xfs_mount.c

来自「Linux Kernel 2.6.9 for OMAP1710」· C语言 代码 · 共 1,582 行 · 第 1/3 页

			goto fail;
		}
		ASSERT(XFS_BUF_ISBUSY(bp));
		ASSERT(XFS_BUF_VALUSEMA(bp) <= 0);
	}

	mp->m_sb_bp = bp;
	xfs_buf_relse(bp);
	ASSERT(XFS_BUF_VALUSEMA(bp) > 0);
	return 0;

 fail:
	if (bp) {
		XFS_BUF_UNMANAGE(bp);
		xfs_buf_relse(bp);
	}
	return error;
}


/*
 * xfs_mount_common
 *
 * Mount initialization code establishing various mount
 * fields from the superblock associated with the given
 * mount structure
 */
void
xfs_mount_common(xfs_mount_t *mp, xfs_sb_t *sbp)
{
	int	i;

	mp->m_agfrotor = mp->m_agirotor = 0;
	spinlock_init(&mp->m_agirotor_lock, "m_agirotor_lock");
	mp->m_maxagi = mp->m_sb.sb_agcount;
	mp->m_blkbit_log = sbp->sb_blocklog + XFS_NBBYLOG;
	mp->m_blkbb_log = sbp->sb_blocklog - BBSHIFT;
	mp->m_sectbb_log = sbp->sb_sectlog - BBSHIFT;
	mp->m_agno_log = xfs_highbit32(sbp->sb_agcount - 1) + 1;
	mp->m_agino_log = sbp->sb_inopblog + sbp->sb_agblklog;
	mp->m_litino = sbp->sb_inodesize -
		((uint)sizeof(xfs_dinode_core_t) + (uint)sizeof(xfs_agino_t));
	mp->m_blockmask = sbp->sb_blocksize - 1;
	mp->m_blockwsize = sbp->sb_blocksize >> XFS_WORDLOG;
	mp->m_blockwmask = mp->m_blockwsize - 1;
	INIT_LIST_HEAD(&mp->m_del_inodes);

	/*
	 * Setup for attributes, in case they get created.
	 * This value is for inodes getting attributes for the first time,
	 * the per-inode value is for old attribute values.
	 */
	ASSERT(sbp->sb_inodesize >= 256 && sbp->sb_inodesize <= 2048);
	switch (sbp->sb_inodesize) {
	case 256:
		mp->m_attroffset = XFS_LITINO(mp) - XFS_BMDR_SPACE_CALC(2);
		break;
	case 512:
	case 1024:
	case 2048:
		mp->m_attroffset = XFS_BMDR_SPACE_CALC(12);
		break;
	default:
		ASSERT(0);
	}
	ASSERT(mp->m_attroffset < XFS_LITINO(mp));

	for (i = 0; i < 2; i++) {
		mp->m_alloc_mxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize,
			xfs_alloc, i == 0);
		mp->m_alloc_mnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize,
			xfs_alloc, i == 0);
	}
	for (i = 0; i < 2; i++) {
		mp->m_bmap_dmxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize,
			xfs_bmbt, i == 0);
		mp->m_bmap_dmnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize,
			xfs_bmbt, i == 0);
	}
	for (i = 0; i < 2; i++) {
		mp->m_inobt_mxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize,
			xfs_inobt, i == 0);
		mp->m_inobt_mnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize,
			xfs_inobt, i == 0);
	}

	mp->m_bsize = XFS_FSB_TO_BB(mp, 1);
	mp->m_ialloc_inos = (int)MAX((__uint16_t)XFS_INODES_PER_CHUNK,
					sbp->sb_inopblock);
	mp->m_ialloc_blks = mp->m_ialloc_inos >> sbp->sb_inopblog;
}
/*
 * xfs_mountfs
 *
 * This function does the following on an initial mount of a file system:
 *	- reads the superblock from disk and init the mount struct
 *	- if we're a 32-bit kernel, do a size check on the superblock
 *		so we don't mount terabyte filesystems
 *	- init mount struct realtime fields
 *	- allocate inode hash table for fs
 *	- init directory manager
 *	- perform recovery and init the log manager
 */
int
xfs_mountfs(
	vfs_t		*vfsp,
	xfs_mount_t	*mp,
	int		mfsi_flags)
{
	xfs_buf_t	*bp;
	xfs_sb_t	*sbp = &(mp->m_sb);
	xfs_inode_t	*rip;
	vnode_t		*rvp = NULL;
	int		readio_log, writeio_log;
	xfs_daddr_t	d;
	__uint64_t	ret64;
	__int64_t	update_flags;
	uint		quotamount, quotaflags;
	int		agno;
	int		uuid_mounted = 0;
	int		error = 0;

	if (mp->m_sb_bp == NULL) {
		if ((error = xfs_readsb(mp))) {
			return (error);
		}
	}
	xfs_mount_common(mp, sbp);

	/*
	 * Check if sb_agblocks is aligned at stripe boundary
	 * If sb_agblocks is NOT aligned turn off m_dalign since
	 * allocator alignment is within an ag, therefore ag has
	 * to be aligned at stripe boundary.
	 */
	update_flags = 0LL;
	if (mp->m_dalign && !(mfsi_flags & XFS_MFSI_SECOND)) {
		/*
		 * If stripe unit and stripe width are not multiples
		 * of the fs blocksize turn off alignment.
		 */
		if ((BBTOB(mp->m_dalign) & mp->m_blockmask) ||
		    (BBTOB(mp->m_swidth) & mp->m_blockmask)) {
			if (mp->m_flags & XFS_MOUNT_RETERR) {
				cmn_err(CE_WARN,
					"XFS: alignment check 1 failed");
				error = XFS_ERROR(EINVAL);
				goto error1;
			}
			mp->m_dalign = mp->m_swidth = 0;
		} else {
			/*
			 * Convert the stripe unit and width to FSBs.
			 */
			mp->m_dalign = XFS_BB_TO_FSBT(mp, mp->m_dalign);
			if (mp->m_dalign && (sbp->sb_agblocks % mp->m_dalign)) {
				if (mp->m_flags & XFS_MOUNT_RETERR) {
					error = XFS_ERROR(EINVAL);
					goto error1;
				}
				xfs_fs_cmn_err(CE_WARN, mp,
"stripe alignment turned off: sunit(%d)/swidth(%d) incompatible with agsize(%d)",
					mp->m_dalign, mp->m_swidth,
					sbp->sb_agblocks);

				mp->m_dalign = 0;
				mp->m_swidth = 0;
			} else if (mp->m_dalign) {
				mp->m_swidth = XFS_BB_TO_FSBT(mp, mp->m_swidth);
			} else {
				if (mp->m_flags & XFS_MOUNT_RETERR) {
					xfs_fs_cmn_err(CE_WARN, mp,
"stripe alignment turned off: sunit(%d) less than bsize(%d)",
                                        	mp->m_dalign,
						mp->m_blockmask +1);
					error = XFS_ERROR(EINVAL);
					goto error1;
				}
				mp->m_swidth = 0;
			}
		}

		/*
		 * Update superblock with new values
		 * and log changes
		 */
		if (XFS_SB_VERSION_HASDALIGN(sbp)) {
			if (sbp->sb_unit != mp->m_dalign) {
				sbp->sb_unit = mp->m_dalign;
				update_flags |= XFS_SB_UNIT;
			}
			if (sbp->sb_width != mp->m_swidth) {
				sbp->sb_width = mp->m_swidth;
				update_flags |= XFS_SB_WIDTH;
			}
		}
	} else if ((mp->m_flags & XFS_MOUNT_NOALIGN) != XFS_MOUNT_NOALIGN &&
		    XFS_SB_VERSION_HASDALIGN(&mp->m_sb)) {
			mp->m_dalign = sbp->sb_unit;
			mp->m_swidth = sbp->sb_width;
	}

	xfs_alloc_compute_maxlevels(mp);
	xfs_bmap_compute_maxlevels(mp, XFS_DATA_FORK);
	xfs_bmap_compute_maxlevels(mp, XFS_ATTR_FORK);
	xfs_ialloc_compute_maxlevels(mp);

	if (sbp->sb_imax_pct) {
		__uint64_t	icount;

		/* Make sure the maximum inode count is a multiple of the
		 * units we allocate inodes in.
		 */

		icount = sbp->sb_dblocks * sbp->sb_imax_pct;
		do_div(icount, 100);
		do_div(icount, mp->m_ialloc_blks);
		mp->m_maxicount = (icount * mp->m_ialloc_blks)  <<
				   sbp->sb_inopblog;
	} else
		mp->m_maxicount = 0;

	mp->m_maxioffset = xfs_max_file_offset(sbp->sb_blocklog);

	/*
	 * XFS uses the uuid from the superblock as the unique
	 * identifier for fsid.  We can not use the uuid from the volume
	 * since a single partition filesystem is identical to a single
	 * partition volume/filesystem.
	 */
	if ((mfsi_flags & XFS_MFSI_SECOND) == 0 &&
	    (mp->m_flags & XFS_MOUNT_NOUUID) == 0) {
		if (xfs_uuid_mount(mp)) {
			error = XFS_ERROR(EINVAL);
			goto error1;
		}
		uuid_mounted=1;
		ret64 = uuid_hash64(&sbp->sb_uuid);
		memcpy(&vfsp->vfs_fsid, &ret64, sizeof(ret64));
	}

	/*
	 * Set the default minimum read and write sizes unless
	 * already specified in a mount option.
	 * We use smaller I/O sizes when the file system
	 * is being used for NFS service (wsync mount option).
	 */
	if (!(mp->m_flags & XFS_MOUNT_DFLT_IOSIZE)) {
		if (mp->m_flags & XFS_MOUNT_WSYNC) {
			readio_log = XFS_WSYNC_READIO_LOG;
			writeio_log = XFS_WSYNC_WRITEIO_LOG;
		} else {
			readio_log = XFS_READIO_LOG_LARGE;
			writeio_log = XFS_WRITEIO_LOG_LARGE;
		}
	} else {
		readio_log = mp->m_readio_log;
		writeio_log = mp->m_writeio_log;
	}

	/*
	 * Set the number of readahead buffers to use based on
	 * physical memory size.
	 */
	if (xfs_physmem <= 4096)		/* <= 16MB */
		mp->m_nreadaheads = XFS_RW_NREADAHEAD_16MB;
	else if (xfs_physmem <= 8192)	/* <= 32MB */
		mp->m_nreadaheads = XFS_RW_NREADAHEAD_32MB;
	else
		mp->m_nreadaheads = XFS_RW_NREADAHEAD_K32;
	if (sbp->sb_blocklog > readio_log) {
		mp->m_readio_log = sbp->sb_blocklog;
	} else {
		mp->m_readio_log = readio_log;
	}
	mp->m_readio_blocks = 1 << (mp->m_readio_log - sbp->sb_blocklog);
	if (sbp->sb_blocklog > writeio_log) {
		mp->m_writeio_log = sbp->sb_blocklog;
	} else {
		mp->m_writeio_log = writeio_log;
	}
	mp->m_writeio_blocks = 1 << (mp->m_writeio_log - sbp->sb_blocklog);

	/*
	 * Set the inode cluster size based on the physical memory
	 * size.  This may still be overridden by the file system
	 * block size if it is larger than the chosen cluster size.
	 */
	if (xfs_physmem <= btoc(32 * 1024 * 1024)) { /* <= 32 MB */
		mp->m_inode_cluster_size = XFS_INODE_SMALL_CLUSTER_SIZE;
	} else {
		mp->m_inode_cluster_size = XFS_INODE_BIG_CLUSTER_SIZE;
	}
	/*
	 * Set whether we're using inode alignment.
	 */
	if (XFS_SB_VERSION_HASALIGN(&mp->m_sb) &&
	    mp->m_sb.sb_inoalignmt >=
	    XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size))
		mp->m_inoalign_mask = mp->m_sb.sb_inoalignmt - 1;
	else
		mp->m_inoalign_mask = 0;
	/*
	 * If we are using stripe alignment, check whether
	 * the stripe unit is a multiple of the inode alignment
	 */
	if (mp->m_dalign && mp->m_inoalign_mask &&
	    !(mp->m_dalign & mp->m_inoalign_mask))
		mp->m_sinoalign = mp->m_dalign;
	else
		mp->m_sinoalign = 0;
	/*
	 * Check that the data (and log if separate) are an ok size.
	 */
	d = (xfs_daddr_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks);
	if (XFS_BB_TO_FSB(mp, d) != mp->m_sb.sb_dblocks) {
		cmn_err(CE_WARN, "XFS: size check 1 failed");
		error = XFS_ERROR(E2BIG);
		goto error1;
	}
	error = xfs_read_buf(mp, mp->m_ddev_targp,
			     d - XFS_FSS_TO_BB(mp, 1),
			     XFS_FSS_TO_BB(mp, 1), 0, &bp);
	if (!error) {
		xfs_buf_relse(bp);
	} else {
		cmn_err(CE_WARN, "XFS: size check 2 failed");
		if (error == ENOSPC) {
			error = XFS_ERROR(E2BIG);
		}
		goto error1;
	}

	if (((mfsi_flags & XFS_MFSI_CLIENT) == 0) &&
	    mp->m_logdev_targp != mp->m_ddev_targp) {
		d = (xfs_daddr_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_logblocks);
		if (XFS_BB_TO_FSB(mp, d) != mp->m_sb.sb_logblocks) {
			cmn_err(CE_WARN, "XFS: size check 3 failed");
			error = XFS_ERROR(E2BIG);
			goto error1;
		}
		error = xfs_read_buf(mp, mp->m_logdev_targp,
				     d - XFS_FSB_TO_BB(mp, 1),
				     XFS_FSB_TO_BB(mp, 1), 0, &bp);
		if (!error) {
			xfs_buf_relse(bp);
		} else {
			cmn_err(CE_WARN, "XFS: size check 3 failed");
			if (error == ENOSPC) {
				error = XFS_ERROR(E2BIG);
			}
			goto error1;
		}
	}

	/*
	 * Initialize realtime fields in the mount structure
	 */
	if ((error = xfs_rtmount_init(mp))) {
		cmn_err(CE_WARN, "XFS: RT mount failed");
		goto error1;
	}

	/*
	 * For client case we are done now
	 */
	if (mfsi_flags & XFS_MFSI_CLIENT) {
		return(0);
	}

	/*
	 *  Copies the low order bits of the timestamp and the randomly
	 *  set "sequence" number out of a UUID.
	 */
	uuid_getnodeuniq(&sbp->sb_uuid, mp->m_fixedfsid);

	/*
	 *  The vfs structure needs to have a file system independent
	 *  way of checking for the invariant file system ID.  Since it
	 *  can't look at mount structures it has a pointer to the data
	 *  in the mount structure.
	 *
	 *  File systems that don't support user level file handles (i.e.
	 *  all of them except for XFS) will leave vfs_altfsid as NULL.
	 */
	vfsp->vfs_altfsid = (xfs_fsid_t *)mp->m_fixedfsid;
	mp->m_dmevmask = 0;	/* not persistent; set after each mount */

	/*
	 * Select the right directory manager.
	 */
	mp->m_dirops =
		XFS_SB_VERSION_HASDIRV2(&mp->m_sb) ?
			xfsv2_dirops :
			xfsv1_dirops;

	/*
	 * Initialize directory manager's entries.
	 */
	XFS_DIR_MOUNT(mp);

	/*
	 * Initialize the attribute manager's entries.
	 */
	mp->m_attr_magicpct = (mp->m_sb.sb_blocksize * 37) / 100;

	/*
	 * Initialize the precomputed transaction reservations values.
	 */
	xfs_trans_init(mp);

	/*
	 * Allocate and initialize the inode hash table for this
	 * file system.
	 */
	xfs_ihash_init(mp);
	xfs_chash_init(mp);

	/*
	 * Allocate and initialize the per-ag data.
	 */
	init_rwsem(&mp->m_peraglock);
	mp->m_perag =
		kmem_zalloc(sbp->sb_agcount * sizeof(xfs_perag_t), KM_SLEEP);

	mp->m_maxagi = xfs_initialize_perag(mp, sbp->sb_agcount);

	/*
	 * log's mount-time initialization. Perform 1st part recovery if needed
	 */
	if (likely(sbp->sb_logblocks > 0)) {	/* check for volume case */
		error = xfs_log_mount(mp, mp->m_logdev_targp,
				      XFS_FSB_TO_DADDR(mp, sbp->sb_logstart),
				      XFS_FSB_TO_BB(mp, sbp->sb_logblocks));
		if (error) {
			cmn_err(CE_WARN, "XFS: log mount failed");
			goto error2;
		}
	} else {	/* No log has been defined */
		cmn_err(CE_WARN, "XFS: no log defined");
		XFS_ERROR_REPORT("xfs_mountfs_int(1)", XFS_ERRLEVEL_LOW, mp);
		error = XFS_ERROR(EFSCORRUPTED);
		goto error2;
	}

	/*
	 * Get and sanity-check the root inode.
	 * Save the pointer to it in the mount structure.
	 */
	error = xfs_iget(mp, NULL, sbp->sb_rootino, XFS_ILOCK_EXCL, &rip, 0);
	if (error) {
		cmn_err(CE_WARN, "XFS: failed to read root inode");
		goto error3;
	}

	ASSERT(rip != NULL);
	rvp = XFS_ITOV(rip);

	if (unlikely((rip->i_d.di_mode & S_IFMT) != S_IFDIR)) {
		cmn_err(CE_WARN, "XFS: corrupted root inode");
		prdev("Root inode %llu is not a directory",
		      mp->m_ddev_targp, (unsigned long long)rip->i_ino);
		xfs_iunlock(rip, XFS_ILOCK_EXCL);
		XFS_ERROR_REPORT("xfs_mountfs_int(2)", XFS_ERRLEVEL_LOW,
				 mp);
		error = XFS_ERROR(EFSCORRUPTED);
		goto error4;
	}
	mp->m_rootip = rip;	/* save it */

	xfs_iunlock(rip, XFS_ILOCK_EXCL);

	/*
	 * Initialize realtime inode pointers in the mount structure
	 */
	if ((error = xfs_rtmount_inodes(mp))) {
		/*
		 * Free up the root inode.
		 */
		cmn_err(CE_WARN, "XFS: failed to read RT inodes");
		goto error4;
	}

	/*
	 * If fs is not mounted readonly, then update the superblock
	 * unit and width changes.
	 */
	if (update_flags && !(vfsp->vfs_flag & VFS_RDONLY))
		xfs_mount_log_sbunit(mp, update_flags);

	/*
	 * Initialise the XFS quota management subsystem for this mount
	 */
	if ((error = XFS_QM_INIT(mp, &quotamount, &quotaflags)))
		goto error4;

	/*
	 * Finish recovering the file system.  This part needed to be
	 * delayed until after the root and real-time bitmap inodes
	 * were consistently read in.
	 */
	error = xfs_log_mount_finish(mp, mfsi_flags);
	if (error) {
		cmn_err(CE_WARN, "XFS: log mount finish failed");
		goto error4;
	}

	/*
	 * Complete the quota initialisation, post-log-replay component.
	 */
	if ((error = XFS_QM_MOUNT(mp, quotamount, quotaflags)))
		goto error4;

	return 0;

 error4:
	/*
	 * Free up the root inode.
	 */
	VN_RELE(rvp);
 error3:
	xfs_log_unmount_dealloc(mp);
 error2:
	xfs_ihash_free(mp);
	xfs_chash_free(mp);
	for (agno = 0; agno < sbp->sb_agcount; agno++)
		if (mp->m_perag[agno].pagb_list)
			kmem_free(mp->m_perag[agno].pagb_list,