[1/3] xfs_repair: fix maximum file offset comparison

Commit Message

Darrick J. Wong Dec. 23, 2024, 9:36 p.m. UTC

From: Darrick J. Wong <[e-mail]>

When running generic/525 with rtinherit=1 and rextsize=28k, generic/525
trips over the following block mapping:

data offset 2251799813685247 startblock 7 (0/7) count 1 flag 0
data offset 2251799813685248 startblock 8 (0/8) count 6 flag 1

with this error:

inode 155 - extent exceeds max offset - start 2251799813685248, count 6,
physical block 8

This is due to an incorrect check in xfs_repair, which tries to validate
that a block mapping cannot exceed what it thinks is the maximum file
offset.  Unfortunately, the check is wrong, because only br_startoff is
subject to the 2^52-1 limit -- not br_startoff + br_blockcount.

Nowadays libxfs provides a symbol XFS_MAX_FILEOFF for the maximum
allowable file block offset that can be mapped into a file.  Use this
instead of the open-coded logic in versions.c and correct all the other
checks.  Note that this problem only surfaced when rtgroups were enabled
because hch changed xfs_repair to use the same tree-based block state
data structure that we use for AGs when rtgroups are enabled.

Signed-off-by: "Darrick J. Wong" <[e-mail]>
Reviewed-by: Christoph Hellwig <[e-mail]>
---
 repair/dinode.c   |    2 +-
 repair/globals.c  |    1 -
 repair/globals.h  |    1 -
 repair/prefetch.c |    2 +-
 repair/versions.c |    7 +------
 5 files changed, 3 insertions(+), 10 deletions(-)

Patch

diff --git a/repair/dinode.c b/repair/dinode.c
index ac81c487a20b8a..2c9d9acfa10be5 100644
--- a/repair/dinode.c
+++ b/repair/dinode.c
@@ -491,7 +491,7 @@  _("inode %" PRIu64 " - bad extent overflows - start %" PRIu64 ", "
 		}
 		/* Ensure this extent does not extend beyond the max offset */
 		if (irec.br_startoff + irec.br_blockcount - 1 >
-							fs_max_file_offset) {
+							XFS_MAX_FILEOFF) {
 			do_warn(
 _("inode %" PRIu64 " - extent exceeds max offset - start %" PRIu64 ", "
   "count %" PRIu64 ", physical block %" PRIu64 "\n"),
diff --git a/repair/globals.c b/repair/globals.c
index c0c45df51d562a..7388090a7d39f3 100644
--- a/repair/globals.c
+++ b/repair/globals.c
@@ -83,7 +83,6 @@  int		inodes_per_block;
 unsigned int	glob_agcount;
 int		chunks_pblock;	/* # of 64-ino chunks per allocation */
 int		max_symlink_blocks;
-int64_t		fs_max_file_offset;
 
 /* realtime info */
 
diff --git a/repair/globals.h b/repair/globals.h
index 1eadfdbf9ae4f2..fa53502f98bbcd 100644
--- a/repair/globals.h
+++ b/repair/globals.h
@@ -124,7 +124,6 @@  extern int		inodes_per_block;
 extern unsigned int	glob_agcount;
 extern int		chunks_pblock;	/* # of 64-ino chunks per allocation */
 extern int		max_symlink_blocks;
-extern int64_t		fs_max_file_offset;
 
 /* realtime info */
 
diff --git a/repair/prefetch.c b/repair/prefetch.c
index 0772ecef9d73eb..5ecf19ae9cb111 100644
--- a/repair/prefetch.c
+++ b/repair/prefetch.c
@@ -185,7 +185,7 @@  pf_read_bmbt_reclist(
 
 		if (((i > 0) && (op + cp > irec.br_startoff)) ||
 				(irec.br_blockcount == 0) ||
-				(irec.br_startoff >= fs_max_file_offset))
+				(irec.br_startoff + irec.br_blockcount - 1 >= XFS_MAX_FILEOFF))
 			goto out_free;
 
 		if (!libxfs_verify_fsbno(mp, irec.br_startblock) ||
diff --git a/repair/versions.c b/repair/versions.c
index b24965b263a183..7dc91b4597eece 100644
--- a/repair/versions.c
+++ b/repair/versions.c
@@ -180,10 +180,5 @@  _("WARNING: you have a V1 inode filesystem. It would be converted to a\n"
 		fs_ino_alignment = mp->m_sb.sb_inoalignmt;
 	}
 
-	/*
-	 * calculate maximum file offset for this geometry
-	 */
-	fs_max_file_offset = 0x7fffffffffffffffLL >> mp->m_sb.sb_blocklog;
-
-	return(0);
+	return 0;
 }