1 files changed, 50 insertions, 49 deletions

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 70fc4e7cc5a0..9247a58894de 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -116,7 +116,7 @@ static void csum_tree_block(struct extent_buffer *buf, u8 *result)
  * detect blocks that either didn't get written at all or got written
  * in the wrong place.
  */
-int btrfs_buffer_uptodate(struct extent_buffer *eb, u64 parent_transid, int atomic)
+int btrfs_buffer_uptodate(struct extent_buffer *eb, u64 parent_transid, bool atomic)
 {
 	if (!extent_buffer_uptodate(eb))
 		return 0;
@@ -370,21 +370,21 @@ int btrfs_validate_extent_buffer(struct extent_buffer *eb,
 	ASSERT(check);
 
 	found_start = btrfs_header_bytenr(eb);
-	if (found_start != eb->start) {
+	if (unlikely(found_start != eb->start)) {
 		btrfs_err_rl(fs_info,
 			"bad tree block start, mirror %u want %llu have %llu",
 			     eb->read_mirror, eb->start, found_start);
 		ret = -EIO;
 		goto out;
 	}
-	if (check_tree_block_fsid(eb)) {
+	if (unlikely(check_tree_block_fsid(eb))) {
 		btrfs_err_rl(fs_info, "bad fsid on logical %llu mirror %u",
 			     eb->start, eb->read_mirror);
 		ret = -EIO;
 		goto out;
 	}
 	found_level = btrfs_header_level(eb);
-	if (found_level >= BTRFS_MAX_LEVEL) {
+	if (unlikely(found_level >= BTRFS_MAX_LEVEL)) {
 		btrfs_err(fs_info,
 			"bad tree block level, mirror %u level %d on logical %llu",
 			eb->read_mirror, btrfs_header_level(eb), eb->start);
@@ -404,13 +404,13 @@ int btrfs_validate_extent_buffer(struct extent_buffer *eb,
 			      CSUM_FMT_VALUE(csum_size, result),
 			      btrfs_header_level(eb),
 			      ignore_csum ? ", ignored" : "");
-		if (!ignore_csum) {
+		if (unlikely(!ignore_csum)) {
 			ret = -EUCLEAN;
 			goto out;
 		}
 	}
 
-	if (found_level != check->level) {
+	if (unlikely(found_level != check->level)) {
 		btrfs_err(fs_info,
 		"level verify failed on logical %llu mirror %u wanted %u found %u",
 			  eb->start, eb->read_mirror, check->level, found_level);
@@ -639,7 +639,6 @@ static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info,
 					   u64 objectid, gfp_t flags)
 {
 	struct btrfs_root *root;
-	bool dummy = btrfs_is_testing(fs_info);
 
 	root = kzalloc(sizeof(*root), flags);
 	if (!root)
@@ -696,7 +695,7 @@ static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info,
 	root->log_transid_committed = -1;
 	btrfs_set_root_last_log_commit(root, 0);
 	root->anon_dev = 0;
-	if (!dummy) {
+	if (!btrfs_is_testing(fs_info)) {
 		btrfs_extent_io_tree_init(fs_info, &root->dirty_log_pages,
 					  IO_TREE_ROOT_DIRTY_LOG_PAGES);
 		btrfs_extent_io_tree_init(fs_info, &root->log_csum_range,
@@ -1047,7 +1046,7 @@ static struct btrfs_root *read_tree_root_path(struct btrfs_root *tree_root,
 		root->node = NULL;
 		goto fail;
 	}
-	if (!btrfs_buffer_uptodate(root->node, generation, 0)) {
+	if (unlikely(!btrfs_buffer_uptodate(root->node, generation, false))) {
 		ret = -EIO;
 		goto fail;
 	}
@@ -1056,10 +1055,10 @@ static struct btrfs_root *read_tree_root_path(struct btrfs_root *tree_root,
 	 * For real fs, and not log/reloc trees, root owner must
 	 * match its root node owner
 	 */
-	if (!btrfs_is_testing(fs_info) &&
-	    btrfs_root_id(root) != BTRFS_TREE_LOG_OBJECTID &&
-	    btrfs_root_id(root) != BTRFS_TREE_RELOC_OBJECTID &&
-	    btrfs_root_id(root) != btrfs_header_owner(root->node)) {
+	if (unlikely(!btrfs_is_testing(fs_info) &&
+		     btrfs_root_id(root) != BTRFS_TREE_LOG_OBJECTID &&
+		     btrfs_root_id(root) != BTRFS_TREE_RELOC_OBJECTID &&
+		     btrfs_root_id(root) != btrfs_header_owner(root->node))) {
 		btrfs_crit(fs_info,
 "root=%llu block=%llu, tree root owner mismatch, have %llu expect %llu",
 			   btrfs_root_id(root), root->node->start,
@@ -1248,6 +1247,7 @@ void btrfs_free_fs_info(struct btrfs_fs_info *fs_info)
 
 	if (fs_info->fs_devices)
 		btrfs_close_devices(fs_info->fs_devices);
+	btrfs_free_compress_wsm(fs_info);
 	percpu_counter_destroy(&fs_info->stats_read_blocks);
 	percpu_counter_destroy(&fs_info->dirty_metadata_bytes);
 	percpu_counter_destroy(&fs_info->delalloc_bytes);
@@ -1958,7 +1958,7 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info)
 {
 	u32 max_active = fs_info->thread_pool_size;
 	unsigned int flags = WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_UNBOUND;
-	unsigned int ordered_flags = WQ_MEM_RECLAIM | WQ_FREEZABLE;
+	unsigned int ordered_flags = WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_PERCPU;
 
 	fs_info->workers =
 		btrfs_alloc_workqueue(fs_info, "worker", flags, max_active, 16);
@@ -2058,7 +2058,7 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 	u64 bytenr = btrfs_super_log_root(disk_super);
 	int level = btrfs_super_log_root_level(disk_super);
 
-	if (fs_devices->rw_devices == 0) {
+	if (unlikely(fs_devices->rw_devices == 0)) {
 		btrfs_warn(fs_info, "log replay required on RO media");
 		return -EIO;
 	}
@@ -2079,7 +2079,7 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 		btrfs_put_root(log_tree_root);
 		return ret;
 	}
-	if (!extent_buffer_uptodate(log_tree_root->node)) {
+	if (unlikely(!extent_buffer_uptodate(log_tree_root->node))) {
 		btrfs_err(fs_info, "failed to read log tree");
 		btrfs_put_root(log_tree_root);
 		return -EIO;
@@ -2087,10 +2087,10 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info,
 
 	/* returns with log_tree_root freed on success */
 	ret = btrfs_recover_log_trees(log_tree_root);
+	btrfs_put_root(log_tree_root);
 	if (ret) {
 		btrfs_handle_fs_error(fs_info, ret,
 				      "Failed to recover log tree");
-		btrfs_put_root(log_tree_root);
 		return ret;
 	}
 
@@ -2324,7 +2324,7 @@ static int validate_sys_chunk_array(const struct btrfs_fs_info *fs_info,
 	const u32 sectorsize = btrfs_super_sectorsize(sb);
 	u32 sys_array_size = btrfs_super_sys_array_size(sb);
 
-	if (sys_array_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) {
+	if (unlikely(sys_array_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE)) {
 		btrfs_err(fs_info, "system chunk array too big %u > %u",
 			  sys_array_size, BTRFS_SYSTEM_CHUNK_ARRAY_SIZE);
 		return -EUCLEAN;
@@ -2342,12 +2342,12 @@ static int validate_sys_chunk_array(const struct btrfs_fs_info *fs_info,
 		disk_key = (struct btrfs_disk_key *)(sb->sys_chunk_array + cur);
 		len = sizeof(*disk_key);
 
-		if (cur + len > sys_array_size)
+		if (unlikely(cur + len > sys_array_size))
 			goto short_read;
 		cur += len;
 
 		btrfs_disk_key_to_cpu(&key, disk_key);
-		if (key.type != BTRFS_CHUNK_ITEM_KEY) {
+		if (unlikely(key.type != BTRFS_CHUNK_ITEM_KEY)) {
 			btrfs_err(fs_info,
 			    "unexpected item type %u in sys_array at offset %u",
 				  key.type, cur);
@@ -2355,10 +2355,10 @@ static int validate_sys_chunk_array(const struct btrfs_fs_info *fs_info,
 		}
 		chunk = (struct btrfs_chunk *)(sb->sys_chunk_array + cur);
 		num_stripes = btrfs_stack_chunk_num_stripes(chunk);
-		if (cur + btrfs_chunk_item_size(num_stripes) > sys_array_size)
+		if (unlikely(cur + btrfs_chunk_item_size(num_stripes) > sys_array_size))
 			goto short_read;
 		type = btrfs_stack_chunk_type(chunk);
-		if (!(type & BTRFS_BLOCK_GROUP_SYSTEM)) {
+		if (unlikely(!(type & BTRFS_BLOCK_GROUP_SYSTEM))) {
 			btrfs_err(fs_info,
 			"invalid chunk type %llu in sys_array at offset %u",
 				  type, cur);
@@ -2438,21 +2438,7 @@ int btrfs_validate_super(const struct btrfs_fs_info *fs_info,
 		ret = -EINVAL;
 	}
 
-	/*
-	 * We only support at most 3 sectorsizes: 4K, PAGE_SIZE, MIN_BLOCKSIZE.
-	 *
-	 * For 4K page sized systems with non-debug builds, all 3 matches (4K).
-	 * For 4K page sized systems with debug builds, there are two block sizes
-	 * supported. (4K and 2K)
-	 *
-	 * We can support 16K sectorsize with 64K page size without problem,
-	 * but such sectorsize/pagesize combination doesn't make much sense.
-	 * 4K will be our future standard, PAGE_SIZE is supported from the very
-	 * beginning.
-	 */
-	if (sectorsize > PAGE_SIZE || (sectorsize != SZ_4K &&
-				       sectorsize != PAGE_SIZE &&
-				       sectorsize != BTRFS_MIN_BLOCKSIZE)) {
+	if (!btrfs_supported_blocksize(sectorsize)) {
 		btrfs_err(fs_info,
 			"sectorsize %llu not yet supported for page size %lu",
 			sectorsize, PAGE_SIZE);
@@ -2619,13 +2605,13 @@ static int btrfs_validate_write_super(struct btrfs_fs_info *fs_info,
 	ret = btrfs_validate_super(fs_info, sb, -1);
 	if (ret < 0)
 		goto out;
-	if (!btrfs_supported_super_csum(btrfs_super_csum_type(sb))) {
+	if (unlikely(!btrfs_supported_super_csum(btrfs_super_csum_type(sb)))) {
 		ret = -EUCLEAN;
 		btrfs_err(fs_info, "invalid csum type, has %u want %u",
 			  btrfs_super_csum_type(sb), BTRFS_CSUM_TYPE_CRC32);
 		goto out;
 	}
-	if (btrfs_super_incompat_flags(sb) & ~BTRFS_FEATURE_INCOMPAT_SUPP) {
+	if (unlikely(btrfs_super_incompat_flags(sb) & ~BTRFS_FEATURE_INCOMPAT_SUPP)) {
 		ret = -EUCLEAN;
 		btrfs_err(fs_info,
 		"invalid incompat flags, has 0x%llx valid mask 0x%llx",
@@ -2655,7 +2641,7 @@ static int load_super_root(struct btrfs_root *root, u64 bytenr, u64 gen, int lev
 		root->node = NULL;
 		return ret;
 	}
-	if (!extent_buffer_uptodate(root->node)) {
+	if (unlikely(!extent_buffer_uptodate(root->node))) {
 		free_extent_buffer(root->node);
 		root->node = NULL;
 		return -EIO;
@@ -3256,18 +3242,24 @@ int btrfs_check_features(struct btrfs_fs_info *fs_info, bool is_rw_mount)
 	}
 
 	/*
-	 * Subpage runtime limitation on v1 cache.
+	 * Subpage/bs > ps runtime limitation on v1 cache.
 	 *
-	 * V1 space cache still has some hard codeed PAGE_SIZE usage, while
+	 * V1 space cache still has some hard coded PAGE_SIZE usage, while
 	 * we're already defaulting to v2 cache, no need to bother v1 as it's
 	 * going to be deprecated anyway.
 	 */
-	if (fs_info->sectorsize < PAGE_SIZE && btrfs_test_opt(fs_info, SPACE_CACHE)) {
+	if (fs_info->sectorsize != PAGE_SIZE && btrfs_test_opt(fs_info, SPACE_CACHE)) {
 		btrfs_warn(fs_info,
 	"v1 space cache is not supported for page size %lu with sectorsize %u",
 			   PAGE_SIZE, fs_info->sectorsize);
 		return -EINVAL;
 	}
+	if (fs_info->sectorsize > PAGE_SIZE && btrfs_fs_incompat(fs_info, RAID56)) {
+		btrfs_err(fs_info,
+		"RAID56 is not supported for page size %lu with sectorsize %u",
+			  PAGE_SIZE, fs_info->sectorsize);
+		return -EINVAL;
+	}
 
 	/* This can be called by remount, we need to protect the super block. */
 	spin_lock(&fs_info->super_lock);
@@ -3396,10 +3388,16 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 	fs_info->nodesize_bits = ilog2(nodesize);
 	fs_info->sectorsize = sectorsize;
 	fs_info->sectorsize_bits = ilog2(sectorsize);
+	fs_info->block_min_order = ilog2(round_up(sectorsize, PAGE_SIZE) >> PAGE_SHIFT);
+	fs_info->block_max_order = ilog2((BITS_PER_LONG << fs_info->sectorsize_bits) >> PAGE_SHIFT);
 	fs_info->csums_per_leaf = BTRFS_MAX_ITEM_SIZE(fs_info) / fs_info->csum_size;
 	fs_info->stripesize = stripesize;
 	fs_info->fs_devices->fs_info = fs_info;
 
+	if (fs_info->sectorsize > PAGE_SIZE)
+		btrfs_warn(fs_info,
+			   "support for block size %u with page size %zu is experimental, some features may be missing",
+			   fs_info->sectorsize, PAGE_SIZE);
 	/*
 	 * Handle the space caching options appropriately now that we have the
 	 * super block loaded and validated.
@@ -3421,6 +3419,9 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 	 */
 	fs_info->max_inline = min_t(u64, fs_info->max_inline, fs_info->sectorsize);
 
+	ret = btrfs_alloc_compress_wsm(fs_info);
+	if (ret)
+		goto fail_sb_buffer;
 	ret = btrfs_init_workqueues(fs_info);
 	if (ret)
 		goto fail_sb_buffer;
@@ -3468,7 +3469,7 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 	 * below in btrfs_init_dev_replace().
 	 */
 	btrfs_free_extra_devids(fs_devices);
-	if (!fs_devices->latest_dev->bdev) {
+	if (unlikely(!fs_devices->latest_dev->bdev)) {
 		btrfs_err(fs_info, "failed to read devices");
 		ret = -EIO;
 		goto fail_tree_roots;
@@ -3962,7 +3963,7 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 	 * Checks last_flush_error of disks in order to determine the device
 	 * state.
 	 */
-	if (errors_wait && !btrfs_check_rw_degradable(info, NULL))
+	if (unlikely(errors_wait && !btrfs_check_rw_degradable(info, NULL)))
 		return -EIO;
 
 	return 0;
@@ -4064,7 +4065,7 @@ int write_all_supers(struct btrfs_fs_info *fs_info, int max_mirrors)
 		btrfs_set_super_flags(sb, flags | BTRFS_HEADER_FLAG_WRITTEN);
 
 		ret = btrfs_validate_write_super(fs_info, sb);
-		if (ret < 0) {
+		if (unlikely(ret < 0)) {
 			mutex_unlock(&fs_info->fs_devices->device_list_mutex);
 			btrfs_handle_fs_error(fs_info, -EUCLEAN,
 				"unexpected superblock corruption detected");
@@ -4075,7 +4076,7 @@ int write_all_supers(struct btrfs_fs_info *fs_info, int max_mirrors)
 		if (ret)
 			total_errors++;
 	}
-	if (total_errors > max_errors) {
+	if (unlikely(total_errors > max_errors)) {
 		btrfs_err(fs_info, "%d errors while writing supers",
 			  total_errors);
 		mutex_unlock(&fs_info->fs_devices->device_list_mutex);
@@ -4100,7 +4101,7 @@ int write_all_supers(struct btrfs_fs_info *fs_info, int max_mirrors)
 			total_errors++;
 	}
 	mutex_unlock(&fs_info->fs_devices->device_list_mutex);
-	if (total_errors > max_errors) {
+	if (unlikely(total_errors > max_errors)) {
 		btrfs_handle_fs_error(fs_info, -EIO,
 				      "%d errors while writing supers",
 				      total_errors);
@@ -4880,7 +4881,7 @@ int btrfs_init_root_free_objectid(struct btrfs_root *root)
 	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
 	if (ret < 0)
 		return ret;
-	if (ret == 0) {
+	if (unlikely(ret == 0)) {
 		/*
 		 * Key with offset -1 found, there would have to exist a root
 		 * with such id, but this is out of valid range.