		if (num_copies > 1)  {
-			repair_io_failure(inode, start, failrec->len,
-					  failrec->logical, page,
-					  pg_offset, failrec->failed_mirror);
+			repair_io_failure(fs_info, ino, start, failrec->len,
+					  failrec->logical, page, pg_offset,
+					  failrec->failed_mirror);
 		}
 	}
 
 out:
-	free_io_failure(inode, failrec);
+	free_io_failure(failure_tree, io_tree, failrec);
 
 	return 0;
 }
@@ -2357,10 +2332,7 @@ struct bio *btrfs_create_repair_bio(struct inode *inode, struct bio *failed_bio,
 	struct btrfs_io_bio *btrfs_failed_bio;
 	struct btrfs_io_bio *btrfs_bio;
 
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-	if (!bio)
-		return NULL;
-
+	bio = btrfs_io_bio_alloc(1);
 	bio->bi_end_io = endio_func;
 	bio->bi_iter.bi_sector = failrec->logical >> 9;
 	bio->bi_bdev = fs_info->fs_devices->latest_bdev;
@@ -2398,6 +2370,7 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 	struct io_failure_record *failrec;
 	struct inode *inode = page->mapping->host;
 	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;
+	struct extent_io_tree *failure_tree = &BTRFS_I(inode)->io_failure_tree;
 	struct bio *bio;
 	int read_mode = 0;
 	blk_status_t status;
@@ -2411,7 +2384,7 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 
 	ret = btrfs_check_repairable(inode, failed_bio, failrec, failed_mirror);
 	if (!ret) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, tree, failrec);
 		return -EIO;
 	}
 
@@ -2424,7 +2397,7 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 				      (int)phy_offset, failed_bio->bi_end_io,
 				      NULL);
 	if (!bio) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, tree, failrec);
 		return -EIO;
 	}
 	bio_set_op_attrs(bio, REQ_OP_READ, read_mode);
@@ -2433,10 +2406,10 @@ static int bio_readpage_error(struct bio *failed_bio, u64 phy_offset,
 		"Repair Read Error: submitting new read[%#x] to this_mirror=%d, in_validation=%d",
 		read_mode, failrec->this_mirror, failrec->in_validation);
 
-	status = tree->ops->submit_bio_hook(inode, bio, failrec->this_mirror,
+	status = tree->ops->submit_bio_hook(tree->private_data, bio, failrec->this_mirror,
 					 failrec->bio_flags, 0);
 	if (status) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, tree, failrec);
 		bio_put(bio);
 		ret = blk_status_to_errno(status);
 	}
@@ -2542,7 +2515,7 @@ static void end_bio_extent_readpage(struct bio *bio)
 	struct bio_vec *bvec;
 	int uptodate = !bio->bi_status;
 	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
-	struct extent_io_tree *tree;
+	struct extent_io_tree *tree, *failure_tree;
 	u64 offset = 0;
 	u64 start;
 	u64 end;
@@ -2563,6 +2536,7 @@ static void end_bio_extent_readpage(struct bio *bio)
 			(u64)bio->bi_iter.bi_sector, bio->bi_status,
 			io_bio->mirror_num);
 		tree = &BTRFS_I(inode)->io_tree;
+		failure_tree = &BTRFS_I(inode)->io_failure_tree;
 
 		/* We always issue full-page reads, but if some block
 		 * in a page fails to read, blk_update_request() will
@@ -2592,8 +2566,10 @@ static void end_bio_extent_readpage(struct bio *bio)
 			if (ret)
 				uptodate = 0;
 			else
-				clean_io_failure(BTRFS_I(inode), start,
-						page, 0);
+				clean_io_failure(BTRFS_I(inode)->root->fs_info,
+						 failure_tree, tree, start,
+						 page,
+						 btrfs_ino(BTRFS_I(inode)), 0);
 		}
 
 		if (likely(uptodate))
@@ -2682,67 +2658,70 @@ readpage_ok:
 }
 
 /*
- * this allocates from the btrfs_bioset.  We're returning a bio right now
- * but you can call btrfs_io_bio for the appropriate container_of magic
+ * Initialize the members up to but not including 'bio'. Use after allocating a
+ * new bio by bio_alloc_bioset as it does not initialize the bytes outside of
+ * 'bio' because use of __GFP_ZERO is not supported.
  */
-struct bio *
-btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
-		gfp_t gfp_flags)
+static inline void btrfs_io_bio_init(struct btrfs_io_bio *btrfs_bio)
 {
-	struct btrfs_io_bio *btrfs_bio;
-	struct bio *bio;
-
-	bio = bio_alloc_bioset(gfp_flags, nr_vecs, btrfs_bioset);
+	memset(btrfs_bio, 0, offsetof(struct btrfs_io_bio, bio));
+}
 
-	if (bio == NULL && (current->flags & PF_MEMALLOC)) {
-		while (!bio && (nr_vecs /= 2)) {
-			bio = bio_alloc_bioset(gfp_flags,
-					       nr_vecs, btrfs_bioset);
-		}
-	}
+/*
+ * The following helpers allocate a bio. As it's backed by a bioset, it'll
+ * never fail.  We're returning a bio right now but you can call btrfs_io_bio
+ * for the appropriate container_of magic
+ */
+struct bio *btrfs_bio_alloc(struct block_device *bdev, u64 first_byte)
+{
+	struct bio *bio;
 
-	if (bio) {
-		bio->bi_bdev = bdev;
-		bio->bi_iter.bi_sector = first_sector;
-		btrfs_bio = btrfs_io_bio(bio);
-		btrfs_bio->csum = NULL;
-		btrfs_bio->csum_allocated = NULL;
-		btrfs_bio->end_io = NULL;
-	}
+	bio = bio_alloc_bioset(GFP_NOFS, BIO_MAX_PAGES, btrfs_bioset);
+	bio->bi_bdev = bdev;
+	bio->bi_iter.bi_sector = first_byte >> 9;
+	btrfs_io_bio_init(btrfs_io_bio(bio));
 	return bio;
 }
 
-struct bio *btrfs_bio_clone(struct bio *bio, gfp_t gfp_mask)
+struct bio *btrfs_bio_clone(struct bio *bio)
 {
 	struct btrfs_io_bio *btrfs_bio;
 	struct bio *new;
 
-	new = bio_clone_bioset(bio, gfp_mask, btrfs_bioset);
-	if (new) {
-		btrfs_bio = btrfs_io_bio(new);
-		btrfs_bio->csum = NULL;
-		btrfs_bio->csum_allocated = NULL;
-		btrfs_bio->end_io = NULL;
-	}
+	/* Bio allocation backed by a bioset does not fail */
+	new = bio_clone_fast(bio, GFP_NOFS, btrfs_bioset);
+	btrfs_bio = btrfs_io_bio(new);
+	btrfs_io_bio_init(btrfs_bio);
+	btrfs_bio->iter = bio->bi_iter;
 	return new;
 }
 
-/* this also allocates from the btrfs_bioset */
-struct bio *btrfs_io_bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs)
+struct bio *btrfs_io_bio_alloc(unsigned int nr_iovecs)
 {
-	struct btrfs_io_bio *btrfs_bio;
 	struct bio *bio;
 
-	bio = bio_alloc_bioset(gfp_mask, nr_iovecs, btrfs_bioset);
-	if (bio) {
-		btrfs_bio = btrfs_io_bio(bio);
-		btrfs_bio->csum = NULL;
-		btrfs_bio->csum_allocated = NULL;
-		btrfs_bio->end_io = NULL;
-	}
+	/* Bio allocation backed by a bioset does not fail */
+	bio = bio_alloc_bioset(GFP_NOFS, nr_iovecs, btrfs_bioset);
+	btrfs_io_bio_init(btrfs_io_bio(bio));
 	return bio;
 }
 
+struct bio *btrfs_bio_clone_partial(struct bio *orig, int offset, int size)
+{
+	struct bio *bio;
+	struct btrfs_io_bio *btrfs_bio;
+
+	/* this will never fail when it's backed by a bioset */
+	bio = bio_clone_fast(orig, GFP_NOFS, btrfs_bioset);
+	ASSERT(bio);
+
+	btrfs_bio = btrfs_io_bio(bio);
+	btrfs_io_bio_init(btrfs_bio);
+
+	bio_trim(bio, offset >> 9, size >> 9);
+	btrfs_bio->iter = bio->bi_iter;
+	return bio;
+}
 
 static int __must_check submit_one_bio(struct bio *bio, int mirror_num,
 				       unsigned long bio_flags)
@@ -2759,7 +2738,7 @@ static int __must_check submit_one_bio(struct bio *bio, int mirror_num,
 	bio_get(bio);
 
 	if (tree->ops)
-		ret = tree->ops->submit_bio_hook(page->mapping->host, bio,
+		ret = tree->ops->submit_bio_hook(tree->private_data, bio,
 					   mirror_num, bio_flags, start);
 	else
 		btrfsic_submit_bio(bio);
@@ -2822,11 +2801,7 @@ static int submit_extent_page(int op, int op_flags, struct extent_io_tree *tree,
 		}
 	}
 
-	bio = btrfs_bio_alloc(bdev, sector, BIO_MAX_PAGES,
-			GFP_NOFS | __GFP_HIGH);
-	if (!bio)
-		return -ENOMEM;
-
+	bio = btrfs_bio_alloc(bdev, sector << 9);
 	bio_add_page(bio, page, page_size, offset);
 	bio->bi_end_io = end_io_func;
 	bio->bi_private = tree;
@@ -3762,7 +3737,7 @@ static noinline_for_stack int write_one_eb(struct extent_buffer *eb,
 		 * header 0 1 2 .. N ... data_N .. data_2 data_1 data_0
 		 */
 		start = btrfs_item_nr_offset(nritems);
-		end = btrfs_leaf_data(eb) + leaf_data_end(fs_info, eb);
+		end = BTRFS_LEAF_DATA_OFFSET + leaf_data_end(fs_info, eb);
 		memzero_extent_buffer(eb, start, end - start);
 	}
 
@@ -4468,29 +4443,25 @@ try_submit_last:
 }
 
 /*
- * Sanity check for fiemap cache
+ * Emit last fiemap cache
  *
- * All fiemap cache should be submitted by emit_fiemap_extent()
- * Iteration should be terminated either by last fiemap extent or
- * fieinfo->fi_extents_max.
- * So no cached fiemap should exist.
+ * The last fiemap cache may still be cached in the following case:
+ * 0		      4k		    8k
+ * |<- Fiemap range ->|
+ * |<------------  First extent ----------->|
+ *
+ * In this case, the first extent range will be cached but not emitted.
+ * So we must emit it before ending extent_fiemap().
  */
-static int check_fiemap_cache(struct btrfs_fs_info *fs_info,
-			       struct fiemap_extent_info *fieinfo,
-			       struct fiemap_cache *cache)
+static int emit_last_fiemap_cache(struct btrfs_fs_info *fs_info,
+				  struct fiemap_extent_info *fieinfo,
+				  struct fiemap_cache *cache)
 {
 	int ret;
 
 	if (!cache->cached)
 		return 0;
 
-	/* Small and recoverbale problem, only to info developer */
-#ifdef CONFIG_BTRFS_DEBUG
-	WARN_ON(1);
-#endif
-	btrfs_warn(fs_info,
-		   "unhandled fiemap cache detected: offset=%llu phys=%llu len=%llu flags=0x%x",
-		   cache->offset, cache->phys, cache->len, cache->flags);
 	ret = fiemap_fill_next_extent(fieinfo, cache->offset, cache->phys,
 				      cache->len, cache->flags);
 	cache->cached = false;
@@ -4706,7 +4677,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	}
 out_free:
 	if (!ret)
-		ret = check_fiemap_cache(root->fs_info, fieinfo, &cache);
+		ret = emit_last_fiemap_cache(root->fs_info, fieinfo, &cache);
 	free_extent_map(em);
 out:
 	btrfs_free_path(path);

fs/btrfs/extent_io.h

@@ -92,9 +92,9 @@ struct btrfs_inode;
 struct btrfs_io_bio;
 struct io_failure_record;
 
-typedef	blk_status_t (extent_submit_bio_hook_t)(struct inode *inode,
-		struct bio *bio, int mirror_num, unsigned long bio_flags,
-		u64 bio_offset);
+typedef	blk_status_t (extent_submit_bio_hook_t)(void *private_data, struct bio *bio,
+				       int mirror_num, unsigned long bio_flags,
+				       u64 bio_offset);
 struct extent_io_ops {
 	/*
 	 * The following callbacks must be allways defined, the function
@@ -108,32 +108,36 @@ struct extent_io_ops {
 			      size_t size, struct bio *bio,
 			      unsigned long bio_flags);
 	int (*readpage_io_failed_hook)(struct page *page, int failed_mirror);
+	struct btrfs_fs_info *(*tree_fs_info)(void *private_data);
+	void (*set_range_writeback)(void *private_data, u64 start, u64 end);
 
 	/*
 	 * Optional hooks, called if the pointer is not NULL
 	 */
-	int (*fill_delalloc)(struct inode *inode, struct page *locked_page,
+	int (*fill_delalloc)(void *private_data, struct page *locked_page,
 			     u64 start, u64 end, int *page_started,
 			     unsigned long *nr_written);
 
 	int (*writepage_start_hook)(struct page *page, u64 start, u64 end);
 	void (*writepage_end_io_hook)(struct page *page, u64 start, u64 end,
 				      struct extent_state *state, int uptodate);
-	void (*set_bit_hook)(struct inode *inode, struct extent_state *state,
+	void (*set_bit_hook)(void *private_data, struct extent_state *state,
 			     unsigned *bits);
-	void (*clear_bit_hook)(struct btrfs_inode *inode,
+	void (*clear_bit_hook)(void *private_data,
 			struct extent_state *state,
 			unsigned *bits);
-	void (*merge_extent_hook)(struct inode *inode,
+	void (*merge_extent_hook)(void *private_data,
 				  struct extent_state *new,
 				  struct extent_state *other);
-	void (*split_extent_hook)(struct inode *inode,
+	void (*split_extent_hook)(void *private_data,
 				  struct extent_state *orig, u64 split);
+	void (*check_extent_io_range)(void *private_data, const char *caller,
+				      u64 start, u64 end);
 };
 
 struct extent_io_tree {
 	struct rb_root state;
-	struct address_space *mapping;
+	void *private_data;
 	u64 dirty_bytes;
 	int track_uptodate;
 	spinlock_t lock;
@@ -205,12 +209,46 @@ struct extent_buffer {
  */
 struct extent_changeset {
 	/* How many bytes are set/cleared in this operation */
-	u64 bytes_changed;
+	unsigned int bytes_changed;
 
 	/* Changed ranges */
 	struct ulist range_changed;
 };
 
+static inline void extent_changeset_init(struct extent_changeset *changeset)
+{
+	changeset->bytes_changed = 0;
+	ulist_init(&changeset->range_changed);
+}
+
+static inline struct extent_changeset *extent_changeset_alloc(void)
+{
+	struct extent_changeset *ret;
+
+	ret = kmalloc(sizeof(*ret), GFP_KERNEL);
+	if (!ret)
+		return NULL;
+
+	extent_changeset_init(ret);
+	return ret;
+}
+
+static inline void extent_changeset_release(struct extent_changeset *changeset)
+{
+	if (!changeset)
+		return;
+	changeset->bytes_changed = 0;
+	ulist_release(&changeset->range_changed);
+}
+
+static inline void extent_changeset_free(struct extent_changeset *changeset)
+{
+	if (!changeset)
+		return;
+	extent_changeset_release(changeset);
+	kfree(changeset);
+}
+
 static inline void extent_set_compress_type(unsigned long *bio_flags,
 					    int compress_type)
 {
@@ -230,8 +268,7 @@ typedef struct extent_map *(get_extent_t)(struct btrfs_inode *inode,
 					  u64 start, u64 len,
 					  int create);
 
-void extent_io_tree_init(struct extent_io_tree *tree,
-			 struct address_space *mapping);
+void extent_io_tree_init(struct extent_io_tree *tree, void *private_data);
 int try_release_extent_mapping(struct extent_map_tree *map,
 			       struct extent_io_tree *tree, struct page *page,
 			       gfp_t mask);
@@ -459,20 +496,21 @@ void extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end,
 				 u64 delalloc_end, struct page *locked_page,
 				 unsigned bits_to_clear,
 				 unsigned long page_ops);
-struct bio *
-btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
-		gfp_t gfp_flags);
-struct bio *btrfs_io_bio_alloc(gfp_t gfp_mask, unsigned int nr_iovecs);
-struct bio *btrfs_bio_clone(struct bio *bio, gfp_t gfp_mask);
+struct bio *btrfs_bio_alloc(struct block_device *bdev, u64 first_byte);
+struct bio *btrfs_io_bio_alloc(unsigned int nr_iovecs);
+struct bio *btrfs_bio_clone(struct bio *bio);
+struct bio *btrfs_bio_clone_partial(struct bio *orig, int offset, int size);
 
 struct btrfs_fs_info;
 struct btrfs_inode;
 
-int repair_io_failure(struct btrfs_inode *inode, u64 start, u64 length,
-		u64 logical, struct page *page,
-		unsigned int pg_offset, int mirror_num);
-int clean_io_failure(struct btrfs_inode *inode, u64 start,
-		struct page *page, unsigned int pg_offset);
+int repair_io_failure(struct btrfs_fs_info *fs_info, u64 ino, u64 start,
+		      u64 length, u64 logical, struct page *page,
+		      unsigned int pg_offset, int mirror_num);
+int clean_io_failure(struct btrfs_fs_info *fs_info,
+		     struct extent_io_tree *failure_tree,
+		     struct extent_io_tree *io_tree, u64 start,
+		     struct page *page, u64 ino, unsigned int pg_offset);
 void end_extent_writepage(struct page *page, int err, u64 start, u64 end);
 int repair_eb_io_failure(struct btrfs_fs_info *fs_info,
 			 struct extent_buffer *eb, int mirror_num);
@@ -507,7 +545,9 @@ struct bio *btrfs_create_repair_bio(struct inode *inode, struct bio *failed_bio,
 				    struct io_failure_record *failrec,
 				    struct page *page, int pg_offset, int icsum,
 				    bio_end_io_t *endio_func, void *data);
-int free_io_failure(struct btrfs_inode *inode, struct io_failure_record *rec);
+int free_io_failure(struct extent_io_tree *failure_tree,
+		    struct extent_io_tree *io_tree,
+		    struct io_failure_record *rec);
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 noinline u64 find_lock_delalloc_range(struct inode *inode,
 				      struct extent_io_tree *tree,

fs/btrfs/file-item.c

@@ -164,7 +164,8 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio
 				   u64 logical_offset, u32 *dst, int dio)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
-	struct bio_vec *bvec;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
 	struct btrfs_io_bio *btrfs_bio = btrfs_io_bio(bio);
 	struct btrfs_csum_item *item = NULL;
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
@@ -177,7 +178,7 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio
 	u64 page_bytes_left;
 	u32 diff;
 	int nblocks;
-	int count = 0, i;
+	int count = 0;
 	u16 csum_size = btrfs_super_csum_size(fs_info->super_copy);
 
 	path = btrfs_alloc_path();
@@ -206,8 +207,6 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio
 	if (bio->bi_iter.bi_size > PAGE_SIZE * 8)
 		path->reada = READA_FORWARD;
 
-	WARN_ON(bio->bi_vcnt <= 0);
-
 	/*
 	 * the free space stuff is only read when it hasn't been
 	 * updated in the current transaction.  So, we can safely
@@ -223,13 +222,13 @@ static blk_status_t __btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio
 	if (dio)
 		offset = logical_offset;
 
-	bio_for_each_segment_all(bvec, bio, i) {
-		page_bytes_left = bvec->bv_len;
+	bio_for_each_segment(bvec, bio, iter) {
+		page_bytes_left = bvec.bv_len;
 		if (count)
 			goto next;
 
 		if (!dio)
-			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
+			offset = page_offset(bvec.bv_page) + bvec.bv_offset;
 		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
 					       (u32 *)csum, nblocks);
 		if (count)
@@ -440,15 +439,15 @@ blk_status_t btrfs_csum_one_bio(struct inode *inode, struct bio *bio,
 	struct btrfs_ordered_sum *sums;
 	struct btrfs_ordered_extent *ordered = NULL;
 	char *data;
-	struct bio_vec *bvec;
+	struct bvec_iter iter;
+	struct bio_vec bvec;
 	int index;
 	int nr_sectors;
-	int i, j;
 	unsigned long total_bytes = 0;
 	unsigned long this_sum_bytes = 0;
+	int i;
 	u64 offset;
 
-	WARN_ON(bio->bi_vcnt <= 0);
 	sums = kzalloc(btrfs_ordered_sum_size(fs_info, bio->bi_iter.bi_size),
 		       GFP_NOFS);
 	if (!sums)
@@ -465,19 +464,19 @@ blk_status_t btrfs_csum_one_bio(struct inode *inode, struct bio *bio,
 	sums->bytenr = (u64)bio->bi_iter.bi_sector << 9;
 	index = 0;
 
-	bio_for_each_segment_all(bvec, bio, j) {
+	bio_for_each_segment(bvec, bio, iter) {
 		if (!contig)
-			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
+			offset = page_offset(bvec.bv_page) + bvec.bv_offset;
 
 		if (!ordered) {
 			ordered = btrfs_lookup_ordered_extent(inode, offset);
 			BUG_ON(!ordered); /* Logic error */
 		}
 
-		data = kmap_atomic(bvec->bv_page);
+		data = kmap_atomic(bvec.bv_page);
 
 		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info,
-						 bvec->bv_len + fs_info->sectorsize
+						 bvec.bv_len + fs_info->sectorsize
 						 - 1);
 
 		for (i = 0; i < nr_sectors; i++) {
@@ -504,12 +503,12 @@ blk_status_t btrfs_csum_one_bio(struct inode *inode, struct bio *bio,
 					+ total_bytes;
 				index = 0;
 
-				data = kmap_atomic(bvec->bv_page);
+				data = kmap_atomic(bvec.bv_page);
 			}
 
 			sums->sums[index] = ~(u32)0;
 			sums->sums[index]
-				= btrfs_csum_data(data + bvec->bv_offset
+				= btrfs_csum_data(data + bvec.bv_offset
 						+ (i * fs_info->sectorsize),
 						sums->sums[index],
 						fs_info->sectorsize);

fs/btrfs/file.c

@@ -1581,6 +1581,7 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct page **pages = NULL;
 	struct extent_state *cached_state = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	u64 release_bytes = 0;
 	u64 lockstart;
 	u64 lockend;
@@ -1628,7 +1629,9 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 		reserve_bytes = round_up(write_bytes + sector_offset,
 				fs_info->sectorsize);
 
-		ret = btrfs_check_data_free_space(inode, pos, write_bytes);
+		extent_changeset_release(data_reserved);
+		ret = btrfs_check_data_free_space(inode, &data_reserved, pos,
+						  write_bytes);
 		if (ret < 0) {
 			if ((BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
 						      BTRFS_INODE_PREALLOC)) &&
@@ -1657,8 +1660,9 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file,
 				reserve_bytes);
 		if (ret) {
 			if (!only_release_metadata)
-				btrfs_free_reserved_data_space(inode, pos,
-							       write_bytes);
+				btrfs_free_reserved_data_space(inode,
+						data_reserved, pos,
+						write_bytes);
 			else
 				btrfs_end_write_no_snapshoting(root);
 			break;
@@ -1740,8 +1744,9 @@ again:
 				__pos = round_down(pos,
 						   fs_info->sectorsize) +
 					(dirty_pages << PAGE_SHIFT);
-				btrfs_delalloc_release_space(inode, __pos,
-							     release_bytes);
+				btrfs_delalloc_release_space(inode,
+						data_reserved, __pos,
+						release_bytes);
 			}
 		}
 
@@ -1796,12 +1801,13 @@ again:
 			btrfs_delalloc_release_metadata(BTRFS_I(inode),
 					release_bytes);
 		} else {
-			btrfs_delalloc_release_space(inode,
-						round_down(pos, fs_info->sectorsize),
-						release_bytes);
+			btrfs_delalloc_release_space(inode, data_reserved,
+					round_down(pos, fs_info->sectorsize),
+					release_bytes);
 		}
 	}
 
+	extent_changeset_free(data_reserved);
 	return num_written ? num_written : ret;
 }
 
@@ -2405,10 +2411,13 @@ out:
  */
 static int find_first_non_hole(struct inode *inode, u64 *start, u64 *len)
 {
+	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct extent_map *em;
 	int ret = 0;
 
-	em = btrfs_get_extent(BTRFS_I(inode), NULL, 0, *start, *len, 0);
+	em = btrfs_get_extent(BTRFS_I(inode), NULL, 0,
+			      round_down(*start, fs_info->sectorsize),
+			      round_up(*len, fs_info->sectorsize), 0);
 	if (IS_ERR(em))
 		return PTR_ERR(em);
 
@@ -2784,6 +2793,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 {
 	struct inode *inode = file_inode(file);
 	struct extent_state *cached_state = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	struct falloc_range *range;
 	struct falloc_range *tmp;
 	struct list_head reserve_list;
@@ -2913,8 +2923,8 @@ static long btrfs_fallocate(struct file *file, int mode,
 				free_extent_map(em);
 				break;
 			}
-			ret = btrfs_qgroup_reserve_data(inode, cur_offset,
-					last_byte - cur_offset);
+			ret = btrfs_qgroup_reserve_data(inode, &data_reserved,
+					cur_offset, last_byte - cur_offset);
 			if (ret < 0) {
 				free_extent_map(em);
 				break;
@@ -2925,8 +2935,8 @@ static long btrfs_fallocate(struct file *file, int mode,
 			 * range, free reserved data space first, otherwise
 			 * it'll result in false ENOSPC error.
 			 */
-			btrfs_free_reserved_data_space(inode, cur_offset,
-				last_byte - cur_offset);
+			btrfs_free_reserved_data_space(inode, data_reserved,
+					cur_offset, last_byte - cur_offset);
 		}
 		free_extent_map(em);
 		cur_offset = last_byte;
@@ -2945,8 +2955,9 @@ static long btrfs_fallocate(struct file *file, int mode,
 					range->len, i_blocksize(inode),
 					offset + len, &alloc_hint);
 		else
-			btrfs_free_reserved_data_space(inode, range->start,
-						       range->len);
+			btrfs_free_reserved_data_space(inode,
+					data_reserved, range->start,
+					range->len);
 		list_del(&range->list);
 		kfree(range);
 	}
@@ -2984,8 +2995,9 @@ out:
 	inode_unlock(inode);
 	/* Let go of our reservation. */
 	if (ret != 0)
-		btrfs_free_reserved_data_space(inode, alloc_start,
-				       alloc_end - cur_offset);
+		btrfs_free_reserved_data_space(inode, data_reserved,
+				alloc_start, alloc_end - cur_offset);
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 

fs/btrfs/free-space-tree.c

@@ -17,7 +17,7 @@
  */
 
 #include <linux/kernel.h>
-#include <linux/vmalloc.h>
+#include <linux/sched/mm.h>
 #include "ctree.h"
 #include "disk-io.h"
 #include "locking.h"
@@ -153,21 +153,21 @@ static inline u32 free_space_bitmap_size(u64 size, u32 sectorsize)
 
 static u8 *alloc_bitmap(u32 bitmap_size)
 {
-	void *mem;
+	u8 *ret;
+	unsigned int nofs_flag;
 
 	/*
-	 * The allocation size varies, observed numbers were < 4K up to 16K.
-	 * Using vmalloc unconditionally would be too heavy, we'll try
-	 * contiguous allocations first.
+	 * GFP_NOFS doesn't work with kvmalloc(), but we really can't recurse
+	 * into the filesystem as the free space bitmap can be modified in the
+	 * critical section of a transaction commit.
+	 *
+	 * TODO: push the memalloc_nofs_{save,restore}() to the caller where we
+	 * know that recursion is unsafe.
 	 */
-	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_ZERO, PAGE_KERNEL);
+	nofs_flag = memalloc_nofs_save();
+	ret = kvzalloc(bitmap_size, GFP_KERNEL);
+	memalloc_nofs_restore(nofs_flag);
+	return ret;
 }
 
 int convert_free_space_to_bitmaps(struct btrfs_trans_handle *trans,
@@ -1188,11 +1188,7 @@ int btrfs_create_free_space_tree(struct btrfs_fs_info *fs_info)
 	btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID);
 	clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
 
-	ret = btrfs_commit_transaction(trans);
-	if (ret)
-		return ret;
-
-	return 0;
+	return btrfs_commit_transaction(trans);
 
 abort:
 	clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
@@ -1277,11 +1273,7 @@ int btrfs_clear_free_space_tree(struct btrfs_fs_info *fs_info)
 	free_extent_buffer(free_space_root->commit_root);
 	kfree(free_space_root);
 
-	ret = btrfs_commit_transaction(trans);
-	if (ret)
-		return ret;
-
-	return 0;
+	return btrfs_commit_transaction(trans);
 
 abort:
 	btrfs_abort_transaction(trans, ret);

fs/btrfs/inode-map.c

@@ -400,6 +400,7 @@ int btrfs_save_ino_cache(struct btrfs_root *root,
 	struct btrfs_path *path;
 	struct inode *inode;
 	struct btrfs_block_rsv *rsv;
+	struct extent_changeset *data_reserved = NULL;
 	u64 num_bytes;
 	u64 alloc_hint = 0;
 	int ret;
@@ -492,7 +493,7 @@ again:
 	/* Just to make sure we have enough space */
 	prealloc += 8 * PAGE_SIZE;
 
-	ret = btrfs_delalloc_reserve_space(inode, 0, prealloc);
+	ret = btrfs_delalloc_reserve_space(inode, &data_reserved, 0, prealloc);
 	if (ret)
 		goto out_put;
 
@@ -516,6 +517,7 @@ out:
 	trans->bytes_reserved = num_bytes;
 
 	btrfs_free_path(path);
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 

fs/btrfs/inode.c

@@ -86,7 +86,6 @@ static const struct extent_io_ops btrfs_extent_io_ops;
 
 static struct kmem_cache *btrfs_inode_cachep;
 struct kmem_cache *btrfs_trans_handle_cachep;
-struct kmem_cache *btrfs_transaction_cachep;
 struct kmem_cache *btrfs_path_cachep;
 struct kmem_cache *btrfs_free_space_cachep;
 
@@ -178,7 +177,6 @@ static int insert_inline_extent(struct btrfs_trans_handle *trans,
 	char *kaddr;
 	unsigned long ptr;
 	struct btrfs_file_extent_item *ei;
-	int err = 0;
 	int ret;
 	size_t cur_size = size;
 	unsigned long offset;
@@ -200,10 +198,8 @@ static int insert_inline_extent(struct btrfs_trans_handle *trans,
 		path->leave_spinning = 1;
 		ret = btrfs_insert_empty_item(trans, root, path, &key,
 					      datasize);
-		if (ret) {
-			err = ret;
+		if (ret)
 			goto fail;
-		}
 	}
 	leaf = path->nodes[0];
 	ei = btrfs_item_ptr(leaf, path->slots[0],
@@ -258,9 +254,8 @@ static int insert_inline_extent(struct btrfs_trans_handle *trans,
 	BTRFS_I(inode)->disk_i_size = inode->i_size;
 	ret = btrfs_update_inode(trans, root, inode);
 
-	return ret;
 fail:
-	return err;
+	return ret;
 }
 
 
@@ -350,7 +345,7 @@ out:
 	 * And at reserve time, it's always aligned to page size, so
 	 * just free one page here.
 	 */
-	btrfs_qgroup_free_data(inode, 0, PAGE_SIZE);
+	btrfs_qgroup_free_data(inode, NULL, 0, PAGE_SIZE);
 	btrfs_free_path(path);
 	btrfs_end_transaction(trans);
 	return ret;
@@ -608,12 +603,11 @@ cont:
 
 		/*
 		 * one last check to make sure the compression is really a
-		 * win, compare the page count read with the blocks on disk
+		 * win, compare the page count read with the blocks on disk,
+		 * compression must free at least one sector size
 		 */
 		total_in = ALIGN(total_in, PAGE_SIZE);
-		if (total_compressed >= total_in) {
-			will_compress = 0;
-		} else {
+		if (total_compressed + blocksize <= total_in) {
 			num_bytes = total_in;
 			*num_added += 1;
 
@@ -1568,10 +1562,11 @@ static inline int need_force_cow(struct inode *inode, u64 start, u64 end)
 /*
  * extent_io.c call back to do delayed allocation processing
  */
-static int run_delalloc_range(struct inode *inode, struct page *locked_page,
+static int run_delalloc_range(void *private_data, struct page *locked_page,
 			      u64 start, u64 end, int *page_started,
 			      unsigned long *nr_written)
 {
+	struct inode *inode = private_data;
 	int ret;
 	int force_cow = need_force_cow(inode, start, end);
 
@@ -1595,9 +1590,10 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
 	return ret;
 }
 
-static void btrfs_split_extent_hook(struct inode *inode,
+static void btrfs_split_extent_hook(void *private_data,
 				    struct extent_state *orig, u64 split)
 {
+	struct inode *inode = private_data;
 	u64 size;
 
 	/* not delalloc, ignore it */
@@ -1632,10 +1628,11 @@ static void btrfs_split_extent_hook(struct inode *inode,
  * extents, such as when we are doing sequential writes, so we can properly
  * account for the metadata space we'll need.
  */
-static void btrfs_merge_extent_hook(struct inode *inode,
+static void btrfs_merge_extent_hook(void *private_data,
 				    struct extent_state *new,
 				    struct extent_state *other)
 {
+	struct inode *inode = private_data;
 	u64 new_size, old_size;
 	u32 num_extents;
 
@@ -1735,9 +1732,10 @@ static void btrfs_del_delalloc_inode(struct btrfs_root *root,
  * bytes in this file, and to maintain the list of inodes that
  * have pending delalloc work to be done.
  */
-static void btrfs_set_bit_hook(struct inode *inode,
+static void btrfs_set_bit_hook(void *private_data,
 			       struct extent_state *state, unsigned *bits)
 {
+	struct inode *inode = private_data;
 
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 
@@ -1789,10 +1787,11 @@ static void btrfs_set_bit_hook(struct inode *inode,
 /*
  * extent_io.c clear_bit_hook, see set_bit_hook for why
  */
-static void btrfs_clear_bit_hook(struct btrfs_inode *inode,
+static void btrfs_clear_bit_hook(void *private_data,
 				 struct extent_state *state,
 				 unsigned *bits)
 {
+	struct btrfs_inode *inode = BTRFS_I((struct inode *)private_data);
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->vfs_inode.i_sb);
 	u64 len = state->end + 1 - state->start;
 	u32 num_extents = count_max_extents(len);
@@ -1900,10 +1899,11 @@ int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
  * At IO completion time the cums attached on the ordered extent record
  * are inserted into the btree
  */
-static blk_status_t __btrfs_submit_bio_start(struct inode *inode,
-		struct bio *bio, int mirror_num, unsigned long bio_flags,
-		u64 bio_offset)
+static blk_status_t __btrfs_submit_bio_start(void *private_data, struct bio *bio,
+				    int mirror_num, unsigned long bio_flags,
+				    u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	blk_status_t ret = 0;
 
 	ret = btrfs_csum_one_bio(inode, bio, 0, 0);
@@ -1919,10 +1919,11 @@ static blk_status_t __btrfs_submit_bio_start(struct inode *inode,
  * At IO completion time the cums attached on the ordered extent record
  * are inserted into the btree
  */
-static blk_status_t __btrfs_submit_bio_done(struct inode *inode,
-		struct bio *bio, int mirror_num, unsigned long bio_flags,
-		u64 bio_offset)
+static blk_status_t __btrfs_submit_bio_done(void *private_data, struct bio *bio,
+			  int mirror_num, unsigned long bio_flags,
+			  u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	blk_status_t ret;
 
@@ -1938,10 +1939,11 @@ static blk_status_t __btrfs_submit_bio_done(struct inode *inode,
  * extent_io.c submission hook. This does the right thing for csum calculation
  * on write, or reading the csums from the tree before a read
  */
-static blk_status_t btrfs_submit_bio_hook(struct inode *inode, struct bio *bio,
-			  int mirror_num, unsigned long bio_flags,
-			  u64 bio_offset)
+static blk_status_t btrfs_submit_bio_hook(void *private_data, struct bio *bio,
+				 int mirror_num, unsigned long bio_flags,
+				 u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	enum btrfs_wq_endio_type metadata = BTRFS_WQ_ENDIO_DATA;
@@ -1975,8 +1977,8 @@ static blk_status_t btrfs_submit_bio_hook(struct inode *inode, struct bio *bio,
 		if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
 			goto mapit;
 		/* we're doing a write, do the async checksumming */
-		ret = btrfs_wq_submit_bio(fs_info, inode, bio, mirror_num,
-					  bio_flags, bio_offset,
+		ret = btrfs_wq_submit_bio(fs_info, bio, mirror_num, bio_flags,
+					  bio_offset, inode,
 					  __btrfs_submit_bio_start,
 					  __btrfs_submit_bio_done);
 		goto out;
@@ -2034,6 +2036,7 @@ static void btrfs_writepage_fixup_worker(struct btrfs_work *work)
 	struct btrfs_writepage_fixup *fixup;
 	struct btrfs_ordered_extent *ordered;
 	struct extent_state *cached_state = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	struct page *page;
 	struct inode *inode;
 	u64 page_start;
@@ -2071,7 +2074,7 @@ again:
 		goto again;
 	}
 
-	ret = btrfs_delalloc_reserve_space(inode, page_start,
+	ret = btrfs_delalloc_reserve_space(inode, &data_reserved, page_start,
 					   PAGE_SIZE);
 	if (ret) {
 		mapping_set_error(page->mapping, ret);
@@ -2091,6 +2094,7 @@ out_page:
 	unlock_page(page);
 	put_page(page);
 	kfree(fixup);
+	extent_changeset_free(data_reserved);
 }
 
 /*
@@ -2142,6 +2146,7 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	struct btrfs_path *path;
 	struct extent_buffer *leaf;
 	struct btrfs_key ins;
+	u64 qg_released;
 	int extent_inserted = 0;
 	int ret;
 
@@ -2197,13 +2202,17 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	ins.objectid = disk_bytenr;
 	ins.offset = disk_num_bytes;
 	ins.type = BTRFS_EXTENT_ITEM_KEY;
-	ret = btrfs_alloc_reserved_file_extent(trans, root->root_key.objectid,
-			btrfs_ino(BTRFS_I(inode)), file_pos, ram_bytes, &ins);
+
 	/*
 	 * Release the reserved range from inode dirty range map, as it is
 	 * already moved into delayed_ref_head
 	 */
-	btrfs_qgroup_release_data(inode, file_pos, ram_bytes);
+	ret = btrfs_qgroup_release_data(inode, file_pos, ram_bytes);
+	if (ret < 0)
+		goto out;
+	qg_released = ret;
+	ret = btrfs_alloc_reserved_file_extent(trans, root->root_key.objectid,
+			btrfs_ino(BTRFS_I(inode)), file_pos, qg_released, &ins);
 out:
 	btrfs_free_path(path);
 
@@ -2925,7 +2934,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		 * space for NOCOW range.
 		 * As NOCOW won't cause a new delayed ref, just free the space
 		 */
-		btrfs_qgroup_free_data(inode, ordered_extent->file_offset,
+		btrfs_qgroup_free_data(inode, NULL, ordered_extent->file_offset,
 				       ordered_extent->len);
 		btrfs_ordered_update_i_size(inode, 0, ordered_extent);
 		if (nolock)
@@ -4761,6 +4770,7 @@ int btrfs_truncate_block(struct inode *inode, loff_t from, loff_t len,
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	struct btrfs_ordered_extent *ordered;
 	struct extent_state *cached_state = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	char *kaddr;
 	u32 blocksize = fs_info->sectorsize;
 	pgoff_t index = from >> PAGE_SHIFT;
@@ -4775,7 +4785,7 @@ int btrfs_truncate_block(struct inode *inode, loff_t from, loff_t len,
 	    (!len || ((len & (blocksize - 1)) == 0)))
 		goto out;
 
-	ret = btrfs_delalloc_reserve_space(inode,
+	ret = btrfs_delalloc_reserve_space(inode, &data_reserved,
 			round_down(from, blocksize), blocksize);
 	if (ret)
 		goto out;
@@ -4783,7 +4793,7 @@ int btrfs_truncate_block(struct inode *inode, loff_t from, loff_t len,
 again:
 	page = find_or_create_page(mapping, index, mask);
 	if (!page) {
-		btrfs_delalloc_release_space(inode,
+		btrfs_delalloc_release_space(inode, data_reserved,
 				round_down(from, blocksize),
 				blocksize);
 		ret = -ENOMEM;
@@ -4855,11 +4865,12 @@ again:
 
 out_unlock:
 	if (ret)
-		btrfs_delalloc_release_space(inode, block_start,
+		btrfs_delalloc_release_space(inode, data_reserved, block_start,
 					     blocksize);
 	unlock_page(page);
 	put_page(page);
 out:
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 
@@ -5254,7 +5265,7 @@ static void evict_inode_truncate_pages(struct inode *inode)
 		 * Note, end is the bytenr of last byte, so we need + 1 here.
 		 */
 		if (state->state & EXTENT_DELALLOC)
-			btrfs_qgroup_free_data(inode, start, end - start + 1);
+			btrfs_qgroup_free_data(inode, NULL, start, end - start + 1);
 
 		clear_extent_bit(io_tree, start, end,
 				 EXTENT_LOCKED | EXTENT_DIRTY |
@@ -5867,7 +5878,6 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 	struct inode *inode = file_inode(file);
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct btrfs_item *item;
 	struct btrfs_dir_item *di;
 	struct btrfs_key key;
 	struct btrfs_key found_key;
@@ -5918,7 +5928,6 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 			continue;
 		}
 
-		item = btrfs_item_nr(slot);
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
 
 		if (found_key.objectid != key.objectid)
@@ -5933,7 +5942,7 @@ static int btrfs_real_readdir(struct file *file, struct dir_context *ctx)
 		ctx->pos = found_key.offset;
 
 		di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
-		if (verify_dir_item(fs_info, leaf, di))
+		if (verify_dir_item(fs_info, leaf, slot, di))
 			goto next;
 
 		name_len = btrfs_dir_name_len(leaf, di);
@@ -7479,7 +7488,7 @@ out:
 bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end)
 {
 	struct radix_tree_root *root = &inode->i_mapping->page_tree;
-	int found = false;
+	bool found = false;
 	void **pagep = NULL;
 	struct page *page = NULL;
 	unsigned long start_idx;
@@ -7977,9 +7986,12 @@ static int dio_read_error(struct inode *inode, struct bio *failed_bio,
 			bio_end_io_t *repair_endio, void *repair_arg)
 {
 	struct io_failure_record *failrec;
+	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	struct extent_io_tree *failure_tree = &BTRFS_I(inode)->io_failure_tree;
 	struct bio *bio;
 	int isector;
 	int read_mode = 0;
+	int segs;
 	int ret;
 
 	BUG_ON(bio_op(failed_bio) == REQ_OP_WRITE);
@@ -7991,13 +8003,13 @@ static int dio_read_error(struct inode *inode, struct bio *failed_bio,
 	ret = btrfs_check_dio_repairable(inode, failed_bio, failrec,
 					 failed_mirror);
 	if (!ret) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, io_tree, failrec);
 		return -EIO;
 	}
 
-	if ((failed_bio->bi_vcnt > 1)
-		|| (failed_bio->bi_io_vec->bv_len
-			> btrfs_inode_sectorsize(inode)))
+	segs = bio_segments(failed_bio);
+	if (segs > 1 ||
+	    (failed_bio->bi_io_vec->bv_len > btrfs_inode_sectorsize(inode)))
 		read_mode |= REQ_FAILFAST_DEV;
 
 	isector = start - btrfs_io_bio(failed_bio)->logical;
@@ -8005,7 +8017,7 @@ static int dio_read_error(struct inode *inode, struct bio *failed_bio,
 	bio = btrfs_create_repair_bio(inode, failed_bio, failrec, page,
 				pgoff, isector, repair_endio, repair_arg);
 	if (!bio) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, io_tree, failrec);
 		return -EIO;
 	}
 	bio_set_op_attrs(bio, REQ_OP_READ, read_mode);
@@ -8016,7 +8028,7 @@ static int dio_read_error(struct inode *inode, struct bio *failed_bio,
 
 	ret = submit_dio_repair_bio(inode, bio, failrec->this_mirror);
 	if (ret) {
-		free_io_failure(BTRFS_I(inode), failrec);
+		free_io_failure(failure_tree, io_tree, failrec);
 		bio_put(bio);
 	}
 
@@ -8033,19 +8045,24 @@ struct btrfs_retry_complete {
 static void btrfs_retry_endio_nocsum(struct bio *bio)
 {
 	struct btrfs_retry_complete *done = bio->bi_private;
+	struct inode *inode = done->inode;
 	struct bio_vec *bvec;
+	struct extent_io_tree *io_tree, *failure_tree;
 	int i;
 
 	if (bio->bi_status)
 		goto end;
 
 	ASSERT(bio->bi_vcnt == 1);
-	ASSERT(bio->bi_io_vec->bv_len == btrfs_inode_sectorsize(done->inode));
+	io_tree = &BTRFS_I(inode)->io_tree;
+	failure_tree = &BTRFS_I(inode)->io_failure_tree;
+	ASSERT(bio->bi_io_vec->bv_len == btrfs_inode_sectorsize(inode));
 
 	done->uptodate = 1;
 	bio_for_each_segment_all(bvec, bio, i)
-		clean_io_failure(BTRFS_I(done->inode), done->start,
-				 bvec->bv_page, 0);
+		clean_io_failure(BTRFS_I(inode)->root->fs_info, failure_tree,
+				 io_tree, done->start, bvec->bv_page,
+				 btrfs_ino(BTRFS_I(inode)), 0);
 end:
 	complete(&done->done);
 	bio_put(bio);
@@ -8055,36 +8072,40 @@ static int __btrfs_correct_data_nocsum(struct inode *inode,
 				       struct btrfs_io_bio *io_bio)
 {
 	struct btrfs_fs_info *fs_info;
-	struct bio_vec *bvec;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
 	struct btrfs_retry_complete done;
 	u64 start;
 	unsigned int pgoff;
 	u32 sectorsize;
 	int nr_sectors;
-	int i;
 	int ret;
+	int err = 0;
 
 	fs_info = BTRFS_I(inode)->root->fs_info;
 	sectorsize = fs_info->sectorsize;
 
 	start = io_bio->logical;
 	done.inode = inode;
+	io_bio->bio.bi_iter = io_bio->iter;
 
-	bio_for_each_segment_all(bvec, &io_bio->bio, i) {
-		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info, bvec->bv_len);
-		pgoff = bvec->bv_offset;
+	bio_for_each_segment(bvec, &io_bio->bio, iter) {
+		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info, bvec.bv_len);
+		pgoff = bvec.bv_offset;
 
 next_block_or_try_again:
 		done.uptodate = 0;
 		done.start = start;
 		init_completion(&done.done);
 
-		ret = dio_read_error(inode, &io_bio->bio, bvec->bv_page,
+		ret = dio_read_error(inode, &io_bio->bio, bvec.bv_page,
 				pgoff, start, start + sectorsize - 1,
 				io_bio->mirror_num,
 				btrfs_retry_endio_nocsum, &done);
-		if (ret)
-			return ret;
+		if (ret) {
+			err = ret;
+			goto next;
+		}
 
 		wait_for_completion(&done.done);
 
@@ -8093,6 +8114,7 @@ next_block_or_try_again:
 			goto next_block_or_try_again;
 		}
 
+next:
 		start += sectorsize;
 
 		nr_sectors--;
@@ -8103,13 +8125,15 @@ next_block_or_try_again:
 		}
 	}
 
-	return 0;
+	return err;
 }
 
 static void btrfs_retry_endio(struct bio *bio)
 {
 	struct btrfs_retry_complete *done = bio->bi_private;
 	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
+	struct extent_io_tree *io_tree, *failure_tree;
+	struct inode *inode = done->inode;
 	struct bio_vec *bvec;
 	int uptodate;
 	int ret;
@@ -8123,13 +8147,19 @@ static void btrfs_retry_endio(struct bio *bio)
 	ASSERT(bio->bi_vcnt == 1);
 	ASSERT(bio->bi_io_vec->bv_len == btrfs_inode_sectorsize(done->inode));
 
+	io_tree = &BTRFS_I(inode)->io_tree;
+	failure_tree = &BTRFS_I(inode)->io_failure_tree;
+
 	bio_for_each_segment_all(bvec, bio, i) {
-		ret = __readpage_endio_check(done->inode, io_bio, i,
-					bvec->bv_page, bvec->bv_offset,
-					done->start, bvec->bv_len);
+		ret = __readpage_endio_check(inode, io_bio, i, bvec->bv_page,
+					     bvec->bv_offset, done->start,
+					     bvec->bv_len);
 		if (!ret)
-			clean_io_failure(BTRFS_I(done->inode), done->start,
-					bvec->bv_page, bvec->bv_offset);
+			clean_io_failure(BTRFS_I(inode)->root->fs_info,
+					 failure_tree, io_tree, done->start,
+					 bvec->bv_page,
+					 btrfs_ino(BTRFS_I(inode)),
+					 bvec->bv_offset);
 		else
 			uptodate = 0;
 	}
@@ -8144,7 +8174,8 @@ static blk_status_t __btrfs_subio_endio_read(struct inode *inode,
 		struct btrfs_io_bio *io_bio, blk_status_t err)
 {
 	struct btrfs_fs_info *fs_info;
-	struct bio_vec *bvec;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
 	struct btrfs_retry_complete done;
 	u64 start;
 	u64 offset = 0;
@@ -8152,7 +8183,7 @@ static blk_status_t __btrfs_subio_endio_read(struct inode *inode,
 	int nr_sectors;
 	unsigned int pgoff;
 	int csum_pos;
-	int i;
+	bool uptodate = (err == 0);
 	int ret;
 
 	fs_info = BTRFS_I(inode)->root->fs_info;
@@ -8161,24 +8192,26 @@ static blk_status_t __btrfs_subio_endio_read(struct inode *inode,
 	err = 0;
 	start = io_bio->logical;
 	done.inode = inode;
+	io_bio->bio.bi_iter = io_bio->iter;
 
-	bio_for_each_segment_all(bvec, &io_bio->bio, i) {
-		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info, bvec->bv_len);
+	bio_for_each_segment(bvec, &io_bio->bio, iter) {
+		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info, bvec.bv_len);
 
-		pgoff = bvec->bv_offset;
+		pgoff = bvec.bv_offset;
 next_block:
-		csum_pos = BTRFS_BYTES_TO_BLKS(fs_info, offset);
-		ret = __readpage_endio_check(inode, io_bio, csum_pos,
-					bvec->bv_page, pgoff, start,
-					sectorsize);
-		if (likely(!ret))
-			goto next;
+		if (uptodate) {
+			csum_pos = BTRFS_BYTES_TO_BLKS(fs_info, offset);
+			ret = __readpage_endio_check(inode, io_bio, csum_pos,
+					bvec.bv_page, pgoff, start, sectorsize);
+			if (likely(!ret))
+				goto next;
+		}
 try_again:
 		done.uptodate = 0;
 		done.start = start;
 		init_completion(&done.done);
 
-		ret = dio_read_error(inode, &io_bio->bio, bvec->bv_page,
+		ret = dio_read_error(inode, &io_bio->bio, bvec.bv_page,
 				pgoff, start, start + sectorsize - 1,
 				io_bio->mirror_num,
 				btrfs_retry_endio, &done);
@@ -8233,8 +8266,11 @@ static void btrfs_endio_direct_read(struct bio *bio)
 	struct btrfs_io_bio *io_bio = btrfs_io_bio(bio);
 	blk_status_t err = bio->bi_status;
 
-	if (dip->flags & BTRFS_DIO_ORIG_BIO_SUBMITTED)
+	if (dip->flags & BTRFS_DIO_ORIG_BIO_SUBMITTED) {
 		err = btrfs_subio_endio_read(inode, io_bio, err);
+		if (!err)
+			bio->bi_status = 0;
+	}
 
 	unlock_extent(&BTRFS_I(inode)->io_tree, dip->logical_offset,
 		      dip->logical_offset + dip->bytes - 1);
@@ -8307,10 +8343,11 @@ static void btrfs_endio_direct_write(struct bio *bio)
 	bio_put(bio);
 }
 
-static blk_status_t __btrfs_submit_bio_start_direct_io(struct inode *inode,
+static blk_status_t __btrfs_submit_bio_start_direct_io(void *private_data,
 				    struct bio *bio, int mirror_num,
 				    unsigned long bio_flags, u64 offset)
 {
+	struct inode *inode = private_data;
 	blk_status_t ret;
 	ret = btrfs_csum_one_bio(inode, bio, offset, 1);
 	BUG_ON(ret); /* -ENOMEM */
@@ -8357,16 +8394,6 @@ out:
 	bio_put(bio);
 }
 
-static struct bio *btrfs_dio_bio_alloc(struct block_device *bdev,
-				       u64 first_sector, gfp_t gfp_flags)
-{
-	struct bio *bio;
-	bio = btrfs_bio_alloc(bdev, first_sector, BIO_MAX_PAGES, gfp_flags);
-	if (bio)
-		bio_associate_current(bio);
-	return bio;
-}
-
 static inline blk_status_t btrfs_lookup_and_bind_dio_csum(struct inode *inode,
 						 struct btrfs_dio_private *dip,
 						 struct bio *bio,
@@ -8422,8 +8449,8 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
 		goto map;
 
 	if (write && async_submit) {
-		ret = btrfs_wq_submit_bio(fs_info, inode, bio, 0, 0,
-					  file_offset,
+		ret = btrfs_wq_submit_bio(fs_info, bio, 0, 0,
+					  file_offset, inode,
 					  __btrfs_submit_bio_start_direct_io,
 					  __btrfs_submit_bio_done);
 		goto err;
@@ -8453,103 +8480,83 @@ static int btrfs_submit_direct_hook(struct btrfs_dio_private *dip,
 {
 	struct inode *inode = dip->inode;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
-	struct btrfs_root *root = BTRFS_I(inode)->root;
 	struct bio *bio;
 	struct bio *orig_bio = dip->orig_bio;
-	struct bio_vec *bvec;
 	u64 start_sector = orig_bio->bi_iter.bi_sector;
 	u64 file_offset = dip->logical_offset;
-	u64 submit_len = 0;
 	u64 map_length;
-	u32 blocksize = fs_info->sectorsize;
 	int async_submit = 0;
-	int nr_sectors;
+	u64 submit_len;
+	int clone_offset = 0;
+	int clone_len;
 	int ret;
-	int i, j;
 
 	map_length = orig_bio->bi_iter.bi_size;
+	submit_len = map_length;
 	ret = btrfs_map_block(fs_info, btrfs_op(orig_bio), start_sector << 9,
 			      &map_length, NULL, 0);
 	if (ret)
 		return -EIO;
 
-	if (map_length >= orig_bio->bi_iter.bi_size) {
+	if (map_length >= submit_len) {
 		bio = orig_bio;
 		dip->flags |= BTRFS_DIO_ORIG_BIO_SUBMITTED;
 		goto submit;
 	}
 
 	/* async crcs make it difficult to collect full stripe writes. */
-	if (btrfs_get_alloc_profile(root, 1) & BTRFS_BLOCK_GROUP_RAID56_MASK)
+	if (btrfs_data_alloc_profile(fs_info) & BTRFS_BLOCK_GROUP_RAID56_MASK)
 		async_submit = 0;
 	else
 		async_submit = 1;
 
-	bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS);
-	if (!bio)
-		return -ENOMEM;
-
-	bio->bi_opf = orig_bio->bi_opf;
-	bio->bi_private = dip;
-	bio->bi_end_io = btrfs_end_dio_bio;
-	btrfs_io_bio(bio)->logical = file_offset;
+	/* bio split */
+	ASSERT(map_length <= INT_MAX);
 	atomic_inc(&dip->pending_bios);
+	do {
+		clone_len = min_t(int, submit_len, map_length);
 
-	bio_for_each_segment_all(bvec, orig_bio, j) {
-		nr_sectors = BTRFS_BYTES_TO_BLKS(fs_info, bvec->bv_len);
-		i = 0;
-next_block:
-		if (unlikely(map_length < submit_len + blocksize ||
-		    bio_add_page(bio, bvec->bv_page, blocksize,
-			    bvec->bv_offset + (i * blocksize)) < blocksize)) {
-			/*
-			 * inc the count before we submit the bio so
-			 * we know the end IO handler won't happen before
-			 * we inc the count. Otherwise, the dip might get freed
-			 * before we're done setting it up
-			 */
-			atomic_inc(&dip->pending_bios);
-			ret = __btrfs_submit_dio_bio(bio, inode,
-						     file_offset, skip_sum,
-						     async_submit);
-			if (ret) {
-				bio_put(bio);
-				atomic_dec(&dip->pending_bios);
-				goto out_err;
-			}
-
-			start_sector += submit_len >> 9;
-			file_offset += submit_len;
+		/*
+		 * This will never fail as it's passing GPF_NOFS and
+		 * the allocation is backed by btrfs_bioset.
+		 */
+		bio = btrfs_bio_clone_partial(orig_bio, clone_offset,
+					      clone_len);
+		bio->bi_private = dip;
+		bio->bi_end_io = btrfs_end_dio_bio;
+		btrfs_io_bio(bio)->logical = file_offset;
+
+		ASSERT(submit_len >= clone_len);
+		submit_len -= clone_len;
+		if (submit_len == 0)
+			break;
 
-			submit_len = 0;
+		/*
+		 * Increase the count before we submit the bio so we know
+		 * the end IO handler won't happen before we increase the
+		 * count. Otherwise, the dip might get freed before we're
+		 * done setting it up.
+		 */
+		atomic_inc(&dip->pending_bios);
 
-			bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev,
-						  start_sector, GFP_NOFS);
-			if (!bio)
-				goto out_err;
-			bio->bi_opf = orig_bio->bi_opf;
-			bio->bi_private = dip;
-			bio->bi_end_io = btrfs_end_dio_bio;
-			btrfs_io_bio(bio)->logical = file_offset;
+		ret = __btrfs_submit_dio_bio(bio, inode, file_offset, skip_sum,
+					     async_submit);
+		if (ret) {
+			bio_put(bio);
+			atomic_dec(&dip->pending_bios);
+			goto out_err;
+		}
 
-			map_length = orig_bio->bi_iter.bi_size;
-			ret = btrfs_map_block(fs_info, btrfs_op(orig_bio),
-					      start_sector << 9,
-					      &map_length, NULL, 0);
-			if (ret) {
-				bio_put(bio);
-				goto out_err;
-			}
+		clone_offset += clone_len;
+		start_sector += clone_len >> 9;
+		file_offset += clone_len;
 
-			goto next_block;
-		} else {
-			submit_len += blocksize;
-			if (--nr_sectors) {
-				i++;
-				goto next_block;
-			}
-		}
-	}
+		map_length = submit_len;
+		ret = btrfs_map_block(fs_info, btrfs_op(orig_bio),
+				      start_sector << 9, &map_length, NULL, 0);
+		if (ret)
+			goto out_err;
+	} while (submit_len > 0);
 
 submit:
 	ret = __btrfs_submit_dio_bio(bio, inode, file_offset, skip_sum,
@@ -8576,19 +8583,15 @@ static void btrfs_submit_direct(struct bio *dio_bio, struct inode *inode,
 				loff_t file_offset)
 {
 	struct btrfs_dio_private *dip = NULL;
-	struct bio *io_bio = NULL;
-	struct btrfs_io_bio *btrfs_bio;
+	struct bio *bio = NULL;
+	struct btrfs_io_bio *io_bio;
 	int skip_sum;
 	bool write = (bio_op(dio_bio) == REQ_OP_WRITE);
 	int ret = 0;
 
 	skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
-	io_bio = btrfs_bio_clone(dio_bio, GFP_NOFS);
-	if (!io_bio) {
-		ret = -ENOMEM;
-		goto free_ordered;
-	}
+	bio = btrfs_bio_clone(dio_bio);
 
 	dip = kzalloc(sizeof(*dip), GFP_NOFS);
 	if (!dip) {
@@ -8601,17 +8604,17 @@ static void btrfs_submit_direct(struct bio *dio_bio, struct inode *inode,
 	dip->logical_offset = file_offset;
 	dip->bytes = dio_bio->bi_iter.bi_size;
 	dip->disk_bytenr = (u64)dio_bio->bi_iter.bi_sector << 9;
-	io_bio->bi_private = dip;
-	dip->orig_bio = io_bio;
+	bio->bi_private = dip;
+	dip->orig_bio = bio;
 	dip->dio_bio = dio_bio;
 	atomic_set(&dip->pending_bios, 0);
-	btrfs_bio = btrfs_io_bio(io_bio);
-	btrfs_bio->logical = file_offset;
+	io_bio = btrfs_io_bio(bio);
+	io_bio->logical = file_offset;
 
 	if (write) {
-		io_bio->bi_end_io = btrfs_endio_direct_write;
+		bio->bi_end_io = btrfs_endio_direct_write;
 	} else {
-		io_bio->bi_end_io = btrfs_endio_direct_read;
+		bio->bi_end_io = btrfs_endio_direct_read;
 		dip->subio_endio = btrfs_subio_endio_read;
 	}
 
@@ -8634,8 +8637,8 @@ static void btrfs_submit_direct(struct bio *dio_bio, struct inode *inode,
 	if (!ret)
 		return;
 
-	if (btrfs_bio->end_io)
-		btrfs_bio->end_io(btrfs_bio, ret);
+	if (io_bio->end_io)
+		io_bio->end_io(io_bio, ret);
 
 free_ordered:
 	/*
@@ -8647,16 +8650,15 @@ free_ordered:
 	 * same as btrfs_endio_direct_[write|read] because we can't call these
 	 * callbacks - they require an allocated dip and a clone of dio_bio.
 	 */
-	if (io_bio && dip) {
-		io_bio->bi_status = BLK_STS_IOERR;
-		bio_endio(io_bio);
+	if (bio && dip) {
+		bio_io_error(bio);
 		/*
-		 * The end io callbacks free our dip, do the final put on io_bio
+		 * The end io callbacks free our dip, do the final put on bio
 		 * and all the cleanup and final put for dio_bio (through
 		 * dio_end_io()).
 		 */
 		dip = NULL;
-		io_bio = NULL;
+		bio = NULL;
 	} else {
 		if (write)
 			__endio_write_update_ordered(inode,
@@ -8674,8 +8676,8 @@ free_ordered:
 		 */
 		dio_end_io(dio_bio);
 	}
-	if (io_bio)
-		bio_put(io_bio);
+	if (bio)
+		bio_put(bio);
 	kfree(dip);
 }
 
@@ -8719,6 +8721,7 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct inode *inode = file->f_mapping->host;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct btrfs_dio_data dio_data = { 0 };
+	struct extent_changeset *data_reserved = NULL;
 	loff_t offset = iocb->ki_pos;
 	size_t count = 0;
 	int flags = 0;
@@ -8758,7 +8761,8 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 			ret = -EAGAIN;
 			goto out;
 		}
-		ret = btrfs_delalloc_reserve_space(inode, offset, count);
+		ret = btrfs_delalloc_reserve_space(inode, &data_reserved,
+						   offset, count);
 		if (ret)
 			goto out;
 		dio_data.outstanding_extents = count_max_extents(count);
@@ -8790,8 +8794,8 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 		current->journal_info = NULL;
 		if (ret < 0 && ret != -EIOCBQUEUED) {
 			if (dio_data.reserve)
-				btrfs_delalloc_release_space(inode, offset,
-							     dio_data.reserve);
+				btrfs_delalloc_release_space(inode, data_reserved,
+					offset, dio_data.reserve);
 			/*
 			 * On error we might have left some ordered extents
 			 * without submitting corresponding bios for them, so
@@ -8806,8 +8810,8 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 					dio_data.unsubmitted_oe_range_start,
 					false);
 		} else if (ret >= 0 && (size_t)ret < count)
-			btrfs_delalloc_release_space(inode, offset,
-						     count - (size_t)ret);
+			btrfs_delalloc_release_space(inode, data_reserved,
+					offset, count - (size_t)ret);
 	}
 out:
 	if (wakeup)
@@ -8815,6 +8819,7 @@ out:
 	if (relock)
 		inode_lock(inode);
 
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 
@@ -9005,7 +9010,7 @@ again:
 	 *    free the entire extent.
 	 */
 	if (PageDirty(page))
-		btrfs_qgroup_free_data(inode, page_start, PAGE_SIZE);
+		btrfs_qgroup_free_data(inode, NULL, page_start, PAGE_SIZE);
 	if (!inode_evicting) {
 		clear_extent_bit(tree, page_start, page_end,
 				 EXTENT_LOCKED | EXTENT_DIRTY |
@@ -9047,6 +9052,7 @@ int btrfs_page_mkwrite(struct vm_fault *vmf)
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
 	struct btrfs_ordered_extent *ordered;
 	struct extent_state *cached_state = NULL;
+	struct extent_changeset *data_reserved = NULL;
 	char *kaddr;
 	unsigned long zero_start;
 	loff_t size;
@@ -9072,7 +9078,7 @@ int btrfs_page_mkwrite(struct vm_fault *vmf)
 	 * end up waiting indefinitely to get a lock on the page currently
 	 * being processed by btrfs_page_mkwrite() function.
 	 */
-	ret = btrfs_delalloc_reserve_space(inode, page_start,
+	ret = btrfs_delalloc_reserve_space(inode, &data_reserved, page_start,
 					   reserved_space);
 	if (!ret) {
 		ret = file_update_time(vmf->vma->vm_file);
@@ -9126,8 +9132,8 @@ again:
 			spin_lock(&BTRFS_I(inode)->lock);
 			BTRFS_I(inode)->outstanding_extents++;
 			spin_unlock(&BTRFS_I(inode)->lock);
-			btrfs_delalloc_release_space(inode, page_start,
-						PAGE_SIZE - reserved_space);
+			btrfs_delalloc_release_space(inode, data_reserved,
+					page_start, PAGE_SIZE - reserved_space);
 		}
 	}
 
@@ -9178,13 +9184,16 @@ again:
 out_unlock:
 	if (!ret) {
 		sb_end_pagefault(inode->i_sb);
+		extent_changeset_free(data_reserved);
 		return VM_FAULT_LOCKED;
 	}
 	unlock_page(page);
 out:
-	btrfs_delalloc_release_space(inode, page_start, reserved_space);
+	btrfs_delalloc_release_space(inode, data_reserved, page_start,
+				     reserved_space);
 out_noreserve:
 	sb_end_pagefault(inode->i_sb);
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 
@@ -9406,8 +9415,8 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 
 	inode = &ei->vfs_inode;
 	extent_map_tree_init(&ei->extent_tree);
-	extent_io_tree_init(&ei->io_tree, &inode->i_data);
-	extent_io_tree_init(&ei->io_failure_tree, &inode->i_data);
+	extent_io_tree_init(&ei->io_tree, inode);
+	extent_io_tree_init(&ei->io_failure_tree, inode);
 	ei->io_tree.track_uptodate = 1;
 	ei->io_failure_tree.track_uptodate = 1;
 	atomic_set(&ei->sync_writers, 0);
@@ -9516,7 +9525,6 @@ void btrfs_destroy_cachep(void)
 	rcu_barrier();
 	kmem_cache_destroy(btrfs_inode_cachep);
 	kmem_cache_destroy(btrfs_trans_handle_cachep);
-	kmem_cache_destroy(btrfs_transaction_cachep);
 	kmem_cache_destroy(btrfs_path_cachep);
 	kmem_cache_destroy(btrfs_free_space_cachep);
 }
@@ -9536,12 +9544,6 @@ int btrfs_init_cachep(void)
 	if (!btrfs_trans_handle_cachep)
 		goto fail;
 
-	btrfs_transaction_cachep = kmem_cache_create("btrfs_transaction",
-			sizeof(struct btrfs_transaction), 0,
-			SLAB_TEMPORARY | SLAB_MEM_SPREAD, NULL);
-	if (!btrfs_transaction_cachep)
-		goto fail;
-
 	btrfs_path_cachep = kmem_cache_create("btrfs_path",
 			sizeof(struct btrfs_path), 0,
 			SLAB_MEM_SPREAD, NULL);
@@ -9566,6 +9568,24 @@ static int btrfs_getattr(const struct path *path, struct kstat *stat,
 	u64 delalloc_bytes;
 	struct inode *inode = d_inode(path->dentry);
 	u32 blocksize = inode->i_sb->s_blocksize;
+	u32 bi_flags = BTRFS_I(inode)->flags;
+
+	stat->result_mask |= STATX_BTIME;
+	stat->btime.tv_sec = BTRFS_I(inode)->i_otime.tv_sec;
+	stat->btime.tv_nsec = BTRFS_I(inode)->i_otime.tv_nsec;
+	if (bi_flags & BTRFS_INODE_APPEND)
+		stat->attributes |= STATX_ATTR_APPEND;
+	if (bi_flags & BTRFS_INODE_COMPRESS)
+		stat->attributes |= STATX_ATTR_COMPRESSED;
+	if (bi_flags & BTRFS_INODE_IMMUTABLE)
+		stat->attributes |= STATX_ATTR_IMMUTABLE;
+	if (bi_flags & BTRFS_INODE_NODUMP)
+		stat->attributes |= STATX_ATTR_NODUMP;
+
+	stat->attributes_mask |= (STATX_ATTR_APPEND |
+				  STATX_ATTR_COMPRESSED |
+				  STATX_ATTR_IMMUTABLE |
+				  STATX_ATTR_NODUMP);
 
 	generic_fillattr(inode, stat);
 	stat->dev = BTRFS_I(inode)->root->anon_dev;
@@ -10540,7 +10560,7 @@ next:
 			btrfs_end_transaction(trans);
 	}
 	if (cur_offset < end)
-		btrfs_free_reserved_data_space(inode, cur_offset,
+		btrfs_free_reserved_data_space(inode, NULL, cur_offset,
 			end - cur_offset + 1);
 	return ret;
 }
@@ -10661,6 +10681,42 @@ static int btrfs_readpage_io_failed_hook(struct page *page, int failed_mirror)
 	return -EAGAIN;
 }
 
+static struct btrfs_fs_info *iotree_fs_info(void *private_data)
+{
+	struct inode *inode = private_data;
+	return btrfs_sb(inode->i_sb);
+}
+
+static void btrfs_check_extent_io_range(void *private_data, const char *caller,
+					u64 start, u64 end)
+{
+	struct inode *inode = private_data;
+	u64 isize;
+
+	isize = i_size_read(inode);
+	if (end >= PAGE_SIZE && (end % 2) == 0 && end != isize - 1) {
+		btrfs_debug_rl(BTRFS_I(inode)->root->fs_info,
+		    "%s: ino %llu isize %llu odd range [%llu,%llu]",
+			caller, btrfs_ino(BTRFS_I(inode)), isize, start, end);
+	}
+}
+
+void btrfs_set_range_writeback(void *private_data, u64 start, u64 end)
+{
+	struct inode *inode = private_data;
+	unsigned long index = start >> PAGE_SHIFT;
+	unsigned long end_index = end >> PAGE_SHIFT;
+	struct page *page;
+
+	while (index <= end_index) {
+		page = find_get_page(inode->i_mapping, index);
+		ASSERT(page); /* Pages should be in the extent_io_tree */
+		set_page_writeback(page);
+		put_page(page);
+		index++;
+	}
+}
+
 static const struct inode_operations btrfs_dir_inode_operations = {
 	.getattr	= btrfs_getattr,
 	.lookup		= btrfs_lookup,
@@ -10704,6 +10760,8 @@ static const struct extent_io_ops btrfs_extent_io_ops = {
 	.readpage_end_io_hook = btrfs_readpage_end_io_hook,
 	.merge_bio_hook = btrfs_merge_bio_hook,
 	.readpage_io_failed_hook = btrfs_readpage_io_failed_hook,
+	.tree_fs_info = iotree_fs_info,
+	.set_range_writeback = btrfs_set_range_writeback,
 
 	/* optional callbacks */
 	.fill_delalloc = run_delalloc_range,
@@ -10713,6 +10771,7 @@ static const struct extent_io_ops btrfs_extent_io_ops = {
 	.clear_bit_hook = btrfs_clear_bit_hook,
 	.merge_extent_hook = btrfs_merge_extent_hook,
 	.split_extent_hook = btrfs_split_extent_hook,
+	.check_extent_io_range = btrfs_check_extent_io_range,
 };
 
 /*

fs/btrfs/ioctl.c

@@ -37,7 +37,7 @@
 #include <linux/bit_spinlock.h>
 #include <linux/security.h>
 #include <linux/xattr.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <linux/slab.h>
 #include <linux/blkdev.h>
 #include <linux/uuid.h>
@@ -689,7 +689,7 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir,
 	if (ret)
 		goto dec_and_free;
 
-	btrfs_wait_ordered_extents(root, -1, 0, (u64)-1);
+	btrfs_wait_ordered_extents(root, U64_MAX, 0, (u64)-1);
 
 	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
 			     BTRFS_BLOCK_RSV_TEMP);
@@ -1127,6 +1127,7 @@ static int cluster_pages_for_defrag(struct inode *inode,
 	struct btrfs_ordered_extent *ordered;
 	struct extent_state *cached_state = NULL;
 	struct extent_io_tree *tree;
+	struct extent_changeset *data_reserved = NULL;
 	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
 
 	file_end = (isize - 1) >> PAGE_SHIFT;
@@ -1135,7 +1136,7 @@ static int cluster_pages_for_defrag(struct inode *inode,
 
 	page_cnt = min_t(u64, (u64)num_pages, (u64)file_end - start_index + 1);
 
-	ret = btrfs_delalloc_reserve_space(inode,
+	ret = btrfs_delalloc_reserve_space(inode, &data_reserved,
 			start_index << PAGE_SHIFT,
 			page_cnt << PAGE_SHIFT);
 	if (ret)
@@ -1226,7 +1227,7 @@ again:
 		spin_lock(&BTRFS_I(inode)->lock);
 		BTRFS_I(inode)->outstanding_extents++;
 		spin_unlock(&BTRFS_I(inode)->lock);
-		btrfs_delalloc_release_space(inode,
+		btrfs_delalloc_release_space(inode, data_reserved,
 				start_index << PAGE_SHIFT,
 				(page_cnt - i_done) << PAGE_SHIFT);
 	}
@@ -1247,15 +1248,17 @@ again:
 		unlock_page(pages[i]);
 		put_page(pages[i]);
 	}
+	extent_changeset_free(data_reserved);
 	return i_done;
 out:
 	for (i = 0; i < i_done; i++) {
 		unlock_page(pages[i]);
 		put_page(pages[i]);
 	}
-	btrfs_delalloc_release_space(inode,
+	btrfs_delalloc_release_space(inode, data_reserved,
 			start_index << PAGE_SHIFT,
 			page_cnt << PAGE_SHIFT);
+	extent_changeset_free(data_reserved);
 	return ret;
 
 }
@@ -4588,7 +4591,7 @@ static long btrfs_ioctl_logical_to_ino(struct btrfs_fs_info *fs_info,
 
 out:
 	btrfs_free_path(path);
-	vfree(inodes);
+	kvfree(inodes);
 	kfree(loi);
 
 	return ret;
@@ -4897,7 +4900,6 @@ static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
 		goto out;
 	}
 
-	/* FIXME: check if the IDs really exist */
 	if (sa->assign) {
 		ret = btrfs_add_qgroup_relation(trans, fs_info,
 						sa->src, sa->dst);
@@ -4956,7 +4958,6 @@ static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
 		goto out;
 	}
 
-	/* FIXME: check if the IDs really exist */
 	if (sa->create) {
 		ret = btrfs_create_qgroup(trans, fs_info, sa->qgroupid);
 	} else {
@@ -5010,7 +5011,6 @@ static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
 		qgroupid = root->root_key.objectid;
 	}
 
-	/* FIXME: check if the IDs really exist */
 	ret = btrfs_limit_qgroup(trans, fs_info, qgroupid, &sa->lim);
 
 	err = btrfs_end_transaction(trans);

fs/btrfs/lzo.c

@@ -18,13 +18,14 @@
 
 #include <linux/kernel.h>
 #include <linux/slab.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <linux/init.h>
 #include <linux/err.h>
 #include <linux/sched.h>
 #include <linux/pagemap.h>
 #include <linux/bio.h>
 #include <linux/lzo.h>
+#include <linux/refcount.h>
 #include "compression.h"
 
 #define LZO_LEN	4
@@ -40,9 +41,9 @@ static void lzo_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
 
-	vfree(workspace->buf);
-	vfree(workspace->cbuf);
-	vfree(workspace->mem);
+	kvfree(workspace->buf);
+	kvfree(workspace->cbuf);
+	kvfree(workspace->mem);
 	kfree(workspace);
 }
 
@@ -50,13 +51,13 @@ static struct list_head *lzo_alloc_workspace(void)
 {
 	struct workspace *workspace;
 
-	workspace = kzalloc(sizeof(*workspace), GFP_NOFS);
+	workspace = kzalloc(sizeof(*workspace), GFP_KERNEL);
 	if (!workspace)
 		return ERR_PTR(-ENOMEM);
 
-	workspace->mem = vmalloc(LZO1X_MEM_COMPRESS);
-	workspace->buf = vmalloc(lzo1x_worst_compress(PAGE_SIZE));
-	workspace->cbuf = vmalloc(lzo1x_worst_compress(PAGE_SIZE));
+	workspace->mem = kvmalloc(LZO1X_MEM_COMPRESS, GFP_KERNEL);
+	workspace->buf = kvmalloc(lzo1x_worst_compress(PAGE_SIZE), GFP_KERNEL);
+	workspace->cbuf = kvmalloc(lzo1x_worst_compress(PAGE_SIZE), GFP_KERNEL);
 	if (!workspace->mem || !workspace->buf || !workspace->cbuf)
 		goto fail;
 
@@ -141,7 +142,7 @@ static int lzo_compress_pages(struct list_head *ws,
 		ret = lzo1x_1_compress(data_in, in_len, workspace->cbuf,
 				       &out_len, workspace->mem);
 		if (ret != LZO_E_OK) {
-			pr_debug("BTRFS: deflate in loop returned %d\n",
+			pr_debug("BTRFS: lzo in loop returned %d\n",
 			       ret);
 			ret = -EIO;
 			goto out;
@@ -229,8 +230,10 @@ static int lzo_compress_pages(struct list_head *ws,
 		in_len = min(bytes_left, PAGE_SIZE);
 	}
 
-	if (tot_out > tot_in)
+	if (tot_out >= tot_in) {
+		ret = -E2BIG;
 		goto out;
+	}
 
 	/* store the size of all chunks of compressed data */
 	cpage_out = kmap(pages[0]);
@@ -254,16 +257,13 @@ out:
 	return ret;
 }
 
-static int lzo_decompress_bio(struct list_head *ws,
-				 struct page **pages_in,
-				 u64 disk_start,
-				 struct bio *orig_bio,
-				 size_t srclen)
+static int lzo_decompress_bio(struct list_head *ws, struct compressed_bio *cb)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
 	int ret = 0, ret2;
 	char *data_in;
 	unsigned long page_in_index = 0;
+	size_t srclen = cb->compressed_len;
 	unsigned long total_pages_in = DIV_ROUND_UP(srclen, PAGE_SIZE);
 	unsigned long buf_start;
 	unsigned long buf_offset = 0;
@@ -278,6 +278,9 @@ static int lzo_decompress_bio(struct list_head *ws,
 	unsigned long tot_len;
 	char *buf;
 	bool may_late_unmap, need_unmap;
+	struct page **pages_in = cb->compressed_pages;
+	u64 disk_start = cb->start;
+	struct bio *orig_bio = cb->orig_bio;
 
 	data_in = kmap(pages_in[0]);
 	tot_len = read_compress_length(data_in);

fs/btrfs/ordered-data.c

@@ -663,7 +663,7 @@ static void btrfs_run_ordered_extent_work(struct btrfs_work *work)
  * wait for all the ordered extents in a root.  This is done when balancing
  * space between drives.
  */
-int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr,
+u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr,
 			       const u64 range_start, const u64 range_len)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -671,7 +671,7 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr,
 	LIST_HEAD(skipped);
 	LIST_HEAD(works);
 	struct btrfs_ordered_extent *ordered, *next;
-	int count = 0;
+	u64 count = 0;
 	const u64 range_end = range_start + range_len;
 
 	mutex_lock(&root->ordered_extent_mutex);
@@ -701,7 +701,7 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr,
 
 		cond_resched();
 		spin_lock(&root->ordered_extent_lock);
-		if (nr != -1)
+		if (nr != U64_MAX)
 			nr--;
 		count++;
 	}
@@ -720,13 +720,13 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr,
 	return count;
 }
 
-int btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr,
-			      const u64 range_start, const u64 range_len)
+u64 btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr,
+			     const u64 range_start, const u64 range_len)
 {
 	struct btrfs_root *root;
 	struct list_head splice;
-	int done;
-	int total_done = 0;
+	u64 total_done = 0;
+	u64 done;
 
 	INIT_LIST_HEAD(&splice);
 
@@ -748,9 +748,8 @@ int btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr,
 		total_done += done;
 
 		spin_lock(&fs_info->ordered_root_lock);
-		if (nr != -1) {
+		if (nr != U64_MAX) {
 			nr -= done;
-			WARN_ON(nr < 0);
 		}
 	}
 	list_splice_tail(&splice, &fs_info->ordered_roots);

fs/btrfs/ordered-data.h

@@ -200,9 +200,9 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset,
 				struct btrfs_ordered_extent *ordered);
 int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr,
 			   u32 *sum, int len);
-int btrfs_wait_ordered_extents(struct btrfs_root *root, int nr,
+u64 btrfs_wait_ordered_extents(struct btrfs_root *root, u64 nr,
 			       const u64 range_start, const u64 range_len);
-int btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, int nr,
+u64 btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr,
 			      const u64 range_start, const u64 range_len);
 void btrfs_get_logged_extents(struct btrfs_inode *inode,
 			      struct list_head *logged_list,

fs/btrfs/print-tree.c

@@ -261,8 +261,11 @@ void btrfs_print_leaf(struct btrfs_fs_info *fs_info, struct extent_buffer *l)
 		case BTRFS_BLOCK_GROUP_ITEM_KEY:
 			bi = btrfs_item_ptr(l, i,
 					    struct btrfs_block_group_item);
-			pr_info("\t\tblock group used %llu\n",
-			       btrfs_disk_block_group_used(l, bi));
+			pr_info(
+		   "\t\tblock group used %llu chunk_objectid %llu flags %llu\n",
+				btrfs_disk_block_group_used(l, bi),
+				btrfs_disk_block_group_chunk_objectid(l, bi),
+				btrfs_disk_block_group_flags(l, bi));
 			break;
 		case BTRFS_CHUNK_ITEM_KEY:
 			print_chunk(l, btrfs_item_ptr(l, i,

fs/btrfs/props.c

@@ -164,6 +164,7 @@ static int iterate_object_props(struct btrfs_root *root,
 						 size_t),
 				void *ctx)
 {
+	struct btrfs_fs_info *fs_info = root->fs_info;
 	int ret;
 	char *name_buf = NULL;
 	char *value_buf = NULL;
@@ -214,6 +215,12 @@ static int iterate_object_props(struct btrfs_root *root,
 			name_ptr = (unsigned long)(di + 1);
 			data_ptr = name_ptr + name_len;
 
+			if (verify_dir_item(fs_info, leaf,
+					    path->slots[0], di)) {
+				ret = -EIO;
+				goto out;
+			}
+
 			if (name_len <= XATTR_BTRFS_PREFIX_LEN ||
 			    memcmp_extent_buffer(leaf, XATTR_BTRFS_PREFIX,
 						 name_ptr,

fs/btrfs/qgroup.c

@@ -1406,38 +1406,6 @@ out:
 	return ret;
 }
 
-int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans,
-					 struct btrfs_fs_info *fs_info)
-{
-	struct btrfs_qgroup_extent_record *record;
-	struct btrfs_delayed_ref_root *delayed_refs;
-	struct rb_node *node;
-	u64 qgroup_to_skip;
-	int ret = 0;
-
-	delayed_refs = &trans->transaction->delayed_refs;
-	qgroup_to_skip = delayed_refs->qgroup_to_skip;
-
-	/*
-	 * No need to do lock, since this function will only be called in
-	 * btrfs_commit_transaction().
-	 */
-	node = rb_first(&delayed_refs->dirty_extent_root);
-	while (node) {
-		record = rb_entry(node, struct btrfs_qgroup_extent_record,
-				  node);
-		if (WARN_ON(!record->old_roots))
-			ret = btrfs_find_all_roots(NULL, fs_info,
-					record->bytenr, 0, &record->old_roots);
-		if (ret < 0)
-			break;
-		if (qgroup_to_skip)
-			ulist_del(record->old_roots, qgroup_to_skip, 0);
-		node = rb_next(node);
-	}
-	return ret;
-}
-
 int btrfs_qgroup_trace_extent_nolock(struct btrfs_fs_info *fs_info,
 				struct btrfs_delayed_ref_root *delayed_refs,
 				struct btrfs_qgroup_extent_record *record)
@@ -1559,6 +1527,7 @@ int btrfs_qgroup_trace_leaf_items(struct btrfs_trans_handle *trans,
 		if (ret)
 			return ret;
 	}
+	cond_resched();
 	return 0;
 }
 
@@ -1918,6 +1887,35 @@ static int qgroup_update_counters(struct btrfs_fs_info *fs_info,
 	return 0;
 }
 
+/*
+ * Check if the @roots potentially is a list of fs tree roots
+ *
+ * Return 0 for definitely not a fs/subvol tree roots ulist
+ * Return 1 for possible fs/subvol tree roots in the list (considering an empty
+ *          one as well)
+ */
+static int maybe_fs_roots(struct ulist *roots)
+{
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+
+	/* Empty one, still possible for fs roots */
+	if (!roots || roots->nnodes == 0)
+		return 1;
+
+	ULIST_ITER_INIT(&uiter);
+	unode = ulist_next(roots, &uiter);
+	if (!unode)
+		return 1;
+
+	/*
+	 * If it contains fs tree roots, then it must belong to fs/subvol
+	 * trees.
+	 * If it contains a non-fs tree, it won't be shared with fs/subvol trees.
+	 */
+	return is_fstree(unode->val);
+}
+
 int
 btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans,
 			    struct btrfs_fs_info *fs_info,
@@ -1934,10 +1932,20 @@ btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans,
 	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
 		return 0;
 
-	if (new_roots)
+	if (new_roots) {
+		if (!maybe_fs_roots(new_roots))
+			goto out_free;
 		nr_new_roots = new_roots->nnodes;
-	if (old_roots)
+	}
+	if (old_roots) {
+		if (!maybe_fs_roots(old_roots))
+			goto out_free;
 		nr_old_roots = old_roots->nnodes;
+	}
+
+	/* Quick exit, either not fs tree roots, or won't affect any qgroup */
+	if (nr_old_roots == 0 && nr_new_roots == 0)
+		goto out_free;
 
 	BUG_ON(!fs_info->quota_root);
 
@@ -2016,6 +2024,19 @@ int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans,
 		trace_btrfs_qgroup_account_extents(fs_info, record);
 
 		if (!ret) {
+			/*
+			 * Old roots should be searched when inserting qgroup
+			 * extent record
+			 */
+			if (WARN_ON(!record->old_roots)) {
+				/* Search commit root to find old_roots */
+				ret = btrfs_find_all_roots(NULL, fs_info,
+						record->bytenr, 0,
+						&record->old_roots);
+				if (ret < 0)
+					goto cleanup;
+			}
+
 			/*
 			 * Use SEQ_LAST as time_seq to do special search, which
 			 * doesn't lock tree or delayed_refs and search current
@@ -2025,8 +2046,11 @@ int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans,
 					record->bytenr, SEQ_LAST, &new_roots);
 			if (ret < 0)
 				goto cleanup;
-			if (qgroup_to_skip)
+			if (qgroup_to_skip) {
 				ulist_del(new_roots, qgroup_to_skip, 0);
+				ulist_del(record->old_roots, qgroup_to_skip,
+					  0);
+			}
 			ret = btrfs_qgroup_account_extent(trans, fs_info,
 					record->bytenr, record->num_bytes,
 					record->old_roots, new_roots);
@@ -2338,6 +2362,11 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce)
 
 	if (num_bytes == 0)
 		return 0;
+
+	if (test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags) &&
+	    capable(CAP_SYS_RESOURCE))
+		enforce = false;
+
 retry:
 	spin_lock(&fs_info->qgroup_lock);
 	quota_root = fs_info->quota_root;
@@ -2376,7 +2405,7 @@ retry:
 				ret = btrfs_start_delalloc_inodes(root, 0);
 				if (ret)
 					return ret;
-				btrfs_wait_ordered_extents(root, -1, 0, (u64)-1);
+				btrfs_wait_ordered_extents(root, U64_MAX, 0, (u64)-1);
 				trans = btrfs_join_transaction(root);
 				if (IS_ERR(trans))
 					return PTR_ERR(trans);
@@ -2806,55 +2835,130 @@ btrfs_qgroup_rescan_resume(struct btrfs_fs_info *fs_info)
  * Return <0 for error (including -EQUOT)
  *
  * NOTE: this function may sleep for memory allocation.
+ *       if btrfs_qgroup_reserve_data() is called multiple times with
+ *       same @reserved, caller must ensure when error happens it's OK
+ *       to free *ALL* reserved space.
  */
-int btrfs_qgroup_reserve_data(struct inode *inode, u64 start, u64 len)
+int btrfs_qgroup_reserve_data(struct inode *inode,
+			struct extent_changeset **reserved_ret, u64 start,
+			u64 len)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
-	struct extent_changeset changeset;
 	struct ulist_node *unode;
 	struct ulist_iterator uiter;
+	struct extent_changeset *reserved;
+	u64 orig_reserved;
+	u64 to_reserve;
 	int ret;
 
 	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &root->fs_info->flags) ||
 	    !is_fstree(root->objectid) || len == 0)
 		return 0;
 
-	changeset.bytes_changed = 0;
-	ulist_init(&changeset.range_changed);
+	/* @reserved parameter is mandatory for qgroup */
+	if (WARN_ON(!reserved_ret))
+		return -EINVAL;
+	if (!*reserved_ret) {
+		*reserved_ret = extent_changeset_alloc();
+		if (!*reserved_ret)
+			return -ENOMEM;
+	}
+	reserved = *reserved_ret;
+	/* Record already reserved space */
+	orig_reserved = reserved->bytes_changed;
 	ret = set_record_extent_bits(&BTRFS_I(inode)->io_tree, start,
-			start + len -1, EXTENT_QGROUP_RESERVED, &changeset);
+			start + len -1, EXTENT_QGROUP_RESERVED, reserved);
+
+	/* Newly reserved space */
+	to_reserve = reserved->bytes_changed - orig_reserved;
 	trace_btrfs_qgroup_reserve_data(inode, start, len,
-					changeset.bytes_changed,
-					QGROUP_RESERVE);
+					to_reserve, QGROUP_RESERVE);
 	if (ret < 0)
 		goto cleanup;
-	ret = qgroup_reserve(root, changeset.bytes_changed, true);
+	ret = qgroup_reserve(root, to_reserve, true);
 	if (ret < 0)
 		goto cleanup;
 
-	ulist_release(&changeset.range_changed);
 	return ret;
 
 cleanup:
-	/* cleanup already reserved ranges */
+	/* cleanup *ALL* already reserved ranges */
 	ULIST_ITER_INIT(&uiter);
-	while ((unode = ulist_next(&changeset.range_changed, &uiter)))
+	while ((unode = ulist_next(&reserved->range_changed, &uiter)))
 		clear_extent_bit(&BTRFS_I(inode)->io_tree, unode->val,
 				 unode->aux, EXTENT_QGROUP_RESERVED, 0, 0, NULL,
 				 GFP_NOFS);
-	ulist_release(&changeset.range_changed);
+	extent_changeset_release(reserved);
+	return ret;
+}
+
+/* Free ranges specified by @reserved, normally in error path */
+static int qgroup_free_reserved_data(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct ulist_node *unode;
+	struct ulist_iterator uiter;
+	struct extent_changeset changeset;
+	int freed = 0;
+	int ret;
+
+	extent_changeset_init(&changeset);
+	len = round_up(start + len, root->fs_info->sectorsize);
+	start = round_down(start, root->fs_info->sectorsize);
+
+	ULIST_ITER_INIT(&uiter);
+	while ((unode = ulist_next(&reserved->range_changed, &uiter))) {
+		u64 range_start = unode->val;
+		/* unode->aux is the inclusive end */
+		u64 range_len = unode->aux - range_start + 1;
+		u64 free_start;
+		u64 free_len;
+
+		extent_changeset_release(&changeset);
+
+		/* Only free range in range [start, start + len) */
+		if (range_start >= start + len ||
+		    range_start + range_len <= start)
+			continue;
+		free_start = max(range_start, start);
+		free_len = min(start + len, range_start + range_len) -
+			   free_start;
+		/*
+		 * TODO: To also modify reserved->ranges_reserved to reflect
+		 * the modification.
+		 *
+		 * However as long as we free qgroup reserved according to
+		 * EXTENT_QGROUP_RESERVED, we won't double free.
+		 * So not need to rush.
+		 */
+		ret = clear_record_extent_bits(&BTRFS_I(inode)->io_failure_tree,
+				free_start, free_start + free_len - 1,
+				EXTENT_QGROUP_RESERVED, &changeset);
+		if (ret < 0)
+			goto out;
+		freed += changeset.bytes_changed;
+	}
+	btrfs_qgroup_free_refroot(root->fs_info, root->objectid, freed);
+	ret = freed;
+out:
+	extent_changeset_release(&changeset);
 	return ret;
 }
 
-static int __btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len,
-				       int free)
+static int __btrfs_qgroup_release_data(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len,
+			int free)
 {
 	struct extent_changeset changeset;
 	int trace_op = QGROUP_RELEASE;
 	int ret;
 
-	changeset.bytes_changed = 0;
-	ulist_init(&changeset.range_changed);
+	/* In release case, we shouldn't have @reserved */
+	WARN_ON(!free && reserved);
+	if (free && reserved)
+		return qgroup_free_reserved_data(inode, reserved, start, len);
+	extent_changeset_init(&changeset);
 	ret = clear_record_extent_bits(&BTRFS_I(inode)->io_tree, start, 
 			start + len -1, EXTENT_QGROUP_RESERVED, &changeset);
 	if (ret < 0)
@@ -2868,8 +2972,9 @@ static int __btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len,
 		btrfs_qgroup_free_refroot(BTRFS_I(inode)->root->fs_info,
 				BTRFS_I(inode)->root->objectid,
 				changeset.bytes_changed);
+	ret = changeset.bytes_changed;
 out:
-	ulist_release(&changeset.range_changed);
+	extent_changeset_release(&changeset);
 	return ret;
 }
 
@@ -2878,14 +2983,17 @@ out:
  *
  * Should be called when a range of pages get invalidated before reaching disk.
  * Or for error cleanup case.
+ * if @reserved is given, only reserved range in [@start, @start + @len) will
+ * be freed.
  *
  * For data written to disk, use btrfs_qgroup_release_data().
  *
  * NOTE: This function may sleep for memory allocation.
  */
-int btrfs_qgroup_free_data(struct inode *inode, u64 start, u64 len)
+int btrfs_qgroup_free_data(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len)
 {
-	return __btrfs_qgroup_release_data(inode, start, len, 1);
+	return __btrfs_qgroup_release_data(inode, reserved, start, len, 1);
 }
 
 /*
@@ -2905,7 +3013,7 @@ int btrfs_qgroup_free_data(struct inode *inode, u64 start, u64 len)
  */
 int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len)
 {
-	return __btrfs_qgroup_release_data(inode, start, len, 0);
+	return __btrfs_qgroup_release_data(inode, NULL, start, len, 0);
 }
 
 int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
@@ -2969,8 +3077,7 @@ void btrfs_qgroup_check_reserved_leak(struct inode *inode)
 	struct ulist_iterator iter;
 	int ret;
 
-	changeset.bytes_changed = 0;
-	ulist_init(&changeset.range_changed);
+	extent_changeset_init(&changeset);
 	ret = clear_record_extent_bits(&BTRFS_I(inode)->io_tree, 0, (u64)-1,
 			EXTENT_QGROUP_RESERVED, &changeset);
 
@@ -2987,5 +3094,5 @@ void btrfs_qgroup_check_reserved_leak(struct inode *inode)
 				changeset.bytes_changed);
 
 	}
-	ulist_release(&changeset.range_changed);
+	extent_changeset_release(&changeset);
 }

fs/btrfs/qgroup.h

@@ -134,8 +134,7 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans,
 int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info);
 void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info);
 struct btrfs_delayed_extent_op;
-int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans,
-					 struct btrfs_fs_info *fs_info);
+
 /*
  * Inform qgroup to trace one dirty extent, its info is recorded in @record.
  * So qgroup can account it at transaction committing time.
@@ -243,9 +242,11 @@ int btrfs_verify_qgroup_counts(struct btrfs_fs_info *fs_info, u64 qgroupid,
 #endif
 
 /* New io_tree based accurate qgroup reserve API */
-int btrfs_qgroup_reserve_data(struct inode *inode, u64 start, u64 len);
+int btrfs_qgroup_reserve_data(struct inode *inode,
+			struct extent_changeset **reserved, u64 start, u64 len);
 int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len);
-int btrfs_qgroup_free_data(struct inode *inode, u64 start, u64 len);
+int btrfs_qgroup_free_data(struct inode *inode,
+			struct extent_changeset *reserved, u64 start, u64 len);
 
 int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes,
 			      bool enforce);

fs/btrfs/raid56.c

@@ -31,7 +31,7 @@
 #include <linux/hash.h>
 #include <linux/list_sort.h>
 #include <linux/raid/xor.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <asm/div64.h>
 #include "ctree.h"
 #include "extent_map.h"
@@ -218,12 +218,9 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info)
 	 * of a failing mount.
 	 */
 	table_size = sizeof(*table) + sizeof(*h) * num_entries;
-	table = kzalloc(table_size, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
-	if (!table) {
-		table = vzalloc(table_size);
-		if (!table)
-			return -ENOMEM;
-	}
+	table = kvzalloc(table_size, GFP_KERNEL);
+	if (!table)
+		return -ENOMEM;
 
 	spin_lock_init(&table->cache_lock);
 	INIT_LIST_HEAD(&table->stripe_cache);
@@ -1101,10 +1098,7 @@ static int rbio_add_io_page(struct btrfs_raid_bio *rbio,
 	}
 
 	/* put a new bio on the list */
-	bio = btrfs_io_bio_alloc(GFP_NOFS, bio_max_len >> PAGE_SHIFT?:1);
-	if (!bio)
-		return -ENOMEM;
-
+	bio = btrfs_io_bio_alloc(bio_max_len >> PAGE_SHIFT ?: 1);
 	bio->bi_iter.bi_size = 0;
 	bio->bi_bdev = stripe->dev->bdev;
 	bio->bi_iter.bi_sector = disk_start >> 9;

fs/btrfs/reada.c

@@ -66,7 +66,6 @@ struct reada_extctl {
 struct reada_extent {
 	u64			logical;
 	struct btrfs_key	top;
-	int			err;
 	struct list_head	extctl;
 	int 			refcnt;
 	spinlock_t		lock;

fs/btrfs/relocation.c

@@ -3093,11 +3093,12 @@ int prealloc_file_extent_cluster(struct inode *inode,
 	u64 prealloc_start = cluster->start - offset;
 	u64 prealloc_end = cluster->end - offset;
 	u64 cur_offset;
+	struct extent_changeset *data_reserved = NULL;
 
 	BUG_ON(cluster->start != cluster->boundary[0]);
 	inode_lock(inode);
 
-	ret = btrfs_check_data_free_space(inode, prealloc_start,
+	ret = btrfs_check_data_free_space(inode, &data_reserved, prealloc_start,
 					  prealloc_end + 1 - prealloc_start);
 	if (ret)
 		goto out;
@@ -3113,8 +3114,8 @@ int prealloc_file_extent_cluster(struct inode *inode,
 		lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		num_bytes = end + 1 - start;
 		if (cur_offset < start)
-			btrfs_free_reserved_data_space(inode, cur_offset,
-					start - cur_offset);
+			btrfs_free_reserved_data_space(inode, data_reserved,
+					cur_offset, start - cur_offset);
 		ret = btrfs_prealloc_file_range(inode, 0, start,
 						num_bytes, num_bytes,
 						end + 1, &alloc_hint);
@@ -3125,10 +3126,11 @@ int prealloc_file_extent_cluster(struct inode *inode,
 		nr++;
 	}
 	if (cur_offset < prealloc_end)
-		btrfs_free_reserved_data_space(inode, cur_offset,
-				       prealloc_end + 1 - cur_offset);
+		btrfs_free_reserved_data_space(inode, data_reserved,
+				cur_offset, prealloc_end + 1 - cur_offset);
 out:
 	inode_unlock(inode);
+	extent_changeset_free(data_reserved);
 	return ret;
 }
 
@@ -4269,8 +4271,7 @@ static struct reloc_control *alloc_reloc_control(struct btrfs_fs_info *fs_info)
 	INIT_LIST_HEAD(&rc->reloc_roots);
 	backref_cache_init(&rc->backref_cache);
 	mapping_tree_init(&rc->reloc_root_tree);
-	extent_io_tree_init(&rc->processed_blocks,
-			    fs_info->btree_inode->i_mapping);
+	extent_io_tree_init(&rc->processed_blocks, NULL);
 	return rc;
 }
 
@@ -4372,7 +4373,7 @@ int btrfs_relocate_block_group(struct btrfs_fs_info *fs_info, u64 group_start)
 
 	btrfs_wait_block_group_reservations(rc->block_group);
 	btrfs_wait_nocow_writers(rc->block_group);
-	btrfs_wait_ordered_roots(fs_info, -1,
+	btrfs_wait_ordered_roots(fs_info, U64_MAX,
 				 rc->block_group->key.objectid,
 				 rc->block_group->key.offset);
 

fs/btrfs/root-tree.c

@@ -390,6 +390,13 @@ again:
 		WARN_ON(btrfs_root_ref_dirid(leaf, ref) != dirid);
 		WARN_ON(btrfs_root_ref_name_len(leaf, ref) != name_len);
 		ptr = (unsigned long)(ref + 1);
+		ret = btrfs_is_name_len_valid(leaf, path->slots[0], ptr,
+					      name_len);
+		if (!ret) {
+			err = -EIO;
+			goto out;
+		}
+
 		WARN_ON(memcmp_extent_buffer(leaf, name, ptr, name_len));
 		*sequence = btrfs_root_ref_sequence(leaf, ref);
 

fs/btrfs/scrub.c

@@ -18,6 +18,7 @@
 
 #include <linux/blkdev.h>
 #include <linux/ratelimit.h>
+#include <linux/sched/mm.h>
 #include "ctree.h"
 #include "volumes.h"
 #include "disk-io.h"
@@ -161,14 +162,6 @@ struct scrub_parity {
 	unsigned long		bitmap[0];
 };
 
-struct scrub_wr_ctx {
-	struct scrub_bio *wr_curr_bio;
-	struct btrfs_device *tgtdev;
-	int pages_per_wr_bio; /* <= SCRUB_PAGES_PER_WR_BIO */
-	atomic_t flush_all_writes;
-	struct mutex wr_lock;
-};
-
 struct scrub_ctx {
 	struct scrub_bio	*bios[SCRUB_BIOS_PER_SCTX];
 	struct btrfs_fs_info	*fs_info;
@@ -183,11 +176,14 @@ struct scrub_ctx {
 	atomic_t		cancel_req;
 	int			readonly;
 	int			pages_per_rd_bio;
-	u32			sectorsize;
-	u32			nodesize;
 
 	int			is_dev_replace;
-	struct scrub_wr_ctx	wr_ctx;
+
+	struct scrub_bio        *wr_curr_bio;
+	struct mutex            wr_lock;
+	int                     pages_per_wr_bio; /* <= SCRUB_PAGES_PER_WR_BIO */
+	atomic_t                flush_all_writes;
+	struct btrfs_device     *wr_tgtdev;
 
 	/*
 	 * statistics
@@ -289,10 +285,6 @@ static void scrub_remap_extent(struct btrfs_fs_info *fs_info,
 			       u64 *extent_physical,
 			       struct btrfs_device **extent_dev,
 			       int *extent_mirror_num);
-static int scrub_setup_wr_ctx(struct scrub_wr_ctx *wr_ctx,
-			      struct btrfs_device *dev,
-			      int is_dev_replace);
-static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx);
 static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
 				    struct scrub_page *spage);
 static void scrub_wr_submit(struct scrub_ctx *sctx);
@@ -643,8 +635,6 @@ static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx)
 	if (!sctx)
 		return;
 
-	scrub_free_wr_ctx(&sctx->wr_ctx);
-
 	/* this can happen when scrub is cancelled */
 	if (sctx->curr != -1) {
 		struct scrub_bio *sbio = sctx->bios[sctx->curr];
@@ -664,6 +654,7 @@ static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx)
 		kfree(sbio);
 	}
 
+	kfree(sctx->wr_curr_bio);
 	scrub_free_csums(sctx);
 	kfree(sctx);
 }
@@ -680,7 +671,6 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
 	struct scrub_ctx *sctx;
 	int		i;
 	struct btrfs_fs_info *fs_info = dev->fs_info;
-	int ret;
 
 	sctx = kzalloc(sizeof(*sctx), GFP_KERNEL);
 	if (!sctx)
@@ -710,8 +700,6 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
 			sctx->bios[i]->next_free = -1;
 	}
 	sctx->first_free = 0;
-	sctx->nodesize = fs_info->nodesize;
-	sctx->sectorsize = fs_info->sectorsize;
 	atomic_set(&sctx->bios_in_flight, 0);
 	atomic_set(&sctx->workers_pending, 0);
 	atomic_set(&sctx->cancel_req, 0);
@@ -722,12 +710,16 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace)
 	spin_lock_init(&sctx->stat_lock);
 	init_waitqueue_head(&sctx->list_wait);
 
-	ret = scrub_setup_wr_ctx(&sctx->wr_ctx,
-				 fs_info->dev_replace.tgtdev, is_dev_replace);
-	if (ret) {
-		scrub_free_ctx(sctx);
-		return ERR_PTR(ret);
+	WARN_ON(sctx->wr_curr_bio != NULL);
+	mutex_init(&sctx->wr_lock);
+	sctx->wr_curr_bio = NULL;
+	if (is_dev_replace) {
+		WARN_ON(!fs_info->dev_replace.tgtdev);
+		sctx->pages_per_wr_bio = SCRUB_PAGES_PER_WR_BIO;
+		sctx->wr_tgtdev = fs_info->dev_replace.tgtdev;
+		atomic_set(&sctx->flush_all_writes, 0);
 	}
+
 	return sctx;
 
 nomem:
@@ -742,6 +734,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	u32 nlink;
 	int ret;
 	int i;
+	unsigned nofs_flag;
 	struct extent_buffer *eb;
 	struct btrfs_inode_item *inode_item;
 	struct scrub_warning *swarn = warn_ctx;
@@ -780,7 +773,14 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root,
 	nlink = btrfs_inode_nlink(eb, inode_item);
 	btrfs_release_path(swarn->path);
 
+	/*
+	 * init_path might indirectly call vmalloc, or use GFP_KERNEL. Scrub
+	 * uses GFP_NOFS in this context, so we keep it consistent but it does
+	 * not seem to be strictly necessary.
+	 */
+	nofs_flag = memalloc_nofs_save();
 	ipath = init_ipath(4096, local_root, swarn->path);
+	memalloc_nofs_restore(nofs_flag);
 	if (IS_ERR(ipath)) {
 		ret = PTR_ERR(ipath);
 		ipath = NULL;
@@ -954,7 +954,7 @@ static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx)
 			ret = -EIO;
 			goto out;
 		}
-		ret = repair_io_failure(BTRFS_I(inode), offset, PAGE_SIZE,
+		ret = repair_io_failure(fs_info, inum, offset, PAGE_SIZE,
 					fixup->logical, page,
 					offset - page_offset(page),
 					fixup->mirror_num);
@@ -1737,12 +1737,7 @@ static void scrub_recheck_block(struct btrfs_fs_info *fs_info,
 		}
 
 		WARN_ON(!page->page);
-		bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-		if (!bio) {
-			page->io_error = 1;
-			sblock->no_io_error_seen = 0;
-			continue;
-		}
+		bio = btrfs_io_bio_alloc(1);
 		bio->bi_bdev = page->dev->bdev;
 
 		bio_add_page(bio, page->page, PAGE_SIZE, 0);
@@ -1830,9 +1825,7 @@ static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad,
 			return -EIO;
 		}
 
-		bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-		if (!bio)
-			return -EIO;
+		bio = btrfs_io_bio_alloc(1);
 		bio->bi_bdev = page_bad->dev->bdev;
 		bio->bi_iter.bi_sector = page_bad->physical >> 9;
 		bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
@@ -1898,37 +1891,31 @@ static int scrub_write_page_to_dev_replace(struct scrub_block *sblock,
 static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
 				    struct scrub_page *spage)
 {
-	struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx;
 	struct scrub_bio *sbio;
 	int ret;
 
-	mutex_lock(&wr_ctx->wr_lock);
+	mutex_lock(&sctx->wr_lock);
 again:
-	if (!wr_ctx->wr_curr_bio) {
-		wr_ctx->wr_curr_bio = kzalloc(sizeof(*wr_ctx->wr_curr_bio),
+	if (!sctx->wr_curr_bio) {
+		sctx->wr_curr_bio = kzalloc(sizeof(*sctx->wr_curr_bio),
 					      GFP_KERNEL);
-		if (!wr_ctx->wr_curr_bio) {
-			mutex_unlock(&wr_ctx->wr_lock);
+		if (!sctx->wr_curr_bio) {
+			mutex_unlock(&sctx->wr_lock);
 			return -ENOMEM;
 		}
-		wr_ctx->wr_curr_bio->sctx = sctx;
-		wr_ctx->wr_curr_bio->page_count = 0;
+		sctx->wr_curr_bio->sctx = sctx;
+		sctx->wr_curr_bio->page_count = 0;
 	}
-	sbio = wr_ctx->wr_curr_bio;
+	sbio = sctx->wr_curr_bio;
 	if (sbio->page_count == 0) {
 		struct bio *bio;
 
 		sbio->physical = spage->physical_for_dev_replace;
 		sbio->logical = spage->logical;
-		sbio->dev = wr_ctx->tgtdev;
+		sbio->dev = sctx->wr_tgtdev;
 		bio = sbio->bio;
 		if (!bio) {
-			bio = btrfs_io_bio_alloc(GFP_KERNEL,
-					wr_ctx->pages_per_wr_bio);
-			if (!bio) {
-				mutex_unlock(&wr_ctx->wr_lock);
-				return -ENOMEM;
-			}
+			bio = btrfs_io_bio_alloc(sctx->pages_per_wr_bio);
 			sbio->bio = bio;
 		}
 
@@ -1951,7 +1938,7 @@ again:
 		if (sbio->page_count < 1) {
 			bio_put(sbio->bio);
 			sbio->bio = NULL;
-			mutex_unlock(&wr_ctx->wr_lock);
+			mutex_unlock(&sctx->wr_lock);
 			return -EIO;
 		}
 		scrub_wr_submit(sctx);
@@ -1961,23 +1948,22 @@ again:
 	sbio->pagev[sbio->page_count] = spage;
 	scrub_page_get(spage);
 	sbio->page_count++;
-	if (sbio->page_count == wr_ctx->pages_per_wr_bio)
+	if (sbio->page_count == sctx->pages_per_wr_bio)
 		scrub_wr_submit(sctx);
-	mutex_unlock(&wr_ctx->wr_lock);
+	mutex_unlock(&sctx->wr_lock);
 
 	return 0;
 }
 
 static void scrub_wr_submit(struct scrub_ctx *sctx)
 {
-	struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx;
 	struct scrub_bio *sbio;
 
-	if (!wr_ctx->wr_curr_bio)
+	if (!sctx->wr_curr_bio)
 		return;
 
-	sbio = wr_ctx->wr_curr_bio;
-	wr_ctx->wr_curr_bio = NULL;
+	sbio = sctx->wr_curr_bio;
+	sctx->wr_curr_bio = NULL;
 	WARN_ON(!sbio->bio->bi_bdev);
 	scrub_pending_bio_inc(sctx);
 	/* process all writes in a single worker thread. Then the block layer
@@ -2081,7 +2067,7 @@ static int scrub_checksum_data(struct scrub_block *sblock)
 	page = sblock->pagev[0]->page;
 	buffer = kmap_atomic(page);
 
-	len = sctx->sectorsize;
+	len = sctx->fs_info->sectorsize;
 	index = 0;
 	for (;;) {
 		u64 l = min_t(u64, len, PAGE_SIZE);
@@ -2146,7 +2132,7 @@ static int scrub_checksum_tree_block(struct scrub_block *sblock)
 		   BTRFS_UUID_SIZE))
 		sblock->header_error = 1;
 
-	len = sctx->nodesize - BTRFS_CSUM_SIZE;
+	len = sctx->fs_info->nodesize - BTRFS_CSUM_SIZE;
 	mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE;
 	p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE;
 	index = 0;
@@ -2329,10 +2315,7 @@ again:
 		sbio->dev = spage->dev;
 		bio = sbio->bio;
 		if (!bio) {
-			bio = btrfs_io_bio_alloc(GFP_KERNEL,
-					sctx->pages_per_rd_bio);
-			if (!bio)
-				return -ENOMEM;
+			bio = btrfs_io_bio_alloc(sctx->pages_per_rd_bio);
 			sbio->bio = bio;
 		}
 
@@ -2420,10 +2403,10 @@ static void scrub_missing_raid56_worker(struct btrfs_work *work)
 	scrub_block_put(sblock);
 
 	if (sctx->is_dev_replace &&
-	    atomic_read(&sctx->wr_ctx.flush_all_writes)) {
-		mutex_lock(&sctx->wr_ctx.wr_lock);
+	    atomic_read(&sctx->flush_all_writes)) {
+		mutex_lock(&sctx->wr_lock);
 		scrub_wr_submit(sctx);
-		mutex_unlock(&sctx->wr_ctx.wr_lock);
+		mutex_unlock(&sctx->wr_lock);
 	}
 
 	scrub_pending_bio_dec(sctx);
@@ -2458,10 +2441,7 @@ static void scrub_missing_raid56_pages(struct scrub_block *sblock)
 		goto bbio_out;
 	}
 
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 0);
-	if (!bio)
-		goto bbio_out;
-
+	bio = btrfs_io_bio_alloc(0);
 	bio->bi_iter.bi_sector = logical >> 9;
 	bio->bi_private = sblock;
 	bio->bi_end_io = scrub_missing_raid56_end_io;
@@ -2628,10 +2608,10 @@ static void scrub_bio_end_io_worker(struct btrfs_work *work)
 	spin_unlock(&sctx->list_lock);
 
 	if (sctx->is_dev_replace &&
-	    atomic_read(&sctx->wr_ctx.flush_all_writes)) {
-		mutex_lock(&sctx->wr_ctx.wr_lock);
+	    atomic_read(&sctx->flush_all_writes)) {
+		mutex_lock(&sctx->wr_lock);
 		scrub_wr_submit(sctx);
-		mutex_unlock(&sctx->wr_ctx.wr_lock);
+		mutex_unlock(&sctx->wr_lock);
 	}
 
 	scrub_pending_bio_dec(sctx);
@@ -2726,8 +2706,8 @@ static int scrub_find_csum(struct scrub_ctx *sctx, u64 logical, u8 *csum)
 	if (!sum)
 		return 0;
 
-	index = ((u32)(logical - sum->bytenr)) / sctx->sectorsize;
-	num_sectors = sum->len / sctx->sectorsize;
+	index = ((u32)(logical - sum->bytenr)) / sctx->fs_info->sectorsize;
+	num_sectors = sum->len / sctx->fs_info->sectorsize;
 	memcpy(csum, sum->sums + index, sctx->csum_size);
 	if (index == num_sectors - 1) {
 		list_del(&sum->list);
@@ -2746,19 +2726,19 @@ static int scrub_extent(struct scrub_ctx *sctx, u64 logical, u64 len,
 	u32 blocksize;
 
 	if (flags & BTRFS_EXTENT_FLAG_DATA) {
-		blocksize = sctx->sectorsize;
+		blocksize = sctx->fs_info->sectorsize;
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.data_extents_scrubbed++;
 		sctx->stat.data_bytes_scrubbed += len;
 		spin_unlock(&sctx->stat_lock);
 	} else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
-		blocksize = sctx->nodesize;
+		blocksize = sctx->fs_info->nodesize;
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.tree_extents_scrubbed++;
 		sctx->stat.tree_bytes_scrubbed += len;
 		spin_unlock(&sctx->stat_lock);
 	} else {
-		blocksize = sctx->sectorsize;
+		blocksize = sctx->fs_info->sectorsize;
 		WARN_ON(1);
 	}
 
@@ -2892,11 +2872,11 @@ static int scrub_extent_for_parity(struct scrub_parity *sparity,
 	}
 
 	if (flags & BTRFS_EXTENT_FLAG_DATA) {
-		blocksize = sctx->sectorsize;
+		blocksize = sctx->fs_info->sectorsize;
 	} else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
-		blocksize = sctx->nodesize;
+		blocksize = sctx->fs_info->nodesize;
 	} else {
-		blocksize = sctx->sectorsize;
+		blocksize = sctx->fs_info->sectorsize;
 		WARN_ON(1);
 	}
 
@@ -3037,10 +3017,7 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity)
 	if (ret || !bbio || !bbio->raid_map)
 		goto bbio_out;
 
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 0);
-	if (!bio)
-		goto bbio_out;
-
+	bio = btrfs_io_bio_alloc(0);
 	bio->bi_iter.bi_sector = sparity->logic_start >> 9;
 	bio->bi_private = sparity;
 	bio->bi_end_io = scrub_parity_bio_endio;
@@ -3305,9 +3282,9 @@ out:
 						logic_end - logic_start);
 	scrub_parity_put(sparity);
 	scrub_submit(sctx);
-	mutex_lock(&sctx->wr_ctx.wr_lock);
+	mutex_lock(&sctx->wr_lock);
 	scrub_wr_submit(sctx);
-	mutex_unlock(&sctx->wr_ctx.wr_lock);
+	mutex_unlock(&sctx->wr_lock);
 
 	btrfs_release_path(path);
 	return ret < 0 ? ret : 0;
@@ -3463,14 +3440,14 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 		 */
 		if (atomic_read(&fs_info->scrub_pause_req)) {
 			/* push queued extents */
-			atomic_set(&sctx->wr_ctx.flush_all_writes, 1);
+			atomic_set(&sctx->flush_all_writes, 1);
 			scrub_submit(sctx);
-			mutex_lock(&sctx->wr_ctx.wr_lock);
+			mutex_lock(&sctx->wr_lock);
 			scrub_wr_submit(sctx);
-			mutex_unlock(&sctx->wr_ctx.wr_lock);
+			mutex_unlock(&sctx->wr_lock);
 			wait_event(sctx->list_wait,
 				   atomic_read(&sctx->bios_in_flight) == 0);
-			atomic_set(&sctx->wr_ctx.flush_all_writes, 0);
+			atomic_set(&sctx->flush_all_writes, 0);
 			scrub_blocked_if_needed(fs_info);
 		}
 
@@ -3677,9 +3654,9 @@ skip:
 out:
 	/* push queued extents */
 	scrub_submit(sctx);
-	mutex_lock(&sctx->wr_ctx.wr_lock);
+	mutex_lock(&sctx->wr_lock);
 	scrub_wr_submit(sctx);
-	mutex_unlock(&sctx->wr_ctx.wr_lock);
+	mutex_unlock(&sctx->wr_lock);
 
 	blk_finish_plug(&plug);
 	btrfs_free_path(path);
@@ -3859,7 +3836,7 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 			 */
 			btrfs_wait_block_group_reservations(cache);
 			btrfs_wait_nocow_writers(cache);
-			ret = btrfs_wait_ordered_roots(fs_info, -1,
+			ret = btrfs_wait_ordered_roots(fs_info, U64_MAX,
 						       cache->key.objectid,
 						       cache->key.offset);
 			if (ret > 0) {
@@ -3916,11 +3893,11 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		 * write requests are really completed when bios_in_flight
 		 * changes to 0.
 		 */
-		atomic_set(&sctx->wr_ctx.flush_all_writes, 1);
+		atomic_set(&sctx->flush_all_writes, 1);
 		scrub_submit(sctx);
-		mutex_lock(&sctx->wr_ctx.wr_lock);
+		mutex_lock(&sctx->wr_lock);
 		scrub_wr_submit(sctx);
-		mutex_unlock(&sctx->wr_ctx.wr_lock);
+		mutex_unlock(&sctx->wr_lock);
 
 		wait_event(sctx->list_wait,
 			   atomic_read(&sctx->bios_in_flight) == 0);
@@ -3934,7 +3911,7 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		 */
 		wait_event(sctx->list_wait,
 			   atomic_read(&sctx->workers_pending) == 0);
-		atomic_set(&sctx->wr_ctx.flush_all_writes, 0);
+		atomic_set(&sctx->flush_all_writes, 0);
 
 		scrub_pause_off(fs_info);
 
@@ -4337,32 +4314,6 @@ static void scrub_remap_extent(struct btrfs_fs_info *fs_info,
 	btrfs_put_bbio(bbio);
 }
 
-static int scrub_setup_wr_ctx(struct scrub_wr_ctx *wr_ctx,
-			      struct btrfs_device *dev,
-			      int is_dev_replace)
-{
-	WARN_ON(wr_ctx->wr_curr_bio != NULL);
-
-	mutex_init(&wr_ctx->wr_lock);
-	wr_ctx->wr_curr_bio = NULL;
-	if (!is_dev_replace)
-		return 0;
-
-	WARN_ON(!dev->bdev);
-	wr_ctx->pages_per_wr_bio = SCRUB_PAGES_PER_WR_BIO;
-	wr_ctx->tgtdev = dev;
-	atomic_set(&wr_ctx->flush_all_writes, 0);
-	return 0;
-}
-
-static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx)
-{
-	mutex_lock(&wr_ctx->wr_lock);
-	kfree(wr_ctx->wr_curr_bio);
-	wr_ctx->wr_curr_bio = NULL;
-	mutex_unlock(&wr_ctx->wr_lock);
-}
-
 static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len,
 			    int mirror_num, u64 physical_for_dev_replace)
 {
@@ -4665,7 +4616,7 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 	struct btrfs_device *dev;
 	int ret;
 
-	dev = sctx->wr_ctx.tgtdev;
+	dev = sctx->wr_tgtdev;
 	if (!dev)
 		return -EIO;
 	if (!dev->bdev) {
@@ -4673,13 +4624,7 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 			"scrub write_page_nocow(bdev == NULL) is unexpected");
 		return -EIO;
 	}
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 1);
-	if (!bio) {
-		spin_lock(&sctx->stat_lock);
-		sctx->stat.malloc_errors++;
-		spin_unlock(&sctx->stat_lock);
-		return -ENOMEM;
-	}
+	bio = btrfs_io_bio_alloc(1);
 	bio->bi_iter.bi_size = 0;
 	bio->bi_iter.bi_sector = physical_for_dev_replace >> 9;
 	bio->bi_bdev = dev->bdev;

fs/btrfs/send.c

@@ -1069,6 +1069,12 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 			}
 		}
 
+		ret = btrfs_is_name_len_valid(eb, path->slots[0],
+			  (unsigned long)(di + 1), name_len + data_len);
+		if (!ret) {
+			ret = -EIO;
+			goto out;
+		}
 		if (name_len + data_len > buf_len) {
 			buf_len = name_len + data_len;
 			if (is_vmalloc_addr(buf)) {
@@ -1083,7 +1089,7 @@ static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
 				buf = tmp;
 			}
 			if (!buf) {
-				buf = vmalloc(buf_len);
+				buf = kvmalloc(buf_len, GFP_KERNEL);
 				if (!buf) {
 					ret = -ENOMEM;
 					goto out;
@@ -2769,15 +2775,20 @@ out:
 
 struct recorded_ref {
 	struct list_head list;
-	char *dir_path;
 	char *name;
 	struct fs_path *full_path;
 	u64 dir;
 	u64 dir_gen;
-	int dir_path_len;
 	int name_len;
 };
 
+static void set_ref_path(struct recorded_ref *ref, struct fs_path *path)
+{
+	ref->full_path = path;
+	ref->name = (char *)kbasename(ref->full_path->start);
+	ref->name_len = ref->full_path->end - ref->name;
+}
+
 /*
  * We need to process new refs before deleted refs, but compare_tree gives us
  * everything mixed. So we first record all refs and later process them.
@@ -2794,17 +2805,7 @@ static int __record_ref(struct list_head *head, u64 dir,
 
 	ref->dir = dir;
 	ref->dir_gen = dir_gen;
-	ref->full_path = path;
-
-	ref->name = (char *)kbasename(ref->full_path->start);
-	ref->name_len = ref->full_path->end - ref->name;
-	ref->dir_path = ref->full_path->start;
-	if (ref->name == ref->full_path->start)
-		ref->dir_path_len = 0;
-	else
-		ref->dir_path_len = ref->full_path->end -
-				ref->full_path->start - 1 - ref->name_len;
-
+	set_ref_path(ref, path);
 	list_add_tail(&ref->list, head);
 	return 0;
 }
@@ -3546,9 +3547,17 @@ static int is_ancestor(struct btrfs_root *root,
 		       struct fs_path *fs_path)
 {
 	u64 ino = ino2;
+	bool free_path = false;
+	int ret = 0;
+
+	if (!fs_path) {
+		fs_path = fs_path_alloc();
+		if (!fs_path)
+			return -ENOMEM;
+		free_path = true;
+	}
 
 	while (ino > BTRFS_FIRST_FREE_OBJECTID) {
-		int ret;
 		u64 parent;
 		u64 parent_gen;
 
@@ -3557,13 +3566,18 @@ static int is_ancestor(struct btrfs_root *root,
 		if (ret < 0) {
 			if (ret == -ENOENT && ino == ino2)
 				ret = 0;
-			return ret;
+			goto out;
+		}
+		if (parent == ino1) {
+			ret = parent_gen == ino1_gen ? 1 : 0;
+			goto out;
 		}
-		if (parent == ino1)
-			return parent_gen == ino1_gen ? 1 : 0;
 		ino = parent;
 	}
-	return 0;
+ out:
+	if (free_path)
+		fs_path_free(fs_path);
+	return ret;
 }
 
 static int wait_for_parent_move(struct send_ctx *sctx,
@@ -3686,6 +3700,7 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 	int is_orphan = 0;
 	u64 last_dir_ino_rm = 0;
 	bool can_rename = true;
+	bool orphanized_ancestor = false;
 
 	btrfs_debug(fs_info, "process_recorded_refs %llu", sctx->cur_ino);
 
@@ -3837,9 +3852,16 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 				 * might contain the pre-orphanization name of
 				 * ow_inode, which is no longer valid.
 				 */
-				fs_path_reset(valid_path);
-				ret = get_cur_path(sctx, sctx->cur_ino,
-					   sctx->cur_inode_gen, valid_path);
+				ret = is_ancestor(sctx->parent_root,
+						  ow_inode, ow_gen,
+						  sctx->cur_ino, NULL);
+				if (ret > 0) {
+					orphanized_ancestor = true;
+					fs_path_reset(valid_path);
+					ret = get_cur_path(sctx, sctx->cur_ino,
+							   sctx->cur_inode_gen,
+							   valid_path);
+				}
 				if (ret < 0)
 					goto out;
 			} else {
@@ -3960,6 +3982,43 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
 			if (ret < 0)
 				goto out;
 			if (!ret) {
+				/*
+				 * If we orphanized any ancestor before, we need
+				 * to recompute the full path for deleted names,
+				 * since any such path was computed before we
+				 * processed any references and orphanized any
+				 * ancestor inode.
+				 */
+				if (orphanized_ancestor) {
+					struct fs_path *new_path;
+
+					/*
+					 * Our reference's name member points to
+					 * its full_path member string, so we
+					 * use here a new path.
+					 */
+					new_path = fs_path_alloc();
+					if (!new_path) {
+						ret = -ENOMEM;
+						goto out;
+					}
+					ret = get_cur_path(sctx, cur->dir,
+							   cur->dir_gen,
+							   new_path);
+					if (ret < 0) {
+						fs_path_free(new_path);
+						goto out;
+					}
+					ret = fs_path_add(new_path,
+							  cur->name,
+							  cur->name_len);
+					if (ret < 0) {
+						fs_path_free(new_path);
+						goto out;
+					}
+					fs_path_free(cur->full_path);
+					set_ref_path(cur, new_path);
+				}
 				ret = send_unlink(sctx, cur->full_path);
 				if (ret < 0)
 					goto out;
@@ -6397,13 +6456,10 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
 
 	alloc_size = sizeof(struct clone_root) * (arg->clone_sources_count + 1);
 
-	sctx->clone_roots = kzalloc(alloc_size, GFP_KERNEL | __GFP_NOWARN);
+	sctx->clone_roots = kzalloc(alloc_size, GFP_KERNEL);
 	if (!sctx->clone_roots) {
-		sctx->clone_roots = vzalloc(alloc_size);
-		if (!sctx->clone_roots) {
-			ret = -ENOMEM;
-			goto out;
-		}
+		ret = -ENOMEM;
+		goto out;
 	}
 
 	alloc_size = arg->clone_sources_count * sizeof(*arg->clone_sources);

fs/btrfs/super.c

@@ -601,18 +601,8 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 			}
 			break;
 		case Opt_alloc_start:
-			num = match_strdup(&args[0]);
-			if (num) {
-				mutex_lock(&info->chunk_mutex);
-				info->alloc_start = memparse(num, NULL);
-				mutex_unlock(&info->chunk_mutex);
-				kfree(num);
-				btrfs_info(info, "allocations start at %llu",
-					   info->alloc_start);
-			} else {
-				ret = -ENOMEM;
-				goto out;
-			}
+			btrfs_info(info,
+				"option alloc_start is obsolete, ignored");
 			break;
 		case Opt_acl:
 #ifdef CONFIG_BTRFS_FS_POSIX_ACL
@@ -1187,7 +1177,7 @@ int btrfs_sync_fs(struct super_block *sb, int wait)
 		return 0;
 	}
 
-	btrfs_wait_ordered_roots(fs_info, -1, 0, (u64)-1);
+	btrfs_wait_ordered_roots(fs_info, U64_MAX, 0, (u64)-1);
 
 	trans = btrfs_attach_transaction_barrier(root);
 	if (IS_ERR(trans)) {
@@ -1232,8 +1222,6 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",nobarrier");
 	if (info->max_inline != BTRFS_DEFAULT_MAX_INLINE)
 		seq_printf(seq, ",max_inline=%llu", info->max_inline);
-	if (info->alloc_start != 0)
-		seq_printf(seq, ",alloc_start=%llu", info->alloc_start);
 	if (info->thread_pool_size !=  min_t(unsigned long,
 					     num_online_cpus() + 2, 8))
 		seq_printf(seq, ",thread_pool=%d", info->thread_pool_size);
@@ -1716,7 +1704,6 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 	unsigned long old_opts = fs_info->mount_opt;
 	unsigned long old_compress_type = fs_info->compress_type;
 	u64 old_max_inline = fs_info->max_inline;
-	u64 old_alloc_start = fs_info->alloc_start;
 	int old_thread_pool_size = fs_info->thread_pool_size;
 	unsigned int old_metadata_ratio = fs_info->metadata_ratio;
 	int ret;
@@ -1855,9 +1842,6 @@ restore:
 	fs_info->mount_opt = old_opts;
 	fs_info->compress_type = old_compress_type;
 	fs_info->max_inline = old_max_inline;
-	mutex_lock(&fs_info->chunk_mutex);
-	fs_info->alloc_start = old_alloc_start;
-	mutex_unlock(&fs_info->chunk_mutex);
 	btrfs_resize_thread_pool(fs_info,
 		old_thread_pool_size, fs_info->thread_pool_size);
 	fs_info->metadata_ratio = old_metadata_ratio;
@@ -1898,18 +1882,15 @@ static inline void btrfs_descending_sort_devices(
 static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
 				       u64 *free_bytes)
 {
-	struct btrfs_root *root = fs_info->tree_root;
 	struct btrfs_device_info *devices_info;
 	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
 	struct btrfs_device *device;
 	u64 skip_space;
 	u64 type;
 	u64 avail_space;
-	u64 used_space;
 	u64 min_stripe_size;
 	int min_stripes = 1, num_stripes = 1;
 	int i = 0, nr_devices;
-	int ret;
 
 	/*
 	 * We aren't under the device list lock, so this is racy-ish, but good
@@ -1927,12 +1908,12 @@ static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
 	}
 
 	devices_info = kmalloc_array(nr_devices, sizeof(*devices_info),
-			       GFP_NOFS);
+			       GFP_KERNEL);
 	if (!devices_info)
 		return -ENOMEM;
 
 	/* calc min stripe number for data space allocation */
-	type = btrfs_get_alloc_profile(root, 1);
+	type = btrfs_data_alloc_profile(fs_info);
 	if (type & BTRFS_BLOCK_GROUP_RAID0) {
 		min_stripes = 2;
 		num_stripes = nr_devices;
@@ -1949,8 +1930,6 @@ static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
 	else
 		min_stripe_size = BTRFS_STRIPE_LEN;
 
-	if (fs_info->alloc_start)
-		mutex_lock(&fs_devices->device_list_mutex);
 	rcu_read_lock();
 	list_for_each_entry_rcu(device, &fs_devices->devices, dev_list) {
 		if (!device->in_fs_metadata || !device->bdev ||
@@ -1973,34 +1952,6 @@ static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
 		 */
 		skip_space = SZ_1M;
 
-		/* user can set the offset in fs_info->alloc_start. */
-		if (fs_info->alloc_start &&
-		    fs_info->alloc_start + BTRFS_STRIPE_LEN <=
-		    device->total_bytes) {
-			rcu_read_unlock();
-			skip_space = max(fs_info->alloc_start, skip_space);
-
-			/*
-			 * btrfs can not use the free space in
-			 * [0, skip_space - 1], we must subtract it from the
-			 * total. In order to implement it, we account the used
-			 * space in this range first.
-			 */
-			ret = btrfs_account_dev_extents_size(device, 0,
-							     skip_space - 1,
-							     &used_space);
-			if (ret) {
-				kfree(devices_info);
-				mutex_unlock(&fs_devices->device_list_mutex);
-				return ret;
-			}
-
-			rcu_read_lock();
-
-			/* calc the free space in [0, skip_space - 1] */
-			skip_space -= used_space;
-		}
-
 		/*
 		 * we can use the free space in [0, skip_space - 1], subtract
 		 * it from the total.
@@ -2019,8 +1970,6 @@ static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
 		i++;
 	}
 	rcu_read_unlock();
-	if (fs_info->alloc_start)
-		mutex_unlock(&fs_devices->device_list_mutex);
 
 	nr_devices = i;
 
@@ -2057,10 +2006,9 @@ static int btrfs_calc_avail_data_space(struct btrfs_fs_info *fs_info,
  * multiplier to scale the sizes.
  *
  * Unused device space usage is based on simulating the chunk allocator
- * algorithm that respects the device sizes, order of allocations and the
- * 'alloc_start' value, this is a close approximation of the actual use but
- * there are other factors that may change the result (like a new metadata
- * chunk).
+ * algorithm that respects the device sizes and order of allocations.  This is
+ * a close approximation of the actual use but there are other factors that may
+ * change the result (like a new metadata chunk).
  *
  * If metadata is exhausted, f_bavail will be 0.
  */
@@ -2243,7 +2191,7 @@ static int btrfs_freeze(struct super_block *sb)
 	struct btrfs_fs_info *fs_info = btrfs_sb(sb);
 	struct btrfs_root *root = fs_info->tree_root;
 
-	fs_info->fs_frozen = 1;
+	set_bit(BTRFS_FS_FROZEN, &fs_info->flags);
 	/*
 	 * We don't need a barrier here, we'll wait for any transaction that
 	 * could be in progress on other threads (and do delayed iputs that
@@ -2262,7 +2210,9 @@ static int btrfs_freeze(struct super_block *sb)
 
 static int btrfs_unfreeze(struct super_block *sb)
 {
-	btrfs_sb(sb)->fs_frozen = 0;
+	struct btrfs_fs_info *fs_info = btrfs_sb(sb);
+
+	clear_bit(BTRFS_FS_FROZEN, &fs_info->flags);
 	return 0;
 }
 

fs/btrfs/sysfs.c

@@ -447,11 +447,52 @@ static ssize_t btrfs_clone_alignment_show(struct kobject *kobj,
 
 BTRFS_ATTR(clone_alignment, btrfs_clone_alignment_show);
 
+static ssize_t quota_override_show(struct kobject *kobj,
+				   struct kobj_attribute *a, char *buf)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	int quota_override;
+
+	quota_override = test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+	return snprintf(buf, PAGE_SIZE, "%d\n", quota_override);
+}
+
+static ssize_t quota_override_store(struct kobject *kobj,
+				    struct kobj_attribute *a,
+				    const char *buf, size_t len)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	unsigned long knob;
+	int err;
+
+	if (!fs_info)
+		return -EPERM;
+
+	if (!capable(CAP_SYS_RESOURCE))
+		return -EPERM;
+
+	err = kstrtoul(buf, 10, &knob);
+	if (err)
+		return err;
+	if (knob > 1)
+		return -EINVAL;
+
+	if (knob)
+		set_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+	else
+		clear_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+
+	return len;
+}
+
+BTRFS_ATTR_RW(quota_override, quota_override_show, quota_override_store);
+
 static const struct attribute *btrfs_attrs[] = {
 	BTRFS_ATTR_PTR(label),
 	BTRFS_ATTR_PTR(nodesize),
 	BTRFS_ATTR_PTR(sectorsize),
 	BTRFS_ATTR_PTR(clone_alignment),
+	BTRFS_ATTR_PTR(quota_override),
 	NULL,
 };
 

fs/btrfs/tests/extent-io-tests.c

@@ -87,7 +87,7 @@ static int test_find_delalloc(u32 sectorsize)
 		return -ENOMEM;
 	}
 
-	extent_io_tree_init(&tmp, &inode->i_data);
+	extent_io_tree_init(&tmp, inode);
 
 	/*
 	 * First go through and create and mark all of our pages dirty, we pin

fs/btrfs/transaction.c

@@ -93,7 +93,7 @@ void btrfs_put_transaction(struct btrfs_transaction *transaction)
 			btrfs_put_block_group_trimming(cache);
 			btrfs_put_block_group(cache);
 		}
-		kmem_cache_free(btrfs_transaction_cachep, transaction);
+		kfree(transaction);
 	}
 }
 
@@ -228,7 +228,7 @@ loop:
 	 */
 	BUG_ON(type == TRANS_JOIN_NOLOCK);
 
-	cur_trans = kmem_cache_alloc(btrfs_transaction_cachep, GFP_NOFS);
+	cur_trans = kmalloc(sizeof(*cur_trans), GFP_NOFS);
 	if (!cur_trans)
 		return -ENOMEM;
 
@@ -238,11 +238,11 @@ loop:
 		 * someone started a transaction after we unlocked.  Make sure
 		 * to redo the checks above
 		 */
-		kmem_cache_free(btrfs_transaction_cachep, cur_trans);
+		kfree(cur_trans);
 		goto loop;
 	} else if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) {
 		spin_unlock(&fs_info->trans_lock);
-		kmem_cache_free(btrfs_transaction_cachep, cur_trans);
+		kfree(cur_trans);
 		return -EROFS;
 	}
 
@@ -294,7 +294,7 @@ loop:
 	spin_lock_init(&cur_trans->dropped_roots_lock);
 	list_add_tail(&cur_trans->list, &fs_info->trans_list);
 	extent_io_tree_init(&cur_trans->dirty_pages,
-			     fs_info->btree_inode->i_mapping);
+			     fs_info->btree_inode);
 	fs_info->generation++;
 	cur_trans->transid = fs_info->generation;
 	fs_info->running_transaction = cur_trans;
@@ -1374,9 +1374,6 @@ static int qgroup_account_snapshot(struct btrfs_trans_handle *trans,
 	ret = commit_fs_roots(trans, fs_info);
 	if (ret)
 		goto out;
-	ret = btrfs_qgroup_prepare_account_extents(trans, fs_info);
-	if (ret < 0)
-		goto out;
 	ret = btrfs_qgroup_account_extents(trans, fs_info);
 	if (ret < 0)
 		goto out;
@@ -1926,7 +1923,7 @@ static inline int btrfs_start_delalloc_flush(struct btrfs_fs_info *fs_info)
 static inline void btrfs_wait_delalloc_flush(struct btrfs_fs_info *fs_info)
 {
 	if (btrfs_test_opt(fs_info, FLUSHONCOMMIT))
-		btrfs_wait_ordered_roots(fs_info, -1, 0, (u64)-1);
+		btrfs_wait_ordered_roots(fs_info, U64_MAX, 0, (u64)-1);
 }
 
 static inline void
@@ -2180,13 +2177,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
 		goto scrub_continue;
 	}
 
-	ret = btrfs_qgroup_prepare_account_extents(trans, fs_info);
-	if (ret) {
-		mutex_unlock(&fs_info->tree_log_mutex);
-		mutex_unlock(&fs_info->reloc_mutex);
-		goto scrub_continue;
-	}
-
 	/*
 	 * Since fs roots are all committed, we can get a quite accurate
 	 * new_roots. So let's do quota accounting.
@@ -2314,7 +2304,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
 	 * it'll result in deadlock about SB_FREEZE_FS.
 	 */
 	if (current != fs_info->transaction_kthread &&
-	    current != fs_info->cleaner_kthread && !fs_info->fs_frozen)
+	    current != fs_info->cleaner_kthread &&
+	    !test_bit(BTRFS_FS_FROZEN, &fs_info->flags))
 		btrfs_run_delayed_iputs(fs_info);
 
 	return ret;

fs/btrfs/tree-log.c

@@ -1175,15 +1175,19 @@ next:
 	return 0;
 }
 
-static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
-			     u32 *namelen, char **name, u64 *index,
-			     u64 *parent_objectid)
+static int extref_get_fields(struct extent_buffer *eb, int slot,
+			     unsigned long ref_ptr, u32 *namelen, char **name,
+			     u64 *index, u64 *parent_objectid)
 {
 	struct btrfs_inode_extref *extref;
 
 	extref = (struct btrfs_inode_extref *)ref_ptr;
 
 	*namelen = btrfs_inode_extref_name_len(eb, extref);
+	if (!btrfs_is_name_len_valid(eb, slot, (unsigned long)&extref->name,
+				     *namelen))
+		return -EIO;
+
 	*name = kmalloc(*namelen, GFP_NOFS);
 	if (*name == NULL)
 		return -ENOMEM;
@@ -1198,14 +1202,19 @@ static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
 	return 0;
 }
 
-static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
-			  u32 *namelen, char **name, u64 *index)
+static int ref_get_fields(struct extent_buffer *eb, int slot,
+			  unsigned long ref_ptr, u32 *namelen, char **name,
+			  u64 *index)
 {
 	struct btrfs_inode_ref *ref;
 
 	ref = (struct btrfs_inode_ref *)ref_ptr;
 
 	*namelen = btrfs_inode_ref_name_len(eb, ref);
+	if (!btrfs_is_name_len_valid(eb, slot, (unsigned long)(ref + 1),
+				     *namelen))
+		return -EIO;
+
 	*name = kmalloc(*namelen, GFP_NOFS);
 	if (*name == NULL)
 		return -ENOMEM;
@@ -1280,8 +1289,8 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 
 	while (ref_ptr < ref_end) {
 		if (log_ref_ver) {
-			ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
-						&ref_index, &parent_objectid);
+			ret = extref_get_fields(eb, slot, ref_ptr, &namelen,
+					  &name, &ref_index, &parent_objectid);
 			/*
 			 * parent object can change from one array
 			 * item to another.
@@ -1293,8 +1302,8 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
 				goto out;
 			}
 		} else {
-			ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
-					     &ref_index);
+			ret = ref_get_fields(eb, slot, ref_ptr, &namelen,
+					     &name, &ref_index);
 		}
 		if (ret)
 			goto out;
@@ -1841,7 +1850,7 @@ static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
 	ptr_end = ptr + item_size;
 	while (ptr < ptr_end) {
 		di = (struct btrfs_dir_item *)ptr;
-		if (verify_dir_item(fs_info, eb, di))
+		if (verify_dir_item(fs_info, eb, slot, di))
 			return -EIO;
 		name_len = btrfs_dir_name_len(eb, di);
 		ret = replay_one_name(trans, root, path, eb, di, key);
@@ -2017,7 +2026,7 @@ again:
 	ptr_end = ptr + item_size;
 	while (ptr < ptr_end) {
 		di = (struct btrfs_dir_item *)ptr;
-		if (verify_dir_item(fs_info, eb, di)) {
+		if (verify_dir_item(fs_info, eb, slot, di)) {
 			ret = -EIO;
 			goto out;
 		}
@@ -2102,6 +2111,7 @@ static int replay_xattr_deletes(struct btrfs_trans_handle *trans,
 			      struct btrfs_path *path,
 			      const u64 ino)
 {
+	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_key search_key;
 	struct btrfs_path *log_path;
 	int i;
@@ -2143,6 +2153,12 @@ process_leaf:
 			u32 this_len = sizeof(*di) + name_len + data_len;
 			char *name;
 
+			ret = verify_dir_item(fs_info, path->nodes[0],
+					      path->slots[0], di);
+			if (ret) {
+				ret = -EIO;
+				goto out;
+			}
 			name = kmalloc(name_len, GFP_NOFS);
 			if (!name) {
 				ret = -ENOMEM;
@@ -4546,6 +4562,12 @@ static int btrfs_check_ref_name_override(struct extent_buffer *eb,
 			this_len = sizeof(*extref) + this_name_len;
 		}
 
+		ret = btrfs_is_name_len_valid(eb, slot, name_ptr,
+					      this_name_len);
+		if (!ret) {
+			ret = -EIO;
+			goto out;
+		}
 		if (this_name_len > name_len) {
 			char *new_name;
 

fs/btrfs/volumes.c

@@ -242,6 +242,17 @@ static struct btrfs_device *__alloc_device(void)
 	if (!dev)
 		return ERR_PTR(-ENOMEM);
 
+	/*
+	 * Preallocate a bio that's always going to be used for flushing device
+	 * barriers and matches the device lifespan
+	 */
+	dev->flush_bio = bio_alloc_bioset(GFP_KERNEL, 0, NULL);
+	if (!dev->flush_bio) {
+		kfree(dev);
+		return ERR_PTR(-ENOMEM);
+	}
+	bio_get(dev->flush_bio);
+
 	INIT_LIST_HEAD(&dev->dev_list);
 	INIT_LIST_HEAD(&dev->dev_alloc_list);
 	INIT_LIST_HEAD(&dev->resized_list);
@@ -838,6 +849,7 @@ static void __free_device(struct work_struct *work)
 
 	device = container_of(work, struct btrfs_device, rcu_work);
 	rcu_string_free(device->name);
+	bio_put(device->flush_bio);
 	kfree(device);
 }
 
@@ -1353,15 +1365,13 @@ int find_free_dev_extent_start(struct btrfs_transaction *transaction,
 	int ret;
 	int slot;
 	struct extent_buffer *l;
-	u64 min_search_start;
 
 	/*
 	 * We don't want to overwrite the superblock on the drive nor any area
 	 * used by the boot loader (grub for example), so we make sure to start
 	 * at an offset of at least 1MB.
 	 */
-	min_search_start = max(fs_info->alloc_start, 1024ull * 1024);
-	search_start = max(search_start, min_search_start);
+	search_start = max_t(u64, search_start, SZ_1M);
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -2387,7 +2397,8 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	device->io_width = fs_info->sectorsize;
 	device->io_align = fs_info->sectorsize;
 	device->sector_size = fs_info->sectorsize;
-	device->total_bytes = i_size_read(bdev->bd_inode);
+	device->total_bytes = round_down(i_size_read(bdev->bd_inode),
+					 fs_info->sectorsize);
 	device->disk_total_bytes = device->total_bytes;
 	device->commit_total_bytes = device->total_bytes;
 	device->fs_info = fs_info;
@@ -2417,16 +2428,14 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	fs_info->fs_devices->total_devices++;
 	fs_info->fs_devices->total_rw_bytes += device->total_bytes;
 
-	spin_lock(&fs_info->free_chunk_lock);
-	fs_info->free_chunk_space += device->total_bytes;
-	spin_unlock(&fs_info->free_chunk_lock);
+	atomic64_add(device->total_bytes, &fs_info->free_chunk_space);
 
 	if (!blk_queue_nonrot(q))
 		fs_info->fs_devices->rotating = 1;
 
 	tmp = btrfs_super_total_bytes(fs_info->super_copy);
 	btrfs_set_super_total_bytes(fs_info->super_copy,
-				    tmp + device->total_bytes);
+		round_down(tmp + device->total_bytes, fs_info->sectorsize));
 
 	tmp = btrfs_super_num_devices(fs_info->super_copy);
 	btrfs_set_super_num_devices(fs_info->super_copy, tmp + 1);
@@ -2574,7 +2583,7 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 		goto error;
 	}
 
-	name = rcu_string_strdup(device_path, GFP_NOFS);
+	name = rcu_string_strdup(device_path, GFP_KERNEL);
 	if (!name) {
 		kfree(device);
 		ret = -ENOMEM;
@@ -2689,6 +2698,8 @@ int btrfs_grow_device(struct btrfs_trans_handle *trans,
 	if (!device->writeable)
 		return -EACCES;
 
+	new_size = round_down(new_size, fs_info->sectorsize);
+
 	mutex_lock(&fs_info->chunk_mutex);
 	old_total = btrfs_super_total_bytes(super_copy);
 	diff = new_size - device->total_bytes;
@@ -2701,7 +2712,8 @@ int btrfs_grow_device(struct btrfs_trans_handle *trans,
 
 	fs_devices = fs_info->fs_devices;
 
-	btrfs_set_super_total_bytes(super_copy, old_total + diff);
+	btrfs_set_super_total_bytes(super_copy,
+			round_down(old_total + diff, fs_info->sectorsize));
 	device->fs_devices->total_rw_bytes += diff;
 
 	btrfs_device_set_total_bytes(device, new_size);
@@ -2874,9 +2886,7 @@ int btrfs_remove_chunk(struct btrfs_trans_handle *trans,
 			mutex_lock(&fs_info->chunk_mutex);
 			btrfs_device_set_bytes_used(device,
 					device->bytes_used - dev_extent_len);
-			spin_lock(&fs_info->free_chunk_lock);
-			fs_info->free_chunk_space += dev_extent_len;
-			spin_unlock(&fs_info->free_chunk_lock);
+			atomic64_add(dev_extent_len, &fs_info->free_chunk_space);
 			btrfs_clear_space_info_full(fs_info);
 			mutex_unlock(&fs_info->chunk_mutex);
 		}
@@ -4393,7 +4403,10 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
 	struct btrfs_super_block *super_copy = fs_info->super_copy;
 	u64 old_total = btrfs_super_total_bytes(super_copy);
 	u64 old_size = btrfs_device_get_total_bytes(device);
-	u64 diff = old_size - new_size;
+	u64 diff;
+
+	new_size = round_down(new_size, fs_info->sectorsize);
+	diff = old_size - new_size;
 
 	if (device->is_tgtdev_for_dev_replace)
 		return -EINVAL;
@@ -4409,9 +4422,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size)
 	btrfs_device_set_total_bytes(device, new_size);
 	if (device->writeable) {
 		device->fs_devices->total_rw_bytes -= diff;
-		spin_lock(&fs_info->free_chunk_lock);
-		fs_info->free_chunk_space -= diff;
-		spin_unlock(&fs_info->free_chunk_lock);
+		atomic64_sub(diff, &fs_info->free_chunk_space);
 	}
 	mutex_unlock(&fs_info->chunk_mutex);
 
@@ -4522,7 +4533,8 @@ again:
 			      &fs_info->fs_devices->resized_devices);
 
 	WARN_ON(diff > old_total);
-	btrfs_set_super_total_bytes(super_copy, old_total - diff);
+	btrfs_set_super_total_bytes(super_copy,
+			round_down(old_total - diff, fs_info->sectorsize));
 	mutex_unlock(&fs_info->chunk_mutex);
 
 	/* Now btrfs_update_device() will change the on-disk size. */
@@ -4535,9 +4547,7 @@ done:
 		btrfs_device_set_total_bytes(device, old_size);
 		if (device->writeable)
 			device->fs_devices->total_rw_bytes += diff;
-		spin_lock(&fs_info->free_chunk_lock);
-		fs_info->free_chunk_space += diff;
-		spin_unlock(&fs_info->free_chunk_lock);
+		atomic64_add(diff, &fs_info->free_chunk_space);
 		mutex_unlock(&fs_info->chunk_mutex);
 	}
 	return ret;
@@ -4882,9 +4892,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 		btrfs_device_set_bytes_used(map->stripes[i].dev, num_bytes);
 	}
 
-	spin_lock(&info->free_chunk_lock);
-	info->free_chunk_space -= (stripe_size * map->num_stripes);
-	spin_unlock(&info->free_chunk_lock);
+	atomic64_sub(stripe_size * map->num_stripes, &info->free_chunk_space);
 
 	free_extent_map(em);
 	check_raid56_incompat_flag(info, type);
@@ -5029,20 +5037,19 @@ int btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 static noinline int init_first_rw_device(struct btrfs_trans_handle *trans,
 					 struct btrfs_fs_info *fs_info)
 {
-	struct btrfs_root *extent_root = fs_info->extent_root;
 	u64 chunk_offset;
 	u64 sys_chunk_offset;
 	u64 alloc_profile;
 	int ret;
 
 	chunk_offset = find_next_chunk(fs_info);
-	alloc_profile = btrfs_get_alloc_profile(extent_root, 0);
+	alloc_profile = btrfs_metadata_alloc_profile(fs_info);
 	ret = __btrfs_alloc_chunk(trans, chunk_offset, alloc_profile);
 	if (ret)
 		return ret;
 
 	sys_chunk_offset = find_next_chunk(fs_info);
-	alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
+	alloc_profile = btrfs_system_alloc_profile(fs_info);
 	ret = __btrfs_alloc_chunk(trans, sys_chunk_offset, alloc_profile);
 	return ret;
 }
@@ -6267,10 +6274,9 @@ int btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
 			continue;
 		}
 
-		if (dev_nr < total_devs - 1) {
-			bio = btrfs_bio_clone(first_bio, GFP_NOFS);
-			BUG_ON(!bio); /* -ENOMEM */
-		} else
+		if (dev_nr < total_devs - 1)
+			bio = btrfs_bio_clone(first_bio);
+		else
 			bio = first_bio;
 
 		submit_stripe_bio(bbio, bio, bbio->stripes[dev_nr].physical,
@@ -6685,10 +6691,8 @@ static int read_one_dev(struct btrfs_fs_info *fs_info,
 	device->in_fs_metadata = 1;
 	if (device->writeable && !device->is_tgtdev_for_dev_replace) {
 		device->fs_devices->total_rw_bytes += device->total_bytes;
-		spin_lock(&fs_info->free_chunk_lock);
-		fs_info->free_chunk_space += device->total_bytes -
-			device->bytes_used;
-		spin_unlock(&fs_info->free_chunk_lock);
+		atomic64_add(device->total_bytes - device->bytes_used,
+				&fs_info->free_chunk_space);
 	}
 	ret = 0;
 	return ret;

fs/btrfs/volumes.h

@@ -74,6 +74,8 @@ struct btrfs_device {
 	int missing;
 	int can_discard;
 	int is_tgtdev_for_dev_replace;
+	int last_flush_error;
+	int flush_bio_sent;
 
 #ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED
 	seqcount_t data_seqcount;
@@ -279,6 +281,11 @@ struct btrfs_io_bio {
 	u8 csum_inline[BTRFS_BIO_INLINE_CSUM_SIZE];
 	u8 *csum_allocated;
 	btrfs_io_bio_end_io_t *end_io;
+	struct bvec_iter iter;
+	/*
+	 * This member must come last, bio_alloc_bioset will allocate enough
+	 * bytes for entire btrfs_io_bio but relies on bio being last.
+	 */
 	struct bio bio;
 };
 

fs/btrfs/xattr.c

@@ -336,7 +336,7 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size)
 			u32 this_len = sizeof(*di) + name_len + data_len;
 			unsigned long name_ptr = (unsigned long)(di + 1);
 
-			if (verify_dir_item(fs_info, leaf, di)) {
+			if (verify_dir_item(fs_info, leaf, slot, di)) {
 				ret = -EIO;
 				goto err;
 			}

fs/btrfs/zlib.c

@@ -24,12 +24,13 @@
 #include <linux/slab.h>
 #include <linux/zlib.h>
 #include <linux/zutil.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <linux/init.h>
 #include <linux/err.h>
 #include <linux/sched.h>
 #include <linux/pagemap.h>
 #include <linux/bio.h>
+#include <linux/refcount.h>
 #include "compression.h"
 
 struct workspace {
@@ -42,7 +43,7 @@ static void zlib_free_workspace(struct list_head *ws)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
 
-	vfree(workspace->strm.workspace);
+	kvfree(workspace->strm.workspace);
 	kfree(workspace->buf);
 	kfree(workspace);
 }
@@ -52,14 +53,14 @@ static struct list_head *zlib_alloc_workspace(void)
 	struct workspace *workspace;
 	int workspacesize;
 
-	workspace = kzalloc(sizeof(*workspace), GFP_NOFS);
+	workspace = kzalloc(sizeof(*workspace), GFP_KERNEL);
 	if (!workspace)
 		return ERR_PTR(-ENOMEM);
 
 	workspacesize = max(zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL),
 			zlib_inflate_workspacesize());
-	workspace->strm.workspace = vmalloc(workspacesize);
-	workspace->buf = kmalloc(PAGE_SIZE, GFP_NOFS);
+	workspace->strm.workspace = kvmalloc(workspacesize, GFP_KERNEL);
+	workspace->buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
 	if (!workspace->strm.workspace || !workspace->buf)
 		goto fail;
 
@@ -211,10 +212,7 @@ out:
 	return ret;
 }
 
-static int zlib_decompress_bio(struct list_head *ws, struct page **pages_in,
-				  u64 disk_start,
-				  struct bio *orig_bio,
-				  size_t srclen)
+static int zlib_decompress_bio(struct list_head *ws, struct compressed_bio *cb)
 {
 	struct workspace *workspace = list_entry(ws, struct workspace, list);
 	int ret = 0, ret2;
@@ -222,8 +220,12 @@ static int zlib_decompress_bio(struct list_head *ws, struct page **pages_in,
 	char *data_in;
 	size_t total_out = 0;
 	unsigned long page_in_index = 0;
+	size_t srclen = cb->compressed_len;
 	unsigned long total_pages_in = DIV_ROUND_UP(srclen, PAGE_SIZE);
 	unsigned long buf_start;
+	struct page **pages_in = cb->compressed_pages;
+	u64 disk_start = cb->start;
+	struct bio *orig_bio = cb->orig_bio;
 
 	data_in = kmap(pages_in[page_in_index]);
 	workspace->strm.next_in = data_in;

include/trace/events/btrfs.h

@@ -1410,42 +1410,6 @@ DEFINE_EVENT(btrfs__workqueue_done, btrfs_workqueue_destroy,
 	TP_ARGS(wq)
 );
 
-DECLARE_EVENT_CLASS(btrfs__qgroup_data_map,
-
-	TP_PROTO(struct inode *inode, u64 free_reserved),
-
-	TP_ARGS(inode, free_reserved),
-
-	TP_STRUCT__entry_btrfs(
-		__field(	u64,		rootid		)
-		__field(	unsigned long,	ino		)
-		__field(	u64,		free_reserved	)
-	),
-
-	TP_fast_assign_btrfs(btrfs_sb(inode->i_sb),
-		__entry->rootid		=	BTRFS_I(inode)->root->objectid;
-		__entry->ino		=	inode->i_ino;
-		__entry->free_reserved	=	free_reserved;
-	),
-
-	TP_printk_btrfs("rootid=%llu ino=%lu free_reserved=%llu",
-		  __entry->rootid, __entry->ino, __entry->free_reserved)
-);
-
-DEFINE_EVENT(btrfs__qgroup_data_map, btrfs_qgroup_init_data_rsv_map,
-
-	TP_PROTO(struct inode *inode, u64 free_reserved),
-
-	TP_ARGS(inode, free_reserved)
-);
-
-DEFINE_EVENT(btrfs__qgroup_data_map, btrfs_qgroup_free_data_rsv_map,
-
-	TP_PROTO(struct inode *inode, u64 free_reserved),
-
-	TP_ARGS(inode, free_reserved)
-);
-
 #define BTRFS_QGROUP_OPERATIONS				\
 	{ QGROUP_RESERVE,	"reserve"	},	\
 	{ QGROUP_RELEASE,	"release"	},	\

include/uapi/linux/btrfs.h

@@ -426,31 +426,54 @@ struct btrfs_ioctl_ino_lookup_args {
 	char name[BTRFS_INO_LOOKUP_PATH_MAX];
 };
 
+/* Search criteria for the btrfs SEARCH ioctl family. */
 struct btrfs_ioctl_search_key {
-	/* which root are we searching.  0 is the tree of tree roots */
-	__u64 tree_id;
-
-	/* keys returned will be >= min and <= max */
-	__u64 min_objectid;
-	__u64 max_objectid;
-
-	/* keys returned will be >= min and <= max */
-	__u64 min_offset;
-	__u64 max_offset;
-
-	/* max and min transids to search for */
-	__u64 min_transid;
-	__u64 max_transid;
+	/*
+	 * The tree we're searching in. 1 is the tree of tree roots, 2 is the
+	 * extent tree, etc...
+	 *
+	 * A special tree_id value of 0 will cause a search in the subvolume
+	 * tree that the inode which is passed to the ioctl is part of.
+	 */
+	__u64 tree_id;		/* in */
 
-	/* keys returned will be >= min and <= max */
-	__u32 min_type;
-	__u32 max_type;
+	/*
+	 * When doing a tree search, we're actually taking a slice from a
+	 * linear search space of 136-bit keys.
+	 *
+	 * A full 136-bit tree key is composed as:
+	 *   (objectid << 72) + (type << 64) + offset
+	 *
+	 * The individual min and max values for objectid, type and offset
+	 * define the min_key and max_key values for the search range. All
+	 * metadata items with a key in the interval [min_key, max_key] will be
+	 * returned.
+	 *
+	 * Additionally, we can filter the items returned on transaction id of
+	 * the metadata block they're stored in by specifying a transid range.
+	 * Be aware that this transaction id only denotes when the metadata
+	 * page that currently contains the item got written the last time as
+	 * result of a COW operation.  The number does not have any meaning
+	 * related to the transaction in which an individual item that is being
+	 * returned was created or changed.
+	 */
+	__u64 min_objectid;	/* in */
+	__u64 max_objectid;	/* in */
+	__u64 min_offset;	/* in */
+	__u64 max_offset;	/* in */
+	__u64 min_transid;	/* in */
+	__u64 max_transid;	/* in */
+	__u32 min_type;		/* in */
+	__u32 max_type;		/* in */
 
 	/*
-	 * how many items did userland ask for, and how many are we
-	 * returning
+	 * input: The maximum amount of results desired.
+	 * output: The actual amount of items returned, restricted by any of:
+	 *  - reaching the upper bound of the search range
+	 *  - reaching the input nr_items amount of items
+	 *  - completely filling the supplied memory buffer
 	 */
-	__u32 nr_items;
+	__u32 nr_items;		/* in/out */
 
 	/* align to 64 bits */
 	__u32 unused;