fs: btrfs: Add single-device read-only BTRFS implementation

This adds the proper implementation for the BTRFS filesystem.
The implementation currently supports only read-only mode and
the filesystem can be only on a single device.

Checksums of data chunks is unimplemented.

Compression is implemented (ZLIB + LZO).

Signed-off-by: Marek Behun <marek.behun@nic.cz>

 create mode 100644 fs/btrfs/btrfs.h
 create mode 100644 fs/btrfs/chunk-map.c
 create mode 100644 fs/btrfs/compression.c
 create mode 100644 fs/btrfs/ctree.c
 create mode 100644 fs/btrfs/dev.c
 create mode 100644 fs/btrfs/dir-item.c
 create mode 100644 fs/btrfs/extent-io.c
 create mode 100644 fs/btrfs/hash.c
 create mode 100644 fs/btrfs/inode.c
 create mode 100644 fs/btrfs/root.c
 create mode 100644 fs/btrfs/subvolume.c
 create mode 100644 fs/btrfs/super.c
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
new file mode 100644
index 0000000..706286e
--- /dev/null
+++ b/fs/btrfs/super.c
@@ -0,0 +1,233 @@
+/*
+ * BTRFS filesystem implementation for U-Boot
+ *
+ * 2017 Marek Behun, CZ.NIC, marek.behun@nic.cz
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include "btrfs.h"
+
+#define BTRFS_SUPER_FLAG_SUPP	(BTRFS_HEADER_FLAG_WRITTEN	\
+				 | BTRFS_HEADER_FLAG_RELOC	\
+				 | BTRFS_SUPER_FLAG_ERROR	\
+				 | BTRFS_SUPER_FLAG_SEEDING	\
+				 | BTRFS_SUPER_FLAG_METADUMP)
+
+#define BTRFS_SUPER_INFO_SIZE	4096
+
+static int btrfs_newest_root_backup(struct btrfs_super_block *sb)
+{
+	struct btrfs_root_backup *root_backup;
+	int i, newest = -1;
+
+	for (i = 0; i < BTRFS_NUM_BACKUP_ROOTS; ++i) {
+		root_backup = sb->super_roots + i;
+		if (root_backup->tree_root_gen == sb->generation)
+			newest = i;
+	}
+
+	return newest;
+}
+
+static inline int is_power_of_2(u64 x)
+{
+	return !(x & (x - 1));
+}
+
+static int btrfs_check_super_csum(char *raw_disk_sb)
+{
+	struct btrfs_super_block *disk_sb =
+		(struct btrfs_super_block *) raw_disk_sb;
+	u16 csum_type = le16_to_cpu(disk_sb->csum_type);
+
+	if (csum_type == BTRFS_CSUM_TYPE_CRC32) {
+		u32 crc = ~(u32) 0;
+		const int csum_size = sizeof(crc);
+		char result[csum_size];
+
+		crc = btrfs_csum_data(raw_disk_sb + BTRFS_CSUM_SIZE, crc,
+				      BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE);
+		btrfs_csum_final(crc, result);
+
+		if (memcmp(raw_disk_sb, result, csum_size))
+			return -1;
+	} else {
+		return -1;
+	}
+
+	return 0;
+}
+
+static int btrfs_check_super(struct btrfs_super_block *sb)
+{
+	int ret = 0;
+
+	if (sb->flags & ~BTRFS_SUPER_FLAG_SUPP) {
+		printf("%s: Unsupported flags: %llu\n", __func__,
+		       sb->flags & ~BTRFS_SUPER_FLAG_SUPP);
+	}
+
+	if (sb->root_level > BTRFS_MAX_LEVEL) {
+		printf("%s: tree_root level too big: %d >= %d\n", __func__,
+		       sb->root_level, BTRFS_MAX_LEVEL);
+		ret = -1;
+	}
+
+	if (sb->chunk_root_level > BTRFS_MAX_LEVEL) {
+		printf("%s: chunk_root level too big: %d >= %d\n", __func__,
+		       sb->chunk_root_level, BTRFS_MAX_LEVEL);
+		ret = -1;
+	}
+
+	if (sb->log_root_level > BTRFS_MAX_LEVEL) {
+		printf("%s: log_root level too big: %d >= %d\n", __func__,
+		       sb->log_root_level, BTRFS_MAX_LEVEL);
+		ret = -1;
+	}
+
+	if (!is_power_of_2(sb->sectorsize) || sb->sectorsize < 4096 ||
+	    sb->sectorsize > BTRFS_MAX_METADATA_BLOCKSIZE) {
+		printf("%s: invalid sectorsize %u\n", __func__,
+		       sb->sectorsize);
+		ret = -1;
+	}
+
+	if (!is_power_of_2(sb->nodesize) || sb->nodesize < sb->sectorsize ||
+	    sb->nodesize > BTRFS_MAX_METADATA_BLOCKSIZE) {
+		printf("%s: invalid nodesize %u\n", __func__, sb->nodesize);
+		ret = -1;
+	}
+
+	if (sb->nodesize != sb->__unused_leafsize) {
+		printf("%s: invalid leafsize %u, should be %u\n", __func__,
+		       sb->__unused_leafsize, sb->nodesize);
+		ret = -1;
+	}
+
+	if (!IS_ALIGNED(sb->root, sb->sectorsize)) {
+		printf("%s: tree_root block unaligned: %llu\n", __func__,
+		       sb->root);
+		ret = -1;
+	}
+
+	if (!IS_ALIGNED(sb->chunk_root, sb->sectorsize)) {
+		printf("%s: chunk_root block unaligned: %llu\n", __func__,
+		       sb->chunk_root);
+		ret = -1;
+	}
+
+	if (!IS_ALIGNED(sb->log_root, sb->sectorsize)) {
+		printf("%s: log_root block unaligned: %llu\n", __func__,
+		       sb->log_root);
+		ret = -1;
+	}
+
+	if (memcmp(sb->fsid, sb->dev_item.fsid, BTRFS_UUID_SIZE) != 0) {
+		printf("%s: dev_item UUID does not match fsid\n", __func__);
+		ret = -1;
+	}
+
+	if (sb->bytes_used < 6*sb->nodesize) {
+		printf("%s: bytes_used is too small %llu\n", __func__,
+		       sb->bytes_used);
+		ret = -1;
+	}
+
+	if (!is_power_of_2(sb->stripesize)) {
+		printf("%s: invalid stripesize %u\n", __func__, sb->stripesize);
+		ret = -1;
+	}
+
+	if (sb->sys_chunk_array_size > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) {
+		printf("%s: system chunk array too big %u > %u\n", __func__,
+		       sb->sys_chunk_array_size, BTRFS_SYSTEM_CHUNK_ARRAY_SIZE);
+		ret = -1;
+	}
+
+	if (sb->sys_chunk_array_size < sizeof(struct btrfs_key) +
+	    sizeof(struct btrfs_chunk)) {
+		printf("%s: system chunk array too small %u < %u\n", __func__,
+		       sb->sys_chunk_array_size, (u32) sizeof(struct btrfs_key)
+		       + sizeof(struct btrfs_chunk));
+		ret = -1;
+	}
+
+	return ret;
+}
+
+int btrfs_read_superblock(void)
+{
+	const u64 superblock_offsets[4] = {
+		0x10000ull,
+		0x4000000ull,
+		0x4000000000ull,
+		0x4000000000000ull
+	};
+	char raw_sb[BTRFS_SUPER_INFO_SIZE];
+	struct btrfs_super_block *sb = (struct btrfs_super_block *) raw_sb;
+	u64 dev_total_bytes;
+	int i, root_backup_idx;
+
+	dev_total_bytes = (u64) btrfs_part_info->size * btrfs_part_info->blksz;
+
+	btrfs_info.sb.generation = 0;
+
+	for (i = 0; i < 4; ++i) {
+		if (superblock_offsets[i] + sizeof(sb) > dev_total_bytes)
+			break;
+
+		if (!btrfs_devread(superblock_offsets[i], BTRFS_SUPER_INFO_SIZE,
+				   raw_sb))
+			break;
+
+		if (btrfs_check_super_csum(raw_sb)) {
+			printf("%s: invalid checksum at superblock mirror %i\n",
+			       __func__, i);
+			continue;
+		}
+
+		btrfs_super_block_to_cpu(sb);
+
+		if (sb->magic != BTRFS_MAGIC) {
+			printf("%s: invalid BTRFS magic 0x%016llX at "
+			       "superblock mirror %i\n", __func__, sb->magic,
+			       i);
+		} else if (sb->bytenr != superblock_offsets[i]) {
+			printf("%s: invalid bytenr 0x%016llX (expected "
+			       "0x%016llX) at superblock mirror %i\n",
+			       __func__, sb->bytenr, superblock_offsets[i], i);
+		} else if (btrfs_check_super(sb)) {
+			printf("%s: Checking superblock mirror %i failed\n",
+			       __func__, i);
+		} else if (sb->generation > btrfs_info.sb.generation) {
+			memcpy(&btrfs_info.sb, sb, sizeof(*sb));
+		} else {
+			/* Nothing */
+		}
+	}
+
+	if (!btrfs_info.sb.generation) {
+		printf("%s: No valid BTRFS superblock found!\n", __func__);
+		return -1;
+	}
+
+	root_backup_idx = btrfs_newest_root_backup(&btrfs_info.sb);
+	if (root_backup_idx < 0) {
+		printf("%s: No valid root_backup found!\n", __func__);
+		return -1;
+	}
+	btrfs_info.root_backup = btrfs_info.sb.super_roots + root_backup_idx;
+
+	if (btrfs_info.root_backup->num_devices != 1) {
+		printf("%s: Unsupported number of devices (%lli). This driver "
+		       "only supports filesystem on one device.\n", __func__,
+		       btrfs_info.root_backup->num_devices);
+		return -1;
+	}
+
+	debug("Chosen superblock with generation = %llu\n",
+	      btrfs_info.sb.generation);
+
+	return 0;
+}