Merge tag 'for-6.2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs updates from David Sterba:
 "This round there are a lot of cleanups and moved code so the diffstat
  looks huge, otherwise there are some nice performance improvements and
  an update to raid56 reliability.

  User visible features:

   - raid56 reliability vs performance trade off:
      - fix destructive RMW for raid5 data (raid6 still needs work): do
        full checksum verification for all data during RMW cycle, this
        should prevent rewriting potentially corrupted data without
        notice
      - stripes are cached in memory which should reduce the performance
        impact but still can hurt some workloads
      - checksums are verified after repair again
      - this is the last option without introducing additional features
        (write intent bitmap, journal, another tree), the extra checksum
        read/verification was supposed to be avoided by the original
        implementation exactly for performance reasons but that caused
        all the reliability problems

   - discard=async by default for devices that support it

   - implement emergency flush reserve to avoid almost all unnecessary
     transaction aborts due to ENOSPC in cases where there are too many
     delayed refs or delayed allocation

   - skip block group synchronization if there's no change in used
     bytes, can reduce transaction commit count for some workloads

  Performance improvements:

   - fiemap and lseek:
      - overall speedup due to skipping unnecessary or duplicate
        searches (-40% run time)
      - cache some data structures and sharedness of extents (-30% run
        time)

   - send:
      - faster backref resolution when finding clones
      - cached leaf to root mapping for faster backref walking
      - improved clone/sharing detection
      - overall run time improvements (-70%)

  Core:

   - module initialization converted to a table of function pointers run
     in a sequence

   - preparation for fscrypt, extend passing file names across calls,
     dir item can store encryption status

   - raid56 updates:
      - more accurate error tracking of sectors within stripe
      - simplify recovery path and remove dedicated endio worker kthread
      - simplify scrub call paths
      - refactoring to support the extra data checksum verification
        during RMW cycle

   - tree block parentness checks consolidated and done at metadata read
     time

   - improved error handling

   - cleanups:
      - move a lot of code for better synchronization between kernel and
        user space sources, split big files
      - enum cleanups
      - GFP flag cleanups
      - header file cleanups, prototypes, dependencies
      - redundant parameter cleanups
      - inline extent handling simplifications
      - inode parameter conversion
      - data structure cleanups, reductions, renames, merges"

* tag 'for-6.2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux: (249 commits)
  btrfs: print transaction aborted messages with an error level
  btrfs: sync some cleanups from progs into uapi/btrfs.h
  btrfs: do not BUG_ON() on ENOMEM when dropping extent items for a range
  btrfs: fix extent map use-after-free when handling missing device in read_one_chunk
  btrfs: remove outdated logic from overwrite_item() and add assertion
  btrfs: unify overwrite_item() and do_overwrite_item()
  btrfs: replace strncpy() with strscpy()
  btrfs: fix uninitialized variable in find_first_clear_extent_bit
  btrfs: fix uninitialized parent in insert_state
  btrfs: add might_sleep() annotations
  btrfs: add stack helpers for a few btrfs items
  btrfs: add nr_global_roots to the super block definition
  btrfs: remove BTRFS_LEAF_DATA_OFFSET
  btrfs: add helpers for manipulating leaf items and data
  btrfs: add eb to btrfs_node_key_ptr_offset
  btrfs: pass the extent buffer for the btrfs_item_nr helpers
  btrfs: move the csum helpers into ctree.h
  btrfs: move eb offset helpers into extent_io.h
  btrfs: move file_extent_item helpers into file-item.h
  btrfs: move leaf_data_end into ctree.c
  ...

2 files changed, 265 insertions, 6 deletions

diff --git a/include/uapi/linux/btrfs.h b/include/uapi/linux/btrfs.h
index 5655e89b962b..b4f0f9531119 100644
--- a/include/uapi/linux/btrfs.h
+++ b/include/uapi/linux/btrfs.h
@@ -19,8 +19,14 @@
 
 #ifndef _UAPI_LINUX_BTRFS_H
 #define _UAPI_LINUX_BTRFS_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 #include <linux/types.h>
 #include <linux/ioctl.h>
+#include <linux/fs.h>
 
 #define BTRFS_IOCTL_MAGIC 0x94
 #define BTRFS_VOL_NAME_MAX 255
@@ -333,6 +339,12 @@ struct btrfs_ioctl_feature_flags {
  */
 struct btrfs_balance_args {
 	__u64 profiles;
+
+	/*
+	 * usage filter
+	 * BTRFS_BALANCE_ARGS_USAGE with a single value means '0..N'
+	 * BTRFS_BALANCE_ARGS_USAGE_RANGE - range syntax, min..max
+	 */
 	union {
 		__u64 usage;
 		struct {
@@ -549,7 +561,7 @@ struct btrfs_ioctl_search_header {
 	__u64 offset;
 	__u32 type;
 	__u32 len;
-};
+} __attribute__ ((__may_alias__));
 
 #define BTRFS_SEARCH_ARGS_BUFSIZE (4096 - sizeof(struct btrfs_ioctl_search_key))
 /*
@@ -562,6 +574,10 @@ struct btrfs_ioctl_search_args {
 	char buf[BTRFS_SEARCH_ARGS_BUFSIZE];
 };
 
+/*
+ * Extended version of TREE_SEARCH ioctl that can return more than 4k of bytes.
+ * The allocated size of the buffer is set in buf_size.
+ */
 struct btrfs_ioctl_search_args_v2 {
 	struct btrfs_ioctl_search_key key; /* in/out - search parameters */
 	__u64 buf_size;		   /* in - size of buffer
@@ -570,10 +586,11 @@ struct btrfs_ioctl_search_args_v2 {
 	__u64 buf[];                       /* out - found items */
 };
 
+/* With a @src_length of zero, the range from @src_offset->EOF is cloned! */
 struct btrfs_ioctl_clone_range_args {
-  __s64 src_fd;
-  __u64 src_offset, src_length;
-  __u64 dest_offset;
+	__s64 src_fd;
+	__u64 src_offset, src_length;
+	__u64 dest_offset;
 };
 
 /*
@@ -677,8 +694,11 @@ struct btrfs_ioctl_logical_ino_args {
 	/* struct btrfs_data_container	*inodes;	out   */
 	__u64				inodes;
 };
-/* Return every ref to the extent, not just those containing logical block.
- * Requires logical == extent bytenr. */
+
+/*
+ * Return every ref to the extent, not just those containing logical block.
+ * Requires logical == extent bytenr.
+ */
 #define BTRFS_LOGICAL_INO_ARGS_IGNORE_OFFSET	(1ULL << 0)
 
 enum btrfs_dev_stat_values {
@@ -1144,4 +1164,8 @@ enum btrfs_err_code {
 #define BTRFS_IOC_ENCODED_WRITE _IOW(BTRFS_IOCTL_MAGIC, 64, \
 				     struct btrfs_ioctl_encoded_io_args)
 
+#ifdef __cplusplus
+}
+#endif
+
 #endif /* _UAPI_LINUX_BTRFS_H */
diff --git a/include/uapi/linux/btrfs_tree.h b/include/uapi/linux/btrfs_tree.h
index 1f7a38ec6ac3..ab38d0f411fa 100644
--- a/include/uapi/linux/btrfs_tree.h
+++ b/include/uapi/linux/btrfs_tree.h
@@ -10,6 +10,23 @@
 #include <stddef.h>
 #endif
 
+/* ASCII for _BHRfS_M, no terminating nul */
+#define BTRFS_MAGIC 0x4D5F53665248425FULL
+
+#define BTRFS_MAX_LEVEL 8
+
+/*
+ * We can actually store much bigger names, but lets not confuse the rest of
+ * linux.
+ */
+#define BTRFS_NAME_LEN 255
+
+/*
+ * Theoretical limit is larger, but we keep this down to a sane value. That
+ * should limit greatly the possibility of collisions on inode ref items.
+ */
+#define BTRFS_LINK_MAX 65535U
+
 /*
  * This header contains the structure definitions and constants used
  * by file system objects that can be retrieved using
@@ -359,6 +376,50 @@ enum btrfs_csum_type {
 #define BTRFS_FT_SYMLINK	7
 #define BTRFS_FT_XATTR		8
 #define BTRFS_FT_MAX		9
+/* Directory contains encrypted data */
+#define BTRFS_FT_ENCRYPTED	0x80
+
+static inline __u8 btrfs_dir_flags_to_ftype(__u8 flags)
+{
+	return flags & ~BTRFS_FT_ENCRYPTED;
+}
+
+/*
+ * Inode flags
+ */
+#define BTRFS_INODE_NODATASUM		(1U << 0)
+#define BTRFS_INODE_NODATACOW		(1U << 1)
+#define BTRFS_INODE_READONLY		(1U << 2)
+#define BTRFS_INODE_NOCOMPRESS		(1U << 3)
+#define BTRFS_INODE_PREALLOC		(1U << 4)
+#define BTRFS_INODE_SYNC		(1U << 5)
+#define BTRFS_INODE_IMMUTABLE		(1U << 6)
+#define BTRFS_INODE_APPEND		(1U << 7)
+#define BTRFS_INODE_NODUMP		(1U << 8)
+#define BTRFS_INODE_NOATIME		(1U << 9)
+#define BTRFS_INODE_DIRSYNC		(1U << 10)
+#define BTRFS_INODE_COMPRESS		(1U << 11)
+
+#define BTRFS_INODE_ROOT_ITEM_INIT	(1U << 31)
+
+#define BTRFS_INODE_FLAG_MASK						\
+	(BTRFS_INODE_NODATASUM |					\
+	 BTRFS_INODE_NODATACOW |					\
+	 BTRFS_INODE_READONLY |						\
+	 BTRFS_INODE_NOCOMPRESS |					\
+	 BTRFS_INODE_PREALLOC |						\
+	 BTRFS_INODE_SYNC |						\
+	 BTRFS_INODE_IMMUTABLE |					\
+	 BTRFS_INODE_APPEND |						\
+	 BTRFS_INODE_NODUMP |						\
+	 BTRFS_INODE_NOATIME |						\
+	 BTRFS_INODE_DIRSYNC |						\
+	 BTRFS_INODE_COMPRESS |						\
+	 BTRFS_INODE_ROOT_ITEM_INIT)
+
+#define BTRFS_INODE_RO_VERITY		(1U << 0)
+
+#define BTRFS_INODE_RO_FLAG_MASK	(BTRFS_INODE_RO_VERITY)
 
 /*
  * The key defines the order in the tree, and so it also defines (optimal)
@@ -389,6 +450,109 @@ struct btrfs_key {
 	__u64 offset;
 } __attribute__ ((__packed__));
 
+/*
+ * Every tree block (leaf or node) starts with this header.
+ */
+struct btrfs_header {
+	/* These first four must match the super block */
+	__u8 csum[BTRFS_CSUM_SIZE];
+	/* FS specific uuid */
+	__u8 fsid[BTRFS_FSID_SIZE];
+	/* Which block this node is supposed to live in */
+	__le64 bytenr;
+	__le64 flags;
+
+	/* Allowed to be different from the super from here on down */
+	__u8 chunk_tree_uuid[BTRFS_UUID_SIZE];
+	__le64 generation;
+	__le64 owner;
+	__le32 nritems;
+	__u8 level;
+} __attribute__ ((__packed__));
+
+/*
+ * This is a very generous portion of the super block, giving us room to
+ * translate 14 chunks with 3 stripes each.
+ */
+#define BTRFS_SYSTEM_CHUNK_ARRAY_SIZE 2048
+
+/*
+ * Just in case we somehow lose the roots and are not able to mount, we store
+ * an array of the roots from previous transactions in the super.
+ */
+#define BTRFS_NUM_BACKUP_ROOTS 4
+struct btrfs_root_backup {
+	__le64 tree_root;
+	__le64 tree_root_gen;
+
+	__le64 chunk_root;
+	__le64 chunk_root_gen;
+
+	__le64 extent_root;
+	__le64 extent_root_gen;
+
+	__le64 fs_root;
+	__le64 fs_root_gen;
+
+	__le64 dev_root;
+	__le64 dev_root_gen;
+
+	__le64 csum_root;
+	__le64 csum_root_gen;
+
+	__le64 total_bytes;
+	__le64 bytes_used;
+	__le64 num_devices;
+	/* future */
+	__le64 unused_64[4];
+
+	__u8 tree_root_level;
+	__u8 chunk_root_level;
+	__u8 extent_root_level;
+	__u8 fs_root_level;
+	__u8 dev_root_level;
+	__u8 csum_root_level;
+	/* future and to align */
+	__u8 unused_8[10];
+} __attribute__ ((__packed__));
+
+/*
+ * A leaf is full of items. offset and size tell us where to find the item in
+ * the leaf (relative to the start of the data area)
+ */
+struct btrfs_item {
+	struct btrfs_disk_key key;
+	__le32 offset;
+	__le32 size;
+} __attribute__ ((__packed__));
+
+/*
+ * Leaves have an item area and a data area:
+ * [item0, item1....itemN] [free space] [dataN...data1, data0]
+ *
+ * The data is separate from the items to get the keys closer together during
+ * searches.
+ */
+struct btrfs_leaf {
+	struct btrfs_header header;
+	struct btrfs_item items[];
+} __attribute__ ((__packed__));
+
+/*
+ * All non-leaf blocks are nodes, they hold only keys and pointers to other
+ * blocks.
+ */
+struct btrfs_key_ptr {
+	struct btrfs_disk_key key;
+	__le64 blockptr;
+	__le64 generation;
+} __attribute__ ((__packed__));
+
+struct btrfs_node {
+	struct btrfs_header header;
+	struct btrfs_key_ptr ptrs[];
+} __attribute__ ((__packed__));
+
 struct btrfs_dev_item {
 	/* the internal btrfs device id */
 	__le64 devid;
@@ -472,6 +636,69 @@ struct btrfs_chunk {
 	/* additional stripes go here */
 } __attribute__ ((__packed__));
 
+/*
+ * The super block basically lists the main trees of the FS.
+ */
+struct btrfs_super_block {
+	/* The first 4 fields must match struct btrfs_header */
+	__u8 csum[BTRFS_CSUM_SIZE];
+	/* FS specific UUID, visible to user */
+	__u8 fsid[BTRFS_FSID_SIZE];
+	/* This block number */
+	__le64 bytenr;
+	__le64 flags;
+
+	/* Allowed to be different from the btrfs_header from here own down */
+	__le64 magic;
+	__le64 generation;
+	__le64 root;
+	__le64 chunk_root;
+	__le64 log_root;
+
+	/*
+	 * This member has never been utilized since the very beginning, thus
+	 * it's always 0 regardless of kernel version.  We always use
+	 * generation + 1 to read log tree root.  So here we mark it deprecated.
+	 */
+	__le64 __unused_log_root_transid;
+	__le64 total_bytes;
+	__le64 bytes_used;
+	__le64 root_dir_objectid;
+	__le64 num_devices;
+	__le32 sectorsize;
+	__le32 nodesize;
+	__le32 __unused_leafsize;
+	__le32 stripesize;
+	__le32 sys_chunk_array_size;
+	__le64 chunk_root_generation;
+	__le64 compat_flags;
+	__le64 compat_ro_flags;
+	__le64 incompat_flags;
+	__le16 csum_type;
+	__u8 root_level;
+	__u8 chunk_root_level;
+	__u8 log_root_level;
+	struct btrfs_dev_item dev_item;
+
+	char label[BTRFS_LABEL_SIZE];
+
+	__le64 cache_generation;
+	__le64 uuid_tree_generation;
+
+	/* The UUID written into btree blocks */
+	__u8 metadata_uuid[BTRFS_FSID_SIZE];
+
+	__u64 nr_global_roots;
+
+	/* Future expansion */
+	__le64 reserved[27];
+	__u8 sys_chunk_array[BTRFS_SYSTEM_CHUNK_ARRAY_SIZE];
+	struct btrfs_root_backup super_roots[BTRFS_NUM_BACKUP_ROOTS];
+
+	/* Padded to 4096 bytes */
+	__u8 padding[565];
+} __attribute__ ((__packed__));
+
 #define BTRFS_FREE_SPACE_EXTENT	1
 #define BTRFS_FREE_SPACE_BITMAP	2
 
@@ -526,6 +753,14 @@ struct btrfs_extent_item_v0 {
 /* use full backrefs for extent pointers in the block */
 #define BTRFS_BLOCK_FLAG_FULL_BACKREF	(1ULL << 8)
 
+#define BTRFS_BACKREF_REV_MAX		256
+#define BTRFS_BACKREF_REV_SHIFT		56
+#define BTRFS_BACKREF_REV_MASK		(((u64)BTRFS_BACKREF_REV_MAX - 1) << \
+					 BTRFS_BACKREF_REV_SHIFT)
+
+#define BTRFS_OLD_BACKREF_REV		0
+#define BTRFS_MIXED_BACKREF_REV		1
+
 /*
  * this flag is only used internally by scrub and may be changed at any time
  * it is only declared here to avoid collisions