xfs_mount_validate_sb doesn't check sb_inopblock for sanity (as does its xfs_repair counterpart, FWIW). If it's out of bounds, we can go off the rails in i.e. xfs_inode_buf_verify(), which uses sb_inopblock as a loop limit when stepping through a metadata buffer. The problem can be demonstrated easily by corrupting sb_inopblock with xfs_db and trying to mount the result: # mkfs.xfs -dfile,name=fsfile,size=1g # xfs_db -x fsfile xfs_db> sb 0 xfs_db> write inopblock 512 inopblock = 512 xfs_db> quit # mount -o loop fsfile mnt and we blow up in xfs_inode_buf_verify(). With this patch, we get a (very noisy) corruption error, and fail the mount as we should. Signed-off-by: Eric Sandeen <sandeen@xxxxxxxxxx> --- diff --git a/fs/xfs/xfs_sb.c b/fs/xfs/xfs_sb.c index b7c9aea..511cce9 100644 --- a/fs/xfs/xfs_sb.c +++ b/fs/xfs/xfs_sb.c @@ -288,6 +288,7 @@ xfs_mount_validate_sb( sbp->sb_inodelog < XFS_DINODE_MIN_LOG || sbp->sb_inodelog > XFS_DINODE_MAX_LOG || sbp->sb_inodesize != (1 << sbp->sb_inodelog) || + sbp->sb_inopblock != howmany(sbp->sb_blocksize,sbp->sb_inodesize) || (sbp->sb_blocklog - sbp->sb_inodelog != sbp->sb_inopblog) || (sbp->sb_rextsize * sbp->sb_blocksize > XFS_MAX_RTEXTSIZE) || (sbp->sb_rextsize * sbp->sb_blocksize < XFS_MIN_RTEXTSIZE) || _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
