xfs: add support for large btree blocks

Add support for larger btree blocks that contains a CRC32C checksum,
a filesystem uuid and block number for detecting filesystem
consistency and out of place writes.

[dchinner@redhat.com] Also include an owner field to allow reverse
mappings to be implemented for improved repairability and a LSN
field to so that log recovery can easily determine the last
modification that made it to disk for each buffer.

[dchinner@redhat.com] Add buffer log format flags to indicate the
type of buffer to recovery so that we don't have to do blind magic
number tests to determine what the buffer is.

[dchinner@redhat.com] Modified to fit into the verifier structure.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>

1 files changed, 73 insertions, 32 deletions

diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c
index b1ddef6..30c4c14 100644
--- a/fs/xfs/xfs_alloc_btree.c
+++ b/fs/xfs/xfs_alloc_btree.c
@@ -33,6 +33,7 @@
 #include "xfs_extent_busy.h"
 #include "xfs_error.h"
 #include "xfs_trace.h"
+#include "xfs_cksum.h"
 
 
 STATIC struct xfs_btree_cur *
@@ -272,7 +273,7 @@ xfs_allocbt_key_diff(
 	return (__int64_t)be32_to_cpu(kp->ar_startblock) - rec->ar_startblock;
 }
 
-static void
+static bool
 xfs_allocbt_verify(
 	struct xfs_buf		*bp)
 {
@@ -280,66 +281,103 @@ xfs_allocbt_verify(
 	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
 	struct xfs_perag	*pag = bp->b_pag;
 	unsigned int		level;
-	int			sblock_ok; /* block passes checks */
 
 	/*
 	 * magic number and level verification
 	 *
-	 * During growfs operations, we can't verify the exact level as the
-	 * perag is not fully initialised and hence not attached to the buffer.
-	 * In this case, check against the maximum tree depth.
+	 * During growfs operations, we can't verify the exact level or owner as
+	 * the perag is not fully initialised and hence not attached to the
+	 * buffer.  In this case, check against the maximum tree depth.
+	 *
+	 * Similarly, during log recovery we will have a perag structure
+	 * attached, but the agf information will not yet have been initialised
+	 * from the on disk AGF. Again, we can only check against maximum limits
+	 * in this case.
 	 */
 	level = be16_to_cpu(block->bb_level);
 	switch (block->bb_magic) {
+	case cpu_to_be32(XFS_ABTB_CRC_MAGIC):
+		if (!xfs_sb_version_hascrc(&mp->m_sb))
+			return false;
+		if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid))
+			return false;
+		if (block->bb_u.s.bb_blkno != cpu_to_be64(bp->b_bn))
+			return false;
+		if (pag &&
+		    be32_to_cpu(block->bb_u.s.bb_owner) != pag->pag_agno)
+			return false;
+		/* fall through */
 	case cpu_to_be32(XFS_ABTB_MAGIC):
-		if (pag)
-			sblock_ok = level < pag->pagf_levels[XFS_BTNUM_BNOi];
-		else
-			sblock_ok = level < mp->m_ag_maxlevels;
+		if (pag && pag->pagf_init) {
+			if (level >= pag->pagf_levels[XFS_BTNUM_BNOi])
+				return false;
+		} else if (level >= mp->m_ag_maxlevels)
+			return false;
 		break;
+	case cpu_to_be32(XFS_ABTC_CRC_MAGIC):
+		if (!xfs_sb_version_hascrc(&mp->m_sb))
+			return false;
+		if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid))
+			return false;
+		if (block->bb_u.s.bb_blkno != cpu_to_be64(bp->b_bn))
+			return false;
+		if (pag &&
+		    be32_to_cpu(block->bb_u.s.bb_owner) != pag->pag_agno)
+			return false;
+		/* fall through */
 	case cpu_to_be32(XFS_ABTC_MAGIC):
-		if (pag)
-			sblock_ok = level < pag->pagf_levels[XFS_BTNUM_CNTi];
-		else
-			sblock_ok = level < mp->m_ag_maxlevels;
+		if (pag && pag->pagf_init) {
+			if (level >= pag->pagf_levels[XFS_BTNUM_CNTi])
+				return false;
+		} else if (level >= mp->m_ag_maxlevels)
+			return false;
 		break;
 	default:
-		sblock_ok = 0;
-		break;
+		return false;
 	}
 
 	/* numrecs verification */
-	sblock_ok = sblock_ok &&
-		be16_to_cpu(block->bb_numrecs) <= mp->m_alloc_mxr[level != 0];
+	if (be16_to_cpu(block->bb_numrecs) > mp->m_alloc_mxr[level != 0])
+		return false;
 
 	/* sibling pointer verification */
-	sblock_ok = sblock_ok &&
-		(block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) ||
-		 be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) &&
-		block->bb_u.s.bb_leftsib &&
-		(block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
-		 be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) &&
-		block->bb_u.s.bb_rightsib;
-
-	if (!sblock_ok) {
-		trace_xfs_btree_corrupt(bp, _RET_IP_);
-		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block);
-		xfs_buf_ioerror(bp, EFSCORRUPTED);
-	}
+	if (!block->bb_u.s.bb_leftsib ||
+	    (be32_to_cpu(block->bb_u.s.bb_leftsib) >= mp->m_sb.sb_agblocks &&
+	     block->bb_u.s.bb_leftsib != cpu_to_be32(NULLAGBLOCK)))
+		return false;
+	if (!block->bb_u.s.bb_rightsib ||
+	    (be32_to_cpu(block->bb_u.s.bb_rightsib) >= mp->m_sb.sb_agblocks &&
+	     block->bb_u.s.bb_rightsib != cpu_to_be32(NULLAGBLOCK)))
+		return false;
+
+	return true;
 }
 
 static void
 xfs_allocbt_read_verify(
 	struct xfs_buf	*bp)
 {
-	xfs_allocbt_verify(bp);
+	if (!(xfs_btree_sblock_verify_crc(bp) &&
+	      xfs_allocbt_verify(bp))) {
+		trace_xfs_btree_corrupt(bp, _RET_IP_);
+		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+				     bp->b_target->bt_mount, bp->b_addr);
+		xfs_buf_ioerror(bp, EFSCORRUPTED);
+	}
 }
 
 static void
 xfs_allocbt_write_verify(
 	struct xfs_buf	*bp)
 {
-	xfs_allocbt_verify(bp);
+	if (!xfs_allocbt_verify(bp)) {
+		trace_xfs_btree_corrupt(bp, _RET_IP_);
+		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+				     bp->b_target->bt_mount, bp->b_addr);
+		xfs_buf_ioerror(bp, EFSCORRUPTED);
+	}
+	xfs_btree_sblock_calc_crc(bp);
+
 }
 
 const struct xfs_buf_ops xfs_allocbt_buf_ops = {
@@ -444,6 +482,9 @@ xfs_allocbt_init_cursor(
 	cur->bc_private.a.agbp = agbp;
 	cur->bc_private.a.agno = agno;
 
+	if (xfs_sb_version_hascrc(&mp->m_sb))
+		cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
+
 	return cur;
 }