diff options
Diffstat (limited to 'fs/xfs/xfs_mount.c')
| -rw-r--r-- | fs/xfs/xfs_mount.c | 2125 |
1 files changed, 859 insertions, 1266 deletions
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index a66b3980517..3507cd0ec40 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -17,189 +17,148 @@ */ #include "xfs.h" #include "xfs_fs.h" -#include "xfs_types.h" +#include "xfs_shared.h" +#include "xfs_format.h" +#include "xfs_log_format.h" +#include "xfs_trans_resv.h" #include "xfs_bit.h" -#include "xfs_log.h" #include "xfs_inum.h" -#include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" -#include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" +#include "xfs_da_format.h" #include "xfs_inode.h" -#include "xfs_btree.h" +#include "xfs_dir2.h" #include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_rtalloc.h" #include "xfs_bmap.h" +#include "xfs_trans.h" +#include "xfs_trans_priv.h" +#include "xfs_log.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_quota.h" #include "xfs_fsops.h" - -STATIC void xfs_mount_log_sbunit(xfs_mount_t *, __int64_t); -STATIC int xfs_uuid_mount(xfs_mount_t *); -STATIC void xfs_uuid_unmount(xfs_mount_t *mp); -STATIC void xfs_unmountfs_wait(xfs_mount_t *); +#include "xfs_trace.h" +#include "xfs_icache.h" +#include "xfs_dinode.h" #ifdef HAVE_PERCPU_SB -STATIC void xfs_icsb_destroy_counters(xfs_mount_t *); STATIC void xfs_icsb_balance_counter(xfs_mount_t *, xfs_sb_field_t, - int, int); -STATIC void xfs_icsb_sync_counters(xfs_mount_t *); -STATIC int xfs_icsb_modify_counters(xfs_mount_t *, xfs_sb_field_t, - int64_t, int); -STATIC int xfs_icsb_disable_counter(xfs_mount_t *, xfs_sb_field_t); - + int); +STATIC void xfs_icsb_balance_counter_locked(xfs_mount_t *, xfs_sb_field_t, + int); +STATIC void xfs_icsb_disable_counter(xfs_mount_t *, xfs_sb_field_t); #else -#define xfs_icsb_destroy_counters(mp) do { } while (0) -#define xfs_icsb_balance_counter(mp, a, b, c) do { } while (0) -#define xfs_icsb_sync_counters(mp) do { } while (0) -#define xfs_icsb_modify_counters(mp, a, b, c) do { } while (0) - +#define xfs_icsb_balance_counter(mp, a, b) do { } while (0) +#define xfs_icsb_balance_counter_locked(mp, a, b) do { } while (0) #endif -static const struct { - short offset; - short type; /* 0 = integer - * 1 = binary / string (no translation) - */ -} xfs_sb_info[] = { - { offsetof(xfs_sb_t, sb_magicnum), 0 }, - { offsetof(xfs_sb_t, sb_blocksize), 0 }, - { offsetof(xfs_sb_t, sb_dblocks), 0 }, - { offsetof(xfs_sb_t, sb_rblocks), 0 }, - { offsetof(xfs_sb_t, sb_rextents), 0 }, - { offsetof(xfs_sb_t, sb_uuid), 1 }, - { offsetof(xfs_sb_t, sb_logstart), 0 }, - { offsetof(xfs_sb_t, sb_rootino), 0 }, - { offsetof(xfs_sb_t, sb_rbmino), 0 }, - { offsetof(xfs_sb_t, sb_rsumino), 0 }, - { offsetof(xfs_sb_t, sb_rextsize), 0 }, - { offsetof(xfs_sb_t, sb_agblocks), 0 }, - { offsetof(xfs_sb_t, sb_agcount), 0 }, - { offsetof(xfs_sb_t, sb_rbmblocks), 0 }, - { offsetof(xfs_sb_t, sb_logblocks), 0 }, - { offsetof(xfs_sb_t, sb_versionnum), 0 }, - { offsetof(xfs_sb_t, sb_sectsize), 0 }, - { offsetof(xfs_sb_t, sb_inodesize), 0 }, - { offsetof(xfs_sb_t, sb_inopblock), 0 }, - { offsetof(xfs_sb_t, sb_fname[0]), 1 }, - { offsetof(xfs_sb_t, sb_blocklog), 0 }, - { offsetof(xfs_sb_t, sb_sectlog), 0 }, - { offsetof(xfs_sb_t, sb_inodelog), 0 }, - { offsetof(xfs_sb_t, sb_inopblog), 0 }, - { offsetof(xfs_sb_t, sb_agblklog), 0 }, - { offsetof(xfs_sb_t, sb_rextslog), 0 }, - { offsetof(xfs_sb_t, sb_inprogress), 0 }, - { offsetof(xfs_sb_t, sb_imax_pct), 0 }, - { offsetof(xfs_sb_t, sb_icount), 0 }, - { offsetof(xfs_sb_t, sb_ifree), 0 }, - { offsetof(xfs_sb_t, sb_fdblocks), 0 }, - { offsetof(xfs_sb_t, sb_frextents), 0 }, - { offsetof(xfs_sb_t, sb_uquotino), 0 }, - { offsetof(xfs_sb_t, sb_gquotino), 0 }, - { offsetof(xfs_sb_t, sb_qflags), 0 }, - { offsetof(xfs_sb_t, sb_flags), 0 }, - { offsetof(xfs_sb_t, sb_shared_vn), 0 }, - { offsetof(xfs_sb_t, sb_inoalignmt), 0 }, - { offsetof(xfs_sb_t, sb_unit), 0 }, - { offsetof(xfs_sb_t, sb_width), 0 }, - { offsetof(xfs_sb_t, sb_dirblklog), 0 }, - { offsetof(xfs_sb_t, sb_logsectlog), 0 }, - { offsetof(xfs_sb_t, sb_logsectsize),0 }, - { offsetof(xfs_sb_t, sb_logsunit), 0 }, - { offsetof(xfs_sb_t, sb_features2), 0 }, - { sizeof(xfs_sb_t), 0 } -}; +static DEFINE_MUTEX(xfs_uuid_table_mutex); +static int xfs_uuid_table_size; +static uuid_t *xfs_uuid_table; /* - * Return a pointer to an initialized xfs_mount structure. + * See if the UUID is unique among mounted XFS filesystems. + * Mount fails if UUID is nil or a FS with the same UUID is already mounted. */ -xfs_mount_t * -xfs_mount_init(void) +STATIC int +xfs_uuid_mount( + struct xfs_mount *mp) { - xfs_mount_t *mp; + uuid_t *uuid = &mp->m_sb.sb_uuid; + int hole, i; - mp = kmem_zalloc(sizeof(xfs_mount_t), KM_SLEEP); + if (mp->m_flags & XFS_MOUNT_NOUUID) + return 0; - if (xfs_icsb_init_counters(mp)) { - mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB; + if (uuid_is_nil(uuid)) { + xfs_warn(mp, "Filesystem has nil UUID - can't mount"); + return XFS_ERROR(EINVAL); } - AIL_LOCKINIT(&mp->m_ail_lock, "xfs_ail"); - spinlock_init(&mp->m_sb_lock, "xfs_sb"); - mutex_init(&mp->m_ilock); - initnsema(&mp->m_growlock, 1, "xfs_grow"); - /* - * Initialize the AIL. - */ - xfs_trans_ail_init(mp); + mutex_lock(&xfs_uuid_table_mutex); + for (i = 0, hole = -1; i < xfs_uuid_table_size; i++) { + if (uuid_is_nil(&xfs_uuid_table[i])) { + hole = i; + continue; + } + if (uuid_equal(uuid, &xfs_uuid_table[i])) + goto out_duplicate; + } + + if (hole < 0) { + xfs_uuid_table = kmem_realloc(xfs_uuid_table, + (xfs_uuid_table_size + 1) * sizeof(*xfs_uuid_table), + xfs_uuid_table_size * sizeof(*xfs_uuid_table), + KM_SLEEP); + hole = xfs_uuid_table_size++; + } + xfs_uuid_table[hole] = *uuid; + mutex_unlock(&xfs_uuid_table_mutex); + + return 0; + + out_duplicate: + mutex_unlock(&xfs_uuid_table_mutex); + xfs_warn(mp, "Filesystem has duplicate UUID %pU - can't mount", uuid); + return XFS_ERROR(EINVAL); +} + +STATIC void +xfs_uuid_unmount( + struct xfs_mount *mp) +{ + uuid_t *uuid = &mp->m_sb.sb_uuid; + int i; + + if (mp->m_flags & XFS_MOUNT_NOUUID) + return; + + mutex_lock(&xfs_uuid_table_mutex); + for (i = 0; i < xfs_uuid_table_size; i++) { + if (uuid_is_nil(&xfs_uuid_table[i])) + continue; + if (!uuid_equal(uuid, &xfs_uuid_table[i])) + continue; + memset(&xfs_uuid_table[i], 0, sizeof(uuid_t)); + break; + } + ASSERT(i < xfs_uuid_table_size); + mutex_unlock(&xfs_uuid_table_mutex); +} - atomic_set(&mp->m_active_trans, 0); - return mp; +STATIC void +__xfs_free_perag( + struct rcu_head *head) +{ + struct xfs_perag *pag = container_of(head, struct xfs_perag, rcu_head); + + ASSERT(atomic_read(&pag->pag_ref) == 0); + kmem_free(pag); } /* - * Free up the resources associated with a mount structure. Assume that - * the structure was initially zeroed, so we can tell which fields got - * initialized. + * Free up the per-ag resources associated with the mount structure. */ -void -xfs_mount_free( - xfs_mount_t *mp, - int remove_bhv) +STATIC void +xfs_free_perag( + xfs_mount_t *mp) { - if (mp->m_ihash) - xfs_ihash_free(mp); - if (mp->m_chash) - xfs_chash_free(mp); - - if (mp->m_perag) { - int agno; - - for (agno = 0; agno < mp->m_maxagi; agno++) - if (mp->m_perag[agno].pagb_list) - kmem_free(mp->m_perag[agno].pagb_list, - sizeof(xfs_perag_busy_t) * - XFS_PAGB_NUM_SLOTS); - kmem_free(mp->m_perag, - sizeof(xfs_perag_t) * mp->m_sb.sb_agcount); - } + xfs_agnumber_t agno; + struct xfs_perag *pag; - AIL_LOCK_DESTROY(&mp->m_ail_lock); - spinlock_destroy(&mp->m_sb_lock); - mutex_destroy(&mp->m_ilock); - freesema(&mp->m_growlock); - if (mp->m_quotainfo) - XFS_QM_DONE(mp); - - if (mp->m_fsname != NULL) - kmem_free(mp->m_fsname, mp->m_fsname_len); - if (mp->m_rtname != NULL) - kmem_free(mp->m_rtname, strlen(mp->m_rtname) + 1); - if (mp->m_logname != NULL) - kmem_free(mp->m_logname, strlen(mp->m_logname) + 1); - - if (remove_bhv) { - struct bhv_vfs *vfsp = XFS_MTOVFS(mp); - - bhv_remove_all_vfsops(vfsp, 0); - VFS_REMOVEBHV(vfsp, &mp->m_bhv); + for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) { + spin_lock(&mp->m_perag_lock); + pag = radix_tree_delete(&mp->m_perag_tree, agno); + spin_unlock(&mp->m_perag_lock); + ASSERT(pag); + ASSERT(atomic_read(&pag->pag_ref) == 0); + call_rcu(&pag->rcu_head, __xfs_free_perag); } - - xfs_icsb_destroy_counters(mp); - kmem_free(mp, sizeof(xfs_mount_t)); } /* @@ -216,264 +175,96 @@ xfs_sb_validate_fsb_count( #if XFS_BIG_BLKNOS /* Limited by ULONG_MAX of page cache index */ if (nblocks >> (PAGE_CACHE_SHIFT - sbp->sb_blocklog) > ULONG_MAX) - return E2BIG; + return EFBIG; #else /* Limited by UINT_MAX of sectors */ if (nblocks << (sbp->sb_blocklog - BBSHIFT) > UINT_MAX) - return E2BIG; + return EFBIG; #endif return 0; } -/* - * Check the validity of the SB found. - */ -STATIC int -xfs_mount_validate_sb( - xfs_mount_t *mp, - xfs_sb_t *sbp, - int flags) -{ - /* - * If the log device and data device have the - * same device number, the log is internal. - * Consequently, the sb_logstart should be non-zero. If - * we have a zero sb_logstart in this case, we may be trying to mount - * a volume filesystem in a non-volume manner. - */ - if (sbp->sb_magicnum != XFS_SB_MAGIC) { - xfs_fs_mount_cmn_err(flags, "bad magic number"); - return XFS_ERROR(EWRONGFS); - } - - if (!XFS_SB_GOOD_VERSION(sbp)) { - xfs_fs_mount_cmn_err(flags, "bad version"); - return XFS_ERROR(EWRONGFS); - } - - if (unlikely( - sbp->sb_logstart == 0 && mp->m_logdev_targp == mp->m_ddev_targp)) { - xfs_fs_mount_cmn_err(flags, - "filesystem is marked as having an external log; " - "specify logdev on the\nmount command line."); - return XFS_ERROR(EINVAL); - } - - if (unlikely( - sbp->sb_logstart != 0 && mp->m_logdev_targp != mp->m_ddev_targp)) { - xfs_fs_mount_cmn_err(flags, - "filesystem is marked as having an internal log; " - "do not specify logdev on\nthe mount command line."); - return XFS_ERROR(EINVAL); - } - - /* - * More sanity checking. These were stolen directly from - * xfs_repair. - */ - if (unlikely( - sbp->sb_agcount <= 0 || - sbp->sb_sectsize < XFS_MIN_SECTORSIZE || - sbp->sb_sectsize > XFS_MAX_SECTORSIZE || - sbp->sb_sectlog < XFS_MIN_SECTORSIZE_LOG || - sbp->sb_sectlog > XFS_MAX_SECTORSIZE_LOG || - sbp->sb_blocksize < XFS_MIN_BLOCKSIZE || - sbp->sb_blocksize > XFS_MAX_BLOCKSIZE || - sbp->sb_blocklog < XFS_MIN_BLOCKSIZE_LOG || - sbp->sb_blocklog > XFS_MAX_BLOCKSIZE_LOG || - sbp->sb_inodesize < XFS_DINODE_MIN_SIZE || - sbp->sb_inodesize > XFS_DINODE_MAX_SIZE || - sbp->sb_inodelog < XFS_DINODE_MIN_LOG || - sbp->sb_inodelog > XFS_DINODE_MAX_LOG || - (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) || - (sbp->sb_imax_pct > 100 /* zero sb_imax_pct is valid */))) { - xfs_fs_mount_cmn_err(flags, "SB sanity check 1 failed"); - return XFS_ERROR(EFSCORRUPTED); - } - - /* - * Sanity check AG count, size fields against data size field - */ - if (unlikely( - sbp->sb_dblocks == 0 || - sbp->sb_dblocks > - (xfs_drfsbno_t)sbp->sb_agcount * sbp->sb_agblocks || - sbp->sb_dblocks < (xfs_drfsbno_t)(sbp->sb_agcount - 1) * - sbp->sb_agblocks + XFS_MIN_AG_BLOCKS)) { - xfs_fs_mount_cmn_err(flags, "SB sanity check 2 failed"); - return XFS_ERROR(EFSCORRUPTED); - } - - if (xfs_sb_validate_fsb_count(sbp, sbp->sb_dblocks) || - xfs_sb_validate_fsb_count(sbp, sbp->sb_rblocks)) { - xfs_fs_mount_cmn_err(flags, - "file system too large to be mounted on this system."); - return XFS_ERROR(E2BIG); - } - - if (unlikely(sbp->sb_inprogress)) { - xfs_fs_mount_cmn_err(flags, "file system busy"); - return XFS_ERROR(EFSCORRUPTED); - } - - /* - * Version 1 directory format has never worked on Linux. - */ - if (unlikely(!XFS_SB_VERSION_HASDIRV2(sbp))) { - xfs_fs_mount_cmn_err(flags, - "file system using version 1 directory format"); - return XFS_ERROR(ENOSYS); - } - - /* - * Until this is fixed only page-sized or smaller data blocks work. - */ - if (unlikely(sbp->sb_blocksize > PAGE_SIZE)) { - xfs_fs_mount_cmn_err(flags, - "file system with blocksize %d bytes", - sbp->sb_blocksize); - xfs_fs_mount_cmn_err(flags, - "only pagesize (%ld) or less will currently work.", - PAGE_SIZE); - return XFS_ERROR(ENOSYS); - } - - return 0; -} - -xfs_agnumber_t +int xfs_initialize_perag( - bhv_vfs_t *vfs, xfs_mount_t *mp, - xfs_agnumber_t agcount) + xfs_agnumber_t agcount, + xfs_agnumber_t *maxagi) { - xfs_agnumber_t index, max_metadata; + xfs_agnumber_t index; + xfs_agnumber_t first_initialised = 0; xfs_perag_t *pag; xfs_agino_t agino; xfs_ino_t ino; xfs_sb_t *sbp = &mp->m_sb; - xfs_ino_t max_inum = XFS_MAXINUMBER_32; - - /* Check to see if the filesystem can overflow 32 bit inodes */ - agino = XFS_OFFBNO_TO_AGINO(mp, sbp->sb_agblocks - 1, 0); - ino = XFS_AGINO_TO_INO(mp, agcount - 1, agino); + int error = -ENOMEM; - /* Clear the mount flag if no inode can overflow 32 bits - * on this filesystem, or if specifically requested.. + /* + * Walk the current per-ag tree so we don't try to initialise AGs + * that already exist (growfs case). Allocate and insert all the + * AGs we don't find ready for initialisation. */ - if ((vfs->vfs_flag & VFS_32BITINODES) && ino > max_inum) { - mp->m_flags |= XFS_MOUNT_32BITINODES; - } else { - mp->m_flags &= ~XFS_MOUNT_32BITINODES; - } - - /* If we can overflow then setup the ag headers accordingly */ - if (mp->m_flags & XFS_MOUNT_32BITINODES) { - /* Calculate how much should be reserved for inodes to - * meet the max inode percentage. - */ - if (mp->m_maxicount) { - __uint64_t icount; - - icount = sbp->sb_dblocks * sbp->sb_imax_pct; - do_div(icount, 100); - icount += sbp->sb_agblocks - 1; - do_div(icount, sbp->sb_agblocks); - max_metadata = icount; - } else { - max_metadata = agcount; - } - for (index = 0; index < agcount; index++) { - ino = XFS_AGINO_TO_INO(mp, index, agino); - if (ino > max_inum) { - index++; - break; - } - - /* This ag is preferred for inodes */ - pag = &mp->m_perag[index]; - pag->pagi_inodeok = 1; - if (index < max_metadata) - pag->pagf_metadata = 1; + for (index = 0; index < agcount; index++) { + pag = xfs_perag_get(mp, index); + if (pag) { + xfs_perag_put(pag); + continue; } - } else { - /* Setup default behavior for smaller filesystems */ - for (index = 0; index < agcount; index++) { - pag = &mp->m_perag[index]; - pag->pagi_inodeok = 1; + if (!first_initialised) + first_initialised = index; + + pag = kmem_zalloc(sizeof(*pag), KM_MAYFAIL); + if (!pag) + goto out_unwind; + pag->pag_agno = index; + pag->pag_mount = mp; + spin_lock_init(&pag->pag_ici_lock); + mutex_init(&pag->pag_ici_reclaim_lock); + INIT_RADIX_TREE(&pag->pag_ici_root, GFP_ATOMIC); + spin_lock_init(&pag->pag_buf_lock); + pag->pag_buf_tree = RB_ROOT; + + if (radix_tree_preload(GFP_NOFS)) + goto out_unwind; + + spin_lock(&mp->m_perag_lock); + if (radix_tree_insert(&mp->m_perag_tree, index, pag)) { + BUG(); + spin_unlock(&mp->m_perag_lock); + radix_tree_preload_end(); + error = -EEXIST; + goto out_unwind; } + spin_unlock(&mp->m_perag_lock); + radix_tree_preload_end(); } - return index; -} - -/* - * xfs_xlatesb - * - * data - on disk version of sb - * sb - a superblock - * dir - conversion direction: <0 - convert sb to buf - * >0 - convert buf to sb - * fields - which fields to copy (bitmask) - */ -void -xfs_xlatesb( - void *data, - xfs_sb_t *sb, - int dir, - __int64_t fields) -{ - xfs_caddr_t buf_ptr; - xfs_caddr_t mem_ptr; - xfs_sb_field_t f; - int first; - int size; - - ASSERT(dir); - ASSERT(fields); - - if (!fields) - return; - buf_ptr = (xfs_caddr_t)data; - mem_ptr = (xfs_caddr_t)sb; + /* + * If we mount with the inode64 option, or no inode overflows + * the legacy 32-bit address space clear the inode32 option. + */ + agino = XFS_OFFBNO_TO_AGINO(mp, sbp->sb_agblocks - 1, 0); + ino = XFS_AGINO_TO_INO(mp, agcount - 1, agino); - while (fields) { - f = (xfs_sb_field_t)xfs_lowbit64((__uint64_t)fields); - first = xfs_sb_info[f].offset; - size = xfs_sb_info[f + 1].offset - first; + if ((mp->m_flags & XFS_MOUNT_SMALL_INUMS) && ino > XFS_MAXINUMBER_32) + mp->m_flags |= XFS_MOUNT_32BITINODES; + else + mp->m_flags &= ~XFS_MOUNT_32BITINODES; - ASSERT(xfs_sb_info[f].type == 0 || xfs_sb_info[f].type == 1); + if (mp->m_flags & XFS_MOUNT_32BITINODES) + index = xfs_set_inode32(mp); + else + index = xfs_set_inode64(mp); - if (size == 1 || xfs_sb_info[f].type == 1) { - if (dir > 0) { - memcpy(mem_ptr + first, buf_ptr + first, size); - } else { - memcpy(buf_ptr + first, mem_ptr + first, size); - } - } else { - switch (size) { - case 2: - INT_XLATE(*(__uint16_t*)(buf_ptr+first), - *(__uint16_t*)(mem_ptr+first), - dir, ARCH_CONVERT); - break; - case 4: - INT_XLATE(*(__uint32_t*)(buf_ptr+first), - *(__uint32_t*)(mem_ptr+first), - dir, ARCH_CONVERT); - break; - case 8: - INT_XLATE(*(__uint64_t*)(buf_ptr+first), - *(__uint64_t*)(mem_ptr+first), dir, ARCH_CONVERT); - break; - default: - ASSERT(0); - } - } + if (maxagi) + *maxagi = index; + return 0; - fields &= ~(1LL << f); +out_unwind: + kmem_free(pag); + for (; index > first_initialised; index--) { + pag = radix_tree_delete(&mp->m_perag_tree, index); + kmem_free(pag); } + return error; } /* @@ -482,313 +273,143 @@ xfs_xlatesb( * Does the initial read of the superblock. */ int -xfs_readsb(xfs_mount_t *mp, int flags) +xfs_readsb( + struct xfs_mount *mp, + int flags) { unsigned int sector_size; - unsigned int extra_flags; - xfs_buf_t *bp; - xfs_sb_t *sbp; + struct xfs_buf *bp; + struct xfs_sb *sbp = &mp->m_sb; int error; + int loud = !(flags & XFS_MFSI_QUIET); + const struct xfs_buf_ops *buf_ops; ASSERT(mp->m_sb_bp == NULL); ASSERT(mp->m_ddev_targp != NULL); /* + * For the initial read, we must guess at the sector + * size based on the block device. It's enough to + * get the sb_sectsize out of the superblock and + * then reread with the proper length. + * We don't verify it yet, because it may not be complete. + */ + sector_size = xfs_getsize_buftarg(mp->m_ddev_targp); + buf_ops = NULL; + + /* * Allocate a (locked) buffer to hold the superblock. * This will be kept around at all times to optimize * access to the superblock. */ - sector_size = xfs_getsize_buftarg(mp->m_ddev_targp); - extra_flags = XFS_BUF_LOCK | XFS_BUF_MANAGE | XFS_BUF_MAPPED; - - bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, - BTOBB(sector_size), extra_flags); - if (!bp || XFS_BUF_ISERROR(bp)) { - xfs_fs_mount_cmn_err(flags, "SB read failed"); - error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; - goto fail; +reread: + bp = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_SB_DADDR, + BTOBB(sector_size), 0, buf_ops); + if (!bp) { + if (loud) + xfs_warn(mp, "SB buffer read failed"); + return EIO; + } + if (bp->b_error) { + error = bp->b_error; + if (loud) + xfs_warn(mp, "SB validate failed with error %d.", error); + /* bad CRC means corrupted metadata */ + if (error == EFSBADCRC) + error = EFSCORRUPTED; + goto release_buf; } - ASSERT(XFS_BUF_ISBUSY(bp)); - ASSERT(XFS_BUF_VALUSEMA(bp) <= 0); /* * Initialize the mount structure from the superblock. - * But first do some basic consistency checking. */ - sbp = XFS_BUF_TO_SBP(bp); - xfs_xlatesb(XFS_BUF_PTR(bp), &(mp->m_sb), 1, XFS_SB_ALL_BITS); + xfs_sb_from_disk(sbp, XFS_BUF_TO_SBP(bp)); + xfs_sb_quota_from_disk(sbp); - error = xfs_mount_validate_sb(mp, &(mp->m_sb), flags); - if (error) { - xfs_fs_mount_cmn_err(flags, "SB validate failed"); - goto fail; + /* + * If we haven't validated the superblock, do so now before we try + * to check the sector size and reread the superblock appropriately. + */ + if (sbp->sb_magicnum != XFS_SB_MAGIC) { + if (loud) + xfs_warn(mp, "Invalid superblock magic number"); + error = EINVAL; + goto release_buf; } /* * We must be able to do sector-sized and sector-aligned IO. */ - if (sector_size > mp->m_sb.sb_sectsize) { - xfs_fs_mount_cmn_err(flags, - "device supports only %u byte sectors (not %u)", - sector_size, mp->m_sb.sb_sectsize); + if (sector_size > sbp->sb_sectsize) { + if (loud) + xfs_warn(mp, "device supports %u byte sectors (not %u)", + sector_size, sbp->sb_sectsize); error = ENOSYS; - goto fail; + goto release_buf; } - /* - * If device sector size is smaller than the superblock size, - * re-read the superblock so the buffer is correctly sized. - */ - if (sector_size < mp->m_sb.sb_sectsize) { - XFS_BUF_UNMANAGE(bp); + if (buf_ops == NULL) { + /* + * Re-read the superblock so the buffer is correctly sized, + * and properly verified. + */ xfs_buf_relse(bp); - sector_size = mp->m_sb.sb_sectsize; - bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, - BTOBB(sector_size), extra_flags); - if (!bp || XFS_BUF_ISERROR(bp)) { - xfs_fs_mount_cmn_err(flags, "SB re-read failed"); - error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; - goto fail; - } - ASSERT(XFS_BUF_ISBUSY(bp)); - ASSERT(XFS_BUF_VALUSEMA(bp) <= 0); + sector_size = sbp->sb_sectsize; + buf_ops = loud ? &xfs_sb_buf_ops : &xfs_sb_quiet_buf_ops; + goto reread; } /* Initialize per-cpu counters */ xfs_icsb_reinit_counters(mp); + /* no need to be quiet anymore, so reset the buf ops */ + bp->b_ops = &xfs_sb_buf_ops; + mp->m_sb_bp = bp; - xfs_buf_relse(bp); - ASSERT(XFS_BUF_VALUSEMA(bp) > 0); + xfs_buf_unlock(bp); return 0; - fail: - if (bp) { - XFS_BUF_UNMANAGE(bp); - xfs_buf_relse(bp); - } +release_buf: + 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 - */ -STATIC 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(MINABTPTRS); - break; - case 512: - case 1024: - case 2048: - mp->m_attroffset = XFS_BMDR_SPACE_CALC(6 * MINABTPTRS); - 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_initialize_perag_data - * - * Read in each per-ag structure so we can count up the number of - * allocated inodes, free inodes and used filesystem blocks as this - * information is no longer persistent in the superblock. Once we have - * this information, write it into the in-core superblock structure. + * Update alignment values based on mount options and sb values */ STATIC int -xfs_initialize_perag_data(xfs_mount_t *mp, xfs_agnumber_t agcount) -{ - xfs_agnumber_t index; - xfs_perag_t *pag; - xfs_sb_t *sbp = &mp->m_sb; - uint64_t ifree = 0; - uint64_t ialloc = 0; - uint64_t bfree = 0; - uint64_t bfreelst = 0; - uint64_t btree = 0; - int error; - int s; - - for (index = 0; index < agcount; index++) { - /* - * read the agf, then the agi. This gets us - * all the inforamtion we need and populates the - * per-ag structures for us. - */ - error = xfs_alloc_pagf_init(mp, NULL, index, 0); - if (error) - return error; - - error = xfs_ialloc_pagi_init(mp, NULL, index); - if (error) - return error; - pag = &mp->m_perag[index]; - ifree += pag->pagi_freecount; - ialloc += pag->pagi_count; - bfree += pag->pagf_freeblks; - bfreelst += pag->pagf_flcount; - btree += pag->pagf_btreeblks; - } - /* - * Overwrite incore superblock counters with just-read data - */ - s = XFS_SB_LOCK(mp); - sbp->sb_ifree = ifree; - sbp->sb_icount = ialloc; - sbp->sb_fdblocks = bfree + bfreelst + btree; - XFS_SB_UNLOCK(mp, s); - - /* Fixup the per-cpu counters as well. */ - xfs_icsb_reinit_counters(mp); - - return 0; -} - -/* - * 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( - bhv_vfs_t *vfsp, - xfs_mount_t *mp, - int mfsi_flags) +xfs_update_alignment(xfs_mount_t *mp) { - xfs_buf_t *bp; xfs_sb_t *sbp = &(mp->m_sb); - xfs_inode_t *rip; - bhv_vnode_t *rvp = NULL; - int readio_log, writeio_log; - xfs_daddr_t d; - __uint64_t resblks; - __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, mfsi_flags))) { - 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 (mp->m_dalign) { /* * 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; + xfs_warn(mp, + "alignment check failed: sunit/swidth vs. blocksize(%d)", + sbp->sb_blocksize); + return XFS_ERROR(EINVAL); } 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; + xfs_warn(mp, + "alignment check failed: sunit/swidth vs. agsize(%d)", + sbp->sb_agblocks); + return XFS_ERROR(EINVAL); } 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; + xfs_warn(mp, + "alignment check failed: sunit(%d) less than bsize(%d)", + mp->m_dalign, sbp->sb_blocksize); + return XFS_ERROR(EINVAL); } } @@ -796,68 +417,65 @@ xfs_mountfs( * Update superblock with new values * and log changes */ - if (XFS_SB_VERSION_HASDALIGN(sbp)) { + if (xfs_sb_version_hasdalign(sbp)) { if (sbp->sb_unit != mp->m_dalign) { sbp->sb_unit = mp->m_dalign; - update_flags |= XFS_SB_UNIT; + mp->m_update_flags |= XFS_SB_UNIT; } if (sbp->sb_width != mp->m_swidth) { sbp->sb_width = mp->m_swidth; - update_flags |= XFS_SB_WIDTH; + mp->m_update_flags |= XFS_SB_WIDTH; } + } else { + xfs_warn(mp, + "cannot change alignment: superblock does not support data alignment"); + return XFS_ERROR(EINVAL); } } else if ((mp->m_flags & XFS_MOUNT_NOALIGN) != XFS_MOUNT_NOALIGN && - XFS_SB_VERSION_HASDALIGN(&mp->m_sb)) { + 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); + return 0; +} - if (sbp->sb_imax_pct) { - __uint64_t icount; +/* + * Set the maximum inode count for this filesystem + */ +STATIC void +xfs_set_maxicount(xfs_mount_t *mp) +{ + xfs_sb_t *sbp = &(mp->m_sb); + __uint64_t icount; - /* Make sure the maximum inode count is a multiple of the - * units we allocate inodes in. + if (sbp->sb_imax_pct) { + /* + * 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 + } 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) { - __uint64_t ret64; - 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). + */ +STATIC void +xfs_set_rw_sizes(xfs_mount_t *mp) +{ + xfs_sb_t *sbp = &(mp->m_sb); + int readio_log, writeio_log; - /* - * 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; @@ -871,16 +489,6 @@ xfs_mountfs( 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 { @@ -893,21 +501,33 @@ xfs_mountfs( 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; +/* + * precalculate the low space thresholds for dynamic speculative preallocation. + */ +void +xfs_set_low_space_thresholds( + struct xfs_mount *mp) +{ + int i; + + for (i = 0; i < XFS_LOWSP_MAX; i++) { + __uint64_t space = mp->m_sb.sb_dblocks; + + do_div(space, 100); + mp->m_low_space[i] = space * (i + 1); } - /* - * Set whether we're using inode alignment. - */ - if (XFS_SB_VERSION_HASALIGN(&mp->m_sb) && +} + + +/* + * Set whether we're using inode alignment. + */ +STATIC void +xfs_set_inoalignment(xfs_mount_t *mp) +{ + 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; @@ -922,127 +542,293 @@ xfs_mountfs( mp->m_sinoalign = mp->m_dalign; else mp->m_sinoalign = 0; - /* - * Check that the data (and log if separate) are an ok size. - */ +} + +/* + * Check that the data (and log if separate) is an ok size. + */ +STATIC int +xfs_check_sizes(xfs_mount_t *mp) +{ + xfs_buf_t *bp; + xfs_daddr_t d; + 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; + xfs_warn(mp, "filesystem size mismatch detected"); + return XFS_ERROR(EFBIG); } - 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; + bp = xfs_buf_read_uncached(mp->m_ddev_targp, + d - XFS_FSS_TO_BB(mp, 1), + XFS_FSS_TO_BB(mp, 1), 0, NULL); + if (!bp) { + xfs_warn(mp, "last sector read failed"); + return EIO; } + xfs_buf_relse(bp); - if (((mfsi_flags & XFS_MFSI_CLIENT) == 0) && - mp->m_logdev_targp != mp->m_ddev_targp) { + if (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; + xfs_warn(mp, "log size mismatch detected"); + return XFS_ERROR(EFBIG); } - 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; + bp = xfs_buf_read_uncached(mp->m_logdev_targp, + d - XFS_FSB_TO_BB(mp, 1), + XFS_FSB_TO_BB(mp, 1), 0, NULL); + if (!bp) { + xfs_warn(mp, "log device read failed"); + return EIO; } + xfs_buf_relse(bp); } + return 0; +} + +/* + * Clear the quotaflags in memory and in the superblock. + */ +int +xfs_mount_reset_sbqflags( + struct xfs_mount *mp) +{ + int error; + struct xfs_trans *tp; + + mp->m_qflags = 0; /* - * Initialize realtime fields in the mount structure + * It is OK to look at sb_qflags here in mount path, + * without m_sb_lock. */ - if ((error = xfs_rtmount_init(mp))) { - cmn_err(CE_WARN, "XFS: RT mount failed"); - goto error1; - } + if (mp->m_sb.sb_qflags == 0) + return 0; + spin_lock(&mp->m_sb_lock); + mp->m_sb.sb_qflags = 0; + spin_unlock(&mp->m_sb_lock); /* - * For client case we are done now + * If the fs is readonly, let the incore superblock run + * with quotas off but don't flush the update out to disk */ - if (mfsi_flags & XFS_MFSI_CLIENT) { + if (mp->m_flags & XFS_MOUNT_RDONLY) return 0; + + tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE); + error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_sbchange, 0, 0); + if (error) { + xfs_trans_cancel(tp, 0); + xfs_alert(mp, "%s: Superblock update failed!", __func__); + return error; + } + + xfs_mod_sb(tp, XFS_SB_QFLAGS); + return xfs_trans_commit(tp, 0); +} + +__uint64_t +xfs_default_resblks(xfs_mount_t *mp) +{ + __uint64_t resblks; + + /* + * We default to 5% or 8192 fsbs of space reserved, whichever is + * smaller. This is intended to cover concurrent allocation + * transactions when we initially hit enospc. These each require a 4 + * block reservation. Hence by default we cover roughly 2000 concurrent + * allocation reservations. + */ + resblks = mp->m_sb.sb_dblocks; + do_div(resblks, 20); + resblks = min_t(__uint64_t, resblks, 8192); + return resblks; +} + +/* + * 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( + xfs_mount_t *mp) +{ + xfs_sb_t *sbp = &(mp->m_sb); + xfs_inode_t *rip; + __uint64_t resblks; + uint quotamount = 0; + uint quotaflags = 0; + int error = 0; + + xfs_sb_mount_common(mp, sbp); + + /* + * Check for a mismatched features2 values. Older kernels + * read & wrote into the wrong sb offset for sb_features2 + * on some platforms due to xfs_sb_t not being 64bit size aligned + * when sb_features2 was added, which made older superblock + * reading/writing routines swap it as a 64-bit value. + * + * For backwards compatibility, we make both slots equal. + * + * If we detect a mismatched field, we OR the set bits into the + * existing features2 field in case it has already been modified; we + * don't want to lose any features. We then update the bad location + * with the ORed value so that older kernels will see any features2 + * flags, and mark the two fields as needing updates once the + * transaction subsystem is online. + */ + if (xfs_sb_has_mismatched_features2(sbp)) { + xfs_warn(mp, "correcting sb_features alignment problem"); + sbp->sb_features2 |= sbp->sb_bad_features2; + sbp->sb_bad_features2 = sbp->sb_features2; + mp->m_update_flags |= XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2; + + /* + * Re-check for ATTR2 in case it was found in bad_features2 + * slot. + */ + if (xfs_sb_version_hasattr2(&mp->m_sb) && + !(mp->m_flags & XFS_MOUNT_NOATTR2)) + mp->m_flags |= XFS_MOUNT_ATTR2; + } + + if (xfs_sb_version_hasattr2(&mp->m_sb) && + (mp->m_flags & XFS_MOUNT_NOATTR2)) { + xfs_sb_version_removeattr2(&mp->m_sb); + mp->m_update_flags |= XFS_SB_FEATURES2; + + /* update sb_versionnum for the clearing of the morebits */ + if (!sbp->sb_features2) + mp->m_update_flags |= XFS_SB_VERSIONNUM; + } + + /* always use v2 inodes by default now */ + if (!(mp->m_sb.sb_versionnum & XFS_SB_VERSION_NLINKBIT)) { + mp->m_sb.sb_versionnum |= XFS_SB_VERSION_NLINKBIT; + mp->m_update_flags |= XFS_SB_VERSIONNUM; } /* - * Copies the low order bits of the timestamp and the randomly - * set "sequence" number out of a UUID. + * 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. */ - uuid_getnodeuniq(&sbp->sb_uuid, mp->m_fixedfsid); + error = xfs_update_alignment(mp); + if (error) + goto out; + + 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); + + xfs_set_maxicount(mp); + + error = xfs_uuid_mount(mp); + if (error) + goto out; + + /* + * Set the minimum read and write sizes + */ + xfs_set_rw_sizes(mp); + + /* set the low space thresholds for dynamic preallocation */ + xfs_set_low_space_thresholds(mp); /* - * 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. + * Set the inode cluster size. + * This may still be overridden by the file system + * block size if it is larger than the chosen cluster size. * - * File systems that don't support user level file handles (i.e. - * all of them except for XFS) will leave vfs_altfsid as NULL. + * For v5 filesystems, scale the cluster size with the inode size to + * keep a constant ratio of inode per cluster buffer, but only if mkfs + * has set the inode alignment value appropriately for larger cluster + * sizes. */ - vfsp->vfs_altfsid = (xfs_fsid_t *)mp->m_fixedfsid; - mp->m_dmevmask = 0; /* not persistent; set after each mount */ + mp->m_inode_cluster_size = XFS_INODE_BIG_CLUSTER_SIZE; + if (xfs_sb_version_hascrc(&mp->m_sb)) { + int new_size = mp->m_inode_cluster_size; - xfs_dir_mount(mp); + new_size *= mp->m_sb.sb_inodesize / XFS_DINODE_MIN_SIZE; + if (mp->m_sb.sb_inoalignmt >= XFS_B_TO_FSBT(mp, new_size)) + mp->m_inode_cluster_size = new_size; + } /* - * Initialize the attribute manager's entries. + * Set inode alignment fields */ - mp->m_attr_magicpct = (mp->m_sb.sb_blocksize * 37) / 100; + xfs_set_inoalignment(mp); /* - * Initialize the precomputed transaction reservations values. + * Check that the data (and log if separate) is an ok size. */ - xfs_trans_init(mp); + error = xfs_check_sizes(mp); + if (error) + goto out_remove_uuid; + + /* + * Initialize realtime fields in the mount structure + */ + error = xfs_rtmount_init(mp); + if (error) { + xfs_warn(mp, "RT mount failed"); + goto out_remove_uuid; + } + + /* + * 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); + + mp->m_dmevmask = 0; /* not persistent; set after each mount */ + + error = xfs_da_mount(mp); + if (error) { + xfs_warn(mp, "Failed dir/attr init: %d", error); + goto out_remove_uuid; + } /* - * Allocate and initialize the inode hash table for this - * file system. + * Initialize the precomputed transaction reservations values. */ - xfs_ihash_init(mp); - xfs_chash_init(mp); + xfs_trans_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); + spin_lock_init(&mp->m_perag_lock); + INIT_RADIX_TREE(&mp->m_perag_tree, GFP_ATOMIC); + error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi); + if (error) { + xfs_warn(mp, "Failed per-ag init: %d", error); + goto out_free_dir; + } - mp->m_maxagi = xfs_initialize_perag(vfsp, mp, sbp->sb_agcount); + if (!sbp->sb_logblocks) { + xfs_warn(mp, "no log defined"); + XFS_ERROR_REPORT("xfs_mountfs", XFS_ERRLEVEL_LOW, mp); + error = XFS_ERROR(EFSCORRUPTED); + goto out_free_perag; + } /* * 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; + 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) { + xfs_warn(mp, "log mount failed"); + goto out_fail_wait; } /* @@ -1064,38 +850,34 @@ xfs_mountfs( * If we are currently making the filesystem, the initialisation will * fail as the perag data is in an undefined state. */ - if (xfs_sb_version_haslazysbcount(&mp->m_sb) && !XFS_LAST_UNMOUNT_WAS_CLEAN(mp) && !mp->m_sb.sb_inprogress) { error = xfs_initialize_perag_data(mp, sbp->sb_agcount); - if (error) { - goto error2; - } + if (error) + goto out_fail_wait; } + /* * Get and sanity-check the root inode. * Save the pointer to it in the mount structure. */ - error = xfs_iget(mp, NULL, sbp->sb_rootino, 0, XFS_ILOCK_EXCL, &rip, 0); + error = xfs_iget(mp, NULL, sbp->sb_rootino, 0, XFS_ILOCK_EXCL, &rip); if (error) { - cmn_err(CE_WARN, "XFS: failed to read root inode"); - goto error3; + xfs_warn(mp, "failed to read root inode"); + goto out_log_dealloc; } 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"); - cmn_err(CE_WARN, "Device %s - root %llu is not a directory", - XFS_BUFTARG_NAME(mp->m_ddev_targp), + if (unlikely(!S_ISDIR(rip->i_d.di_mode))) { + xfs_warn(mp, "corrupted root inode %llu: not a directory", (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; + goto out_rele_rip; } mp->m_rootip = rip; /* save it */ @@ -1104,43 +886,71 @@ xfs_mountfs( /* * Initialize realtime inode pointers in the mount structure */ - if ((error = xfs_rtmount_inodes(mp))) { + error = xfs_rtmount_inodes(mp); + if (error) { /* * Free up the root inode. */ - cmn_err(CE_WARN, "XFS: failed to read RT inodes"); - goto error4; + xfs_warn(mp, "failed to read RT inodes"); + goto out_rele_rip; } /* - * If fs is not mounted readonly, then update the superblock - * unit and width changes. + * If this is a read-only mount defer the superblock updates until + * the next remount into writeable mode. Otherwise we would never + * perform the update e.g. for the root filesystem. */ - if (update_flags && !(vfsp->vfs_flag & VFS_RDONLY)) - xfs_mount_log_sbunit(mp, update_flags); + if (mp->m_update_flags && !(mp->m_flags & XFS_MOUNT_RDONLY)) { + error = xfs_mount_log_sb(mp, mp->m_update_flags); + if (error) { + xfs_warn(mp, "failed to write sb changes"); + goto out_rtunmount; + } + } /* * Initialise the XFS quota management subsystem for this mount */ - if ((error = XFS_QM_INIT(mp, "amount, "aflags))) - goto error4; + if (XFS_IS_QUOTA_RUNNING(mp)) { + error = xfs_qm_newmount(mp, "amount, "aflags); + if (error) + goto out_rtunmount; + } else { + ASSERT(!XFS_IS_QUOTA_ON(mp)); + + /* + * If a file system had quotas running earlier, but decided to + * mount without -o uquota/pquota/gquota options, revoke the + * quotachecked license. + */ + if (mp->m_sb.sb_qflags & XFS_ALL_QUOTA_ACCT) { + xfs_notice(mp, "resetting quota flags"); + error = xfs_mount_reset_sbqflags(mp); + if (error) + return error; + } + } /* * 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); + error = xfs_log_mount_finish(mp); if (error) { - cmn_err(CE_WARN, "XFS: log mount finish failed"); - goto error4; + xfs_warn(mp, "log mount finish failed"); + goto out_rtunmount; } /* * Complete the quota initialisation, post-log-replay component. */ - if ((error = XFS_QM_MOUNT(mp, quotamount, quotaflags, mfsi_flags))) - goto error4; + if (quotamount) { + ASSERT(mp->m_qflags == 0); + mp->m_qflags = quotaflags; + + xfs_qm_mount_quotas(mp); + } /* * Now we are mounted, reserve a small amount of unused space for @@ -1150,59 +960,59 @@ xfs_mountfs( * attr, unwritten extent conversion at ENOSPC, etc. Data allocations * are not allowed to use this reserved space. * - * We default to 5% or 1024 fsbs of space reserved, whichever is smaller. * This may drive us straight to ENOSPC on mount, but that implies - * we were already there on the last unmount. + * we were already there on the last unmount. Warn if this occurs. */ - resblks = mp->m_sb.sb_dblocks; - do_div(resblks, 20); - resblks = min_t(__uint64_t, resblks, 1024); - xfs_reserve_blocks(mp, &resblks, NULL); + if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { + resblks = xfs_default_resblks(mp); + error = xfs_reserve_blocks(mp, &resblks, NULL); + if (error) + xfs_warn(mp, + "Unable to allocate reserve blocks. Continuing without reserve pool."); + } 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, - sizeof(xfs_perag_busy_t) * XFS_PAGB_NUM_SLOTS); - kmem_free(mp->m_perag, sbp->sb_agcount * sizeof(xfs_perag_t)); - mp->m_perag = NULL; - /* FALLTHROUGH */ - error1: - if (uuid_mounted) - xfs_uuid_unmount(mp); - xfs_freesb(mp); + out_rtunmount: + xfs_rtunmount_inodes(mp); + out_rele_rip: + IRELE(rip); + out_log_dealloc: + xfs_log_unmount(mp); + out_fail_wait: + if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp) + xfs_wait_buftarg(mp->m_logdev_targp); + xfs_wait_buftarg(mp->m_ddev_targp); + out_free_perag: + xfs_free_perag(mp); + out_free_dir: + xfs_da_unmount(mp); + out_remove_uuid: + xfs_uuid_unmount(mp); + out: return error; } /* - * xfs_unmountfs - * * This flushes out the inodes,dquots and the superblock, unmounts the * log and makes sure that incore structures are freed. */ -int -xfs_unmountfs(xfs_mount_t *mp, struct cred *cr) +void +xfs_unmountfs( + struct xfs_mount *mp) { - struct bhv_vfs *vfsp = XFS_MTOVFS(mp); -#if defined(DEBUG) || defined(INDUCE_IO_ERROR) - int64_t fsid; -#endif - __uint64_t resblks; + __uint64_t resblks; + int error; + + cancel_delayed_work_sync(&mp->m_eofblocks_work); + + xfs_qm_unmount_quotas(mp); + xfs_rtunmount_inodes(mp); + IRELE(mp->m_rootip); /* * We can potentially deadlock here if we have an inode cluster - * that has been freed has it's buffer still pinned in memory because + * that has been freed has its buffer still pinned in memory because * the transaction is still sitting in a iclog. The stale inodes * on that buffer will have their flush locks held until the * transaction hits the disk and the callbacks run. the inode @@ -1210,28 +1020,29 @@ xfs_unmountfs(xfs_mount_t *mp, struct cred *cr) * push out the iclog we will never get that unlocked. hence we * need to force the log first. */ - xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC); - xfs_iflush_all(mp); + xfs_log_force(mp, XFS_LOG_SYNC); - XFS_QM_DQPURGEALL(mp, XFS_QMOPT_QUOTALL | XFS_QMOPT_UMOUNTING); + /* + * Flush all pending changes from the AIL. + */ + xfs_ail_push_all_sync(mp->m_ail); /* - * Flush out the log synchronously so that we know for sure - * that nothing is pinned. This is important because bflush() - * will skip pinned buffers. + * And reclaim all inodes. At this point there should be no dirty + * inodes and none should be pinned or locked, but use synchronous + * reclaim just to be sure. We can stop background inode reclaim + * here as well if it is still running. */ - xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC); + cancel_delayed_work_sync(&mp->m_reclaim_work); + xfs_reclaim_inodes(mp, SYNC_WAIT); - xfs_binval(mp->m_ddev_targp); - if (mp->m_rtdev_targp) { - xfs_binval(mp->m_rtdev_targp); - } + xfs_qm_unmount(mp); /* * Unreserve any blocks we have so that when we unmount we don't account * the reserved free space as used. This is really only necessary for * lazy superblock counting because it trusts the incore superblock - * counters to be aboslutely correct on clean unmount. + * counters to be absolutely correct on clean unmount. * * We don't bother correcting this elsewhere for lazy superblock * counting because on mount of an unclean filesystem we reconstruct the @@ -1242,80 +1053,44 @@ xfs_unmountfs(xfs_mount_t *mp, struct cred *cr) * value does not matter.... */ resblks = 0; - xfs_reserve_blocks(mp, &resblks, NULL); - - xfs_log_sbcount(mp, 1); - xfs_unmountfs_writesb(mp); - xfs_unmountfs_wait(mp); /* wait for async bufs */ - xfs_log_unmount(mp); /* Done! No more fs ops. */ - - xfs_freesb(mp); - - /* - * All inodes from this mount point should be freed. - */ - ASSERT(mp->m_inodes == NULL); - - xfs_unmountfs_close(mp, cr); - if ((mp->m_flags & XFS_MOUNT_NOUUID) == 0) - xfs_uuid_unmount(mp); - -#if defined(DEBUG) || defined(INDUCE_IO_ERROR) - /* - * clear all error tags on this filesystem - */ - memcpy(&fsid, &vfsp->vfs_fsid, sizeof(int64_t)); - xfs_errortag_clearall_umount(fsid, mp->m_fsname, 0); + error = xfs_reserve_blocks(mp, &resblks, NULL); + if (error) + xfs_warn(mp, "Unable to free reserved block pool. " + "Freespace may not be correct on next mount."); + + error = xfs_log_sbcount(mp); + if (error) + xfs_warn(mp, "Unable to update superblock counters. " + "Freespace may not be correct on next mount."); + + xfs_log_unmount(mp); + xfs_da_unmount(mp); + xfs_uuid_unmount(mp); + +#if defined(DEBUG) + xfs_errortag_clearall(mp, 0); #endif - XFS_IODONE(vfsp); - xfs_mount_free(mp, 1); - return 0; -} - -void -xfs_unmountfs_close(xfs_mount_t *mp, struct cred *cr) -{ - if (mp->m_logdev_targp != mp->m_ddev_targp) - xfs_free_buftarg(mp->m_logdev_targp, 1); - if (mp->m_rtdev_targp) - xfs_free_buftarg(mp->m_rtdev_targp, 1); - xfs_free_buftarg(mp->m_ddev_targp, 0); -} - -STATIC void -xfs_unmountfs_wait(xfs_mount_t *mp) -{ - if (mp->m_logdev_targp != mp->m_ddev_targp) - xfs_wait_buftarg(mp->m_logdev_targp); - if (mp->m_rtdev_targp) - xfs_wait_buftarg(mp->m_rtdev_targp); - xfs_wait_buftarg(mp->m_ddev_targp); + xfs_free_perag(mp); } int xfs_fs_writable(xfs_mount_t *mp) { - bhv_vfs_t *vfsp = XFS_MTOVFS(mp); - - return !(vfs_test_for_freeze(vfsp) || XFS_FORCED_SHUTDOWN(mp) || - (vfsp->vfs_flag & VFS_RDONLY)); + return !(mp->m_super->s_writers.frozen || XFS_FORCED_SHUTDOWN(mp) || + (mp->m_flags & XFS_MOUNT_RDONLY)); } /* * xfs_log_sbcount * - * Called either periodically to keep the on disk superblock values - * roughly up to date or from unmount to make sure the values are - * correct on a clean unmount. + * Sync the superblock counters to disk. * * Note this code can be called during the process of freezing, so - * we may need to use the transaction allocator which does not not + * we may need to use the transaction allocator which does not * block when the transaction subsystem is in its frozen state. */ int -xfs_log_sbcount( - xfs_mount_t *mp, - uint sync) +xfs_log_sbcount(xfs_mount_t *mp) { xfs_trans_t *tp; int error; @@ -1323,7 +1098,7 @@ xfs_log_sbcount( if (!xfs_fs_writable(mp)) return 0; - xfs_icsb_sync_counters(mp); + xfs_icsb_sync_counters(mp, 0); /* * we don't need to do this if we are updating the superblock @@ -1332,124 +1107,29 @@ xfs_log_sbcount( if (!xfs_sb_version_haslazysbcount(&mp->m_sb)) return 0; - tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT); - error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT); + tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT, KM_SLEEP); + error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0); if (error) { xfs_trans_cancel(tp, 0); return error; } xfs_mod_sb(tp, XFS_SB_IFREE | XFS_SB_ICOUNT | XFS_SB_FDBLOCKS); - if (sync) - xfs_trans_set_sync(tp); - xfs_trans_commit(tp, 0); - - return 0; -} - -int -xfs_unmountfs_writesb(xfs_mount_t *mp) -{ - xfs_buf_t *sbp; - xfs_sb_t *sb; - int error = 0; - - /* - * skip superblock write if fs is read-only, or - * if we are doing a forced umount. - */ - if (!(XFS_MTOVFS(mp)->vfs_flag & VFS_RDONLY || - XFS_FORCED_SHUTDOWN(mp))) { - - sbp = xfs_getsb(mp, 0); - sb = XFS_BUF_TO_SBP(sbp); - - /* - * mark shared-readonly if desired - */ - if (mp->m_mk_sharedro) { - if (!(sb->sb_flags & XFS_SBF_READONLY)) - sb->sb_flags |= XFS_SBF_READONLY; - if (!XFS_SB_VERSION_HASSHARED(sb)) - XFS_SB_VERSION_ADDSHARED(sb); - xfs_fs_cmn_err(CE_NOTE, mp, - "Unmounting, marking shared read-only"); - } - - XFS_BUF_UNDONE(sbp); - XFS_BUF_UNREAD(sbp); - XFS_BUF_UNDELAYWRITE(sbp); - XFS_BUF_WRITE(sbp); - XFS_BUF_UNASYNC(sbp); - ASSERT(XFS_BUF_TARGET(sbp) == mp->m_ddev_targp); - xfsbdstrat(mp, sbp); - /* Nevermind errors we might get here. */ - error = xfs_iowait(sbp); - if (error) - xfs_ioerror_alert("xfs_unmountfs_writesb", - mp, sbp, XFS_BUF_ADDR(sbp)); - if (error && mp->m_mk_sharedro) - xfs_fs_cmn_err(CE_ALERT, mp, "Superblock write error detected while unmounting. Filesystem may not be marked shared readonly"); - xfs_buf_relse(sbp); - } + xfs_trans_set_sync(tp); + error = xfs_trans_commit(tp, 0); return error; } /* - * xfs_mod_sb() can be used to copy arbitrary changes to the - * in-core superblock into the superblock buffer to be logged. - * It does not provide the higher level of locking that is - * needed to protect the in-core superblock from concurrent - * access. - */ -void -xfs_mod_sb(xfs_trans_t *tp, __int64_t fields) -{ - xfs_buf_t *bp; - int first; - int last; - xfs_mount_t *mp; - xfs_sb_t *sbp; - xfs_sb_field_t f; - - ASSERT(fields); - if (!fields) - return; - mp = tp->t_mountp; - bp = xfs_trans_getsb(tp, mp, 0); - sbp = XFS_BUF_TO_SBP(bp); - first = sizeof(xfs_sb_t); - last = 0; - - /* translate/copy */ - - xfs_xlatesb(XFS_BUF_PTR(bp), &(mp->m_sb), -1, fields); - - /* find modified range */ - - f = (xfs_sb_field_t)xfs_lowbit64((__uint64_t)fields); - ASSERT((1LL << f) & XFS_SB_MOD_BITS); - first = xfs_sb_info[f].offset; - - f = (xfs_sb_field_t)xfs_highbit64((__uint64_t)fields); - ASSERT((1LL << f) & XFS_SB_MOD_BITS); - last = xfs_sb_info[f + 1].offset - 1; - - xfs_trans_log_buf(tp, bp, first, last); -} - - -/* - * xfs_mod_incore_sb_unlocked() is a utility routine common used to apply + * xfs_mod_incore_sb_unlocked() is a utility routine commonly used to apply * a delta to a specified field in the in-core superblock. Simply * switch on the field indicated and apply the delta to that field. * Fields are not allowed to dip below zero, so if the delta would * do this do not apply it and return EINVAL. * - * The SB_LOCK must be held when this routine is called. + * The m_sb_lock must be held when this routine is called. */ -int +STATIC int xfs_mod_incore_sb_unlocked( xfs_mount_t *mp, xfs_sb_field_t field, @@ -1499,26 +1179,30 @@ xfs_mod_incore_sb_unlocked( lcounter += rem; } } else { /* Taking blocks away */ - lcounter += delta; + if (lcounter >= 0) { + mp->m_sb.sb_fdblocks = lcounter + + XFS_ALLOC_SET_ASIDE(mp); + return 0; + } - /* - * If were out of blocks, use any available reserved blocks if - * were allowed to. - */ + /* + * We are out of blocks, use any available reserved + * blocks if were allowed to. + */ + if (!rsvd) + return XFS_ERROR(ENOSPC); - if (lcounter < 0) { - if (rsvd) { - lcounter = (long long)mp->m_resblks_avail + delta; - if (lcounter < 0) { - return XFS_ERROR(ENOSPC); - } - mp->m_resblks_avail = lcounter; - return 0; - } else { /* not reserved */ - return XFS_ERROR(ENOSPC); - } + lcounter = (long long)mp->m_resblks_avail + delta; + if (lcounter >= 0) { + mp->m_resblks_avail = lcounter; + return 0; } + printk_once(KERN_WARNING + "Filesystem \"%s\": reserve blocks depleted! " + "Consider increasing reserve pool size.", + mp->m_fsname); + return XFS_ERROR(ENOSPC); } mp->m_sb.sb_fdblocks = lcounter + XFS_ALLOC_SET_ASIDE(mp); @@ -1612,142 +1296,77 @@ xfs_mod_incore_sb_unlocked( /* * xfs_mod_incore_sb() is used to change a field in the in-core * superblock structure by the specified delta. This modification - * is protected by the SB_LOCK. Just use the xfs_mod_incore_sb_unlocked() + * is protected by the m_sb_lock. Just use the xfs_mod_incore_sb_unlocked() * routine to do the work. */ int xfs_mod_incore_sb( - xfs_mount_t *mp, - xfs_sb_field_t field, - int64_t delta, - int rsvd) + struct xfs_mount *mp, + xfs_sb_field_t field, + int64_t delta, + int rsvd) { - unsigned long s; - int status; + int status; - /* check for per-cpu counters */ - switch (field) { #ifdef HAVE_PERCPU_SB - case XFS_SBS_ICOUNT: - case XFS_SBS_IFREE: - case XFS_SBS_FDBLOCKS: - if (!(mp->m_flags & XFS_MOUNT_NO_PERCPU_SB)) { - status = xfs_icsb_modify_counters(mp, field, - delta, rsvd); - break; - } - /* FALLTHROUGH */ + ASSERT(field < XFS_SBS_ICOUNT || field > XFS_SBS_FDBLOCKS); #endif - default: - s = XFS_SB_LOCK(mp); - status = xfs_mod_incore_sb_unlocked(mp, field, delta, rsvd); - XFS_SB_UNLOCK(mp, s); - break; - } + spin_lock(&mp->m_sb_lock); + status = xfs_mod_incore_sb_unlocked(mp, field, delta, rsvd); + spin_unlock(&mp->m_sb_lock); return status; } /* - * xfs_mod_incore_sb_batch() is used to change more than one field - * in the in-core superblock structure at a time. This modification - * is protected by a lock internal to this module. The fields and - * changes to those fields are specified in the array of xfs_mod_sb - * structures passed in. + * Change more than one field in the in-core superblock structure at a time. * - * Either all of the specified deltas will be applied or none of - * them will. If any modified field dips below 0, then all modifications - * will be backed out and EINVAL will be returned. + * The fields and changes to those fields are specified in the array of + * xfs_mod_sb structures passed in. Either all of the specified deltas + * will be applied or none of them will. If any modified field dips below 0, + * then all modifications will be backed out and EINVAL will be returned. + * + * Note that this function may not be used for the superblock values that + * are tracked with the in-memory per-cpu counters - a direct call to + * xfs_icsb_modify_counters is required for these. */ int -xfs_mod_incore_sb_batch(xfs_mount_t *mp, xfs_mod_sb_t *msb, uint nmsb, int rsvd) +xfs_mod_incore_sb_batch( + struct xfs_mount *mp, + xfs_mod_sb_t *msb, + uint nmsb, + int rsvd) { - unsigned long s; - int status=0; - xfs_mod_sb_t *msbp; + xfs_mod_sb_t *msbp; + int error = 0; /* - * Loop through the array of mod structures and apply each - * individually. If any fail, then back out all those - * which have already been applied. Do all of this within - * the scope of the SB_LOCK so that all of the changes will - * be atomic. + * Loop through the array of mod structures and apply each individually. + * If any fail, then back out all those which have already been applied. + * Do all of this within the scope of the m_sb_lock so that all of the + * changes will be atomic. */ - s = XFS_SB_LOCK(mp); - msbp = &msb[0]; - for (msbp = &msbp[0]; msbp < (msb + nmsb); msbp++) { - /* - * Apply the delta at index n. If it fails, break - * from the loop so we'll fall into the undo loop - * below. - */ - switch (msbp->msb_field) { -#ifdef HAVE_PERCPU_SB - case XFS_SBS_ICOUNT: - case XFS_SBS_IFREE: - case XFS_SBS_FDBLOCKS: - if (!(mp->m_flags & XFS_MOUNT_NO_PERCPU_SB)) { - XFS_SB_UNLOCK(mp, s); - status = xfs_icsb_modify_counters(mp, - msbp->msb_field, - msbp->msb_delta, rsvd); - s = XFS_SB_LOCK(mp); - break; - } - /* FALLTHROUGH */ -#endif - default: - status = xfs_mod_incore_sb_unlocked(mp, - msbp->msb_field, - msbp->msb_delta, rsvd); - break; - } + spin_lock(&mp->m_sb_lock); + for (msbp = msb; msbp < (msb + nmsb); msbp++) { + ASSERT(msbp->msb_field < XFS_SBS_ICOUNT || + msbp->msb_field > XFS_SBS_FDBLOCKS); - if (status != 0) { - break; - } + error = xfs_mod_incore_sb_unlocked(mp, msbp->msb_field, + msbp->msb_delta, rsvd); + if (error) + goto unwind; } + spin_unlock(&mp->m_sb_lock); + return 0; - /* - * If we didn't complete the loop above, then back out - * any changes made to the superblock. If you add code - * between the loop above and here, make sure that you - * preserve the value of status. Loop back until - * we step below the beginning of the array. Make sure - * we don't touch anything back there. - */ - if (status != 0) { - msbp--; - while (msbp >= msb) { - switch (msbp->msb_field) { -#ifdef HAVE_PERCPU_SB - case XFS_SBS_ICOUNT: - case XFS_SBS_IFREE: - case XFS_SBS_FDBLOCKS: - if (!(mp->m_flags & XFS_MOUNT_NO_PERCPU_SB)) { - XFS_SB_UNLOCK(mp, s); - status = xfs_icsb_modify_counters(mp, - msbp->msb_field, - -(msbp->msb_delta), - rsvd); - s = XFS_SB_LOCK(mp); - break; - } - /* FALLTHROUGH */ -#endif - default: - status = xfs_mod_incore_sb_unlocked(mp, - msbp->msb_field, - -(msbp->msb_delta), - rsvd); - break; - } - ASSERT(status == 0); - msbp--; - } +unwind: + while (--msbp >= msb) { + error = xfs_mod_incore_sb_unlocked(mp, msbp->msb_field, + -msbp->msb_delta, rsvd); + ASSERT(error == 0); } - XFS_SB_UNLOCK(mp, s); - return status; + spin_unlock(&mp->m_sb_lock); + return error; } /* @@ -1759,23 +1378,20 @@ xfs_mod_incore_sb_batch(xfs_mount_t *mp, xfs_mod_sb_t *msb, uint nmsb, int rsvd) * the superblock buffer if it can be locked without sleeping. * If it can't then we'll return NULL. */ -xfs_buf_t * +struct xfs_buf * xfs_getsb( - xfs_mount_t *mp, - int flags) + struct xfs_mount *mp, + int flags) { - xfs_buf_t *bp; + struct xfs_buf *bp = mp->m_sb_bp; - ASSERT(mp->m_sb_bp != NULL); - bp = mp->m_sb_bp; - if (flags & XFS_BUF_TRYLOCK) { - if (!XFS_BUF_CPSEMA(bp)) { + if (!xfs_buf_trylock(bp)) { + if (flags & XBF_TRYLOCK) return NULL; - } - } else { - XFS_BUF_PSEMA(bp, PRIBIO); + xfs_buf_lock(bp); } - XFS_BUF_HOLD(bp); + + xfs_buf_hold(bp); ASSERT(XFS_BUF_ISDONE(bp)); return bp; } @@ -1785,76 +1401,61 @@ xfs_getsb( */ void xfs_freesb( - xfs_mount_t *mp) + struct xfs_mount *mp) { - xfs_buf_t *bp; + struct xfs_buf *bp = mp->m_sb_bp; - /* - * Use xfs_getsb() so that the buffer will be locked - * when we call xfs_buf_relse(). - */ - bp = xfs_getsb(mp, 0); - XFS_BUF_UNMANAGE(bp); - xfs_buf_relse(bp); + xfs_buf_lock(bp); mp->m_sb_bp = NULL; -} - -/* - * See if the UUID is unique among mounted XFS filesystems. - * Mount fails if UUID is nil or a FS with the same UUID is already mounted. - */ -STATIC int -xfs_uuid_mount( - xfs_mount_t *mp) -{ - if (uuid_is_nil(&mp->m_sb.sb_uuid)) { - cmn_err(CE_WARN, - "XFS: Filesystem %s has nil UUID - can't mount", - mp->m_fsname); - return -1; - } - if (!uuid_table_insert(&mp->m_sb.sb_uuid)) { - cmn_err(CE_WARN, - "XFS: Filesystem %s has duplicate UUID - can't mount", - mp->m_fsname); - return -1; - } - return 0; -} - -/* - * Remove filesystem from the UUID table. - */ -STATIC void -xfs_uuid_unmount( - xfs_mount_t *mp) -{ - uuid_table_remove(&mp->m_sb.sb_uuid); + xfs_buf_relse(bp); } /* * Used to log changes to the superblock unit and width fields which could - * be altered by the mount options. Only the first superblock is updated. + * be altered by the mount options, as well as any potential sb_features2 + * fixup. Only the first superblock is updated. */ -STATIC void -xfs_mount_log_sbunit( +int +xfs_mount_log_sb( xfs_mount_t *mp, __int64_t fields) { xfs_trans_t *tp; + int error; - ASSERT(fields & (XFS_SB_UNIT|XFS_SB_WIDTH|XFS_SB_UUID)); + ASSERT(fields & (XFS_SB_UNIT | XFS_SB_WIDTH | XFS_SB_UUID | + XFS_SB_FEATURES2 | XFS_SB_BAD_FEATURES2 | + XFS_SB_VERSIONNUM)); tp = xfs_trans_alloc(mp, XFS_TRANS_SB_UNIT); - if (xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT)) { + error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0); + if (error) { xfs_trans_cancel(tp, 0); - return; + return error; } xfs_mod_sb(tp, fields); - xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp, 0); + return error; } +/* + * If the underlying (data/log/rt) device is readonly, there are some + * operations that cannot proceed. + */ +int +xfs_dev_is_read_only( + struct xfs_mount *mp, + char *message) +{ + if (xfs_readonly_buftarg(mp->m_ddev_targp) || + xfs_readonly_buftarg(mp->m_logdev_targp) || + (mp->m_rtdev_targp && xfs_readonly_buftarg(mp->m_rtdev_targp))) { + xfs_notice(mp, "%s required on read-only device.", message); + xfs_notice(mp, "write access unavailable, cannot proceed."); + return EROFS; + } + return 0; +} #ifdef HAVE_PERCPU_SB /* @@ -1879,7 +1480,7 @@ xfs_mount_log_sbunit( * we disable the per-cpu counter and go through the slow path. * * The slow path is the current xfs_mod_incore_sb() function. This means that - * when we disable a per-cpu counter, we need to drain it's resources back to + * when we disable a per-cpu counter, we need to drain its resources back to * the global superblock. We do this after disabling the counter to prevent * more threads from queueing up on the counter. * @@ -1894,12 +1495,12 @@ xfs_mount_log_sbunit( * * Locking rules: * - * 1. XFS_SB_LOCK() before picking up per-cpu locks + * 1. m_sb_lock before picking up per-cpu locks * 2. per-cpu locks always picked up via for_each_online_cpu() order - * 3. accurate counter sync requires XFS_SB_LOCK + per cpu locks + * 3. accurate counter sync requires m_sb_lock + per cpu locks * 4. modifying per-cpu counters requires holding per-cpu lock - * 5. modifying global counters requires holding XFS_SB_LOCK - * 6. enabling or disabling a counter requires holding the XFS_SB_LOCK + * 5. modifying global counters requires holding m_sb_lock + * 6. enabling or disabling a counter requires holding the m_sb_lock * and _none_ of the per-cpu locks. * * Disabled counters are only ever re-enabled by a balance operation @@ -1926,7 +1527,6 @@ xfs_icsb_cpu_notify( { xfs_icsb_cnts_t *cntp; xfs_mount_t *mp; - int s; mp = (xfs_mount_t *)container_of(nfb, xfs_mount_t, m_icsb_notifier); cntp = (xfs_icsb_cnts_t *) @@ -1941,9 +1541,9 @@ xfs_icsb_cpu_notify( case CPU_ONLINE: case CPU_ONLINE_FROZEN: xfs_icsb_lock(mp); - xfs_icsb_balance_counter(mp, XFS_SBS_ICOUNT, 0, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_IFREE, 0, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_FDBLOCKS, 0, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_ICOUNT, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_IFREE, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_FDBLOCKS, 0); xfs_icsb_unlock(mp); break; case CPU_DEAD: @@ -1952,7 +1552,7 @@ xfs_icsb_cpu_notify( * count into the total on the global superblock and * re-enable the counters. */ xfs_icsb_lock(mp); - s = XFS_SB_LOCK(mp); + spin_lock(&mp->m_sb_lock); xfs_icsb_disable_counter(mp, XFS_SBS_ICOUNT); xfs_icsb_disable_counter(mp, XFS_SBS_IFREE); xfs_icsb_disable_counter(mp, XFS_SBS_FDBLOCKS); @@ -1963,13 +1563,10 @@ xfs_icsb_cpu_notify( memset(cntp, 0, sizeof(xfs_icsb_cnts_t)); - xfs_icsb_balance_counter(mp, XFS_SBS_ICOUNT, - XFS_ICSB_SB_LOCKED, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_IFREE, - XFS_ICSB_SB_LOCKED, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_FDBLOCKS, - XFS_ICSB_SB_LOCKED, 0); - XFS_SB_UNLOCK(mp, s); + xfs_icsb_balance_counter_locked(mp, XFS_SBS_ICOUNT, 0); + xfs_icsb_balance_counter_locked(mp, XFS_SBS_IFREE, 0); + xfs_icsb_balance_counter_locked(mp, XFS_SBS_FDBLOCKS, 0); + spin_unlock(&mp->m_sb_lock); xfs_icsb_unlock(mp); break; } @@ -1989,12 +1586,6 @@ xfs_icsb_init_counters( if (mp->m_sb_cnts == NULL) return -ENOMEM; -#ifdef CONFIG_HOTPLUG_CPU - mp->m_icsb_notifier.notifier_call = xfs_icsb_cpu_notify; - mp->m_icsb_notifier.priority = 0; - register_hotcpu_notifier(&mp->m_icsb_notifier); -#endif /* CONFIG_HOTPLUG_CPU */ - for_each_online_cpu(i) { cntp = (xfs_icsb_cnts_t *)per_cpu_ptr(mp->m_sb_cnts, i); memset(cntp, 0, sizeof(xfs_icsb_cnts_t)); @@ -2007,6 +1598,13 @@ xfs_icsb_init_counters( * initial balance kicks us off correctly */ mp->m_icsb_counters = -1; + +#ifdef CONFIG_HOTPLUG_CPU + mp->m_icsb_notifier.notifier_call = xfs_icsb_cpu_notify; + mp->m_icsb_notifier.priority = 0; + register_hotcpu_notifier(&mp->m_icsb_notifier); +#endif /* CONFIG_HOTPLUG_CPU */ + return 0; } @@ -2020,13 +1618,13 @@ xfs_icsb_reinit_counters( * initial balance kicks us off correctly */ mp->m_icsb_counters = -1; - xfs_icsb_balance_counter(mp, XFS_SBS_ICOUNT, 0, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_IFREE, 0, 0); - xfs_icsb_balance_counter(mp, XFS_SBS_FDBLOCKS, 0, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_ICOUNT, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_IFREE, 0); + xfs_icsb_balance_counter(mp, XFS_SBS_FDBLOCKS, 0); xfs_icsb_unlock(mp); } -STATIC void +void xfs_icsb_destroy_counters( xfs_mount_t *mp) { @@ -2037,7 +1635,7 @@ xfs_icsb_destroy_counters( mutex_destroy(&mp->m_icsb_mutex); } -STATIC_INLINE void +STATIC void xfs_icsb_lock_cntr( xfs_icsb_cnts_t *icsbp) { @@ -2046,7 +1644,7 @@ xfs_icsb_lock_cntr( } } -STATIC_INLINE void +STATIC void xfs_icsb_unlock_cntr( xfs_icsb_cnts_t *icsbp) { @@ -2054,7 +1652,7 @@ xfs_icsb_unlock_cntr( } -STATIC_INLINE void +STATIC void xfs_icsb_lock_all_counters( xfs_mount_t *mp) { @@ -2067,7 +1665,7 @@ xfs_icsb_lock_all_counters( } } -STATIC_INLINE void +STATIC void xfs_icsb_unlock_all_counters( xfs_mount_t *mp) { @@ -2114,7 +1712,7 @@ xfs_icsb_counter_disabled( return test_bit(field, &mp->m_icsb_counters); } -STATIC int +STATIC void xfs_icsb_disable_counter( xfs_mount_t *mp, xfs_sb_field_t field) @@ -2132,13 +1730,13 @@ xfs_icsb_disable_counter( * the m_icsb_mutex. */ if (xfs_icsb_counter_disabled(mp, field)) - return 0; + return; xfs_icsb_lock_all_counters(mp); if (!test_and_set_bit(field, &mp->m_icsb_counters)) { /* drain back to superblock */ - xfs_icsb_count(mp, &cnt, XFS_ICSB_SB_LOCKED|XFS_ICSB_LAZY_COUNT); + xfs_icsb_count(mp, &cnt, XFS_ICSB_LAZY_COUNT); switch(field) { case XFS_SBS_ICOUNT: mp->m_sb.sb_icount = cnt.icsb_icount; @@ -2155,8 +1753,6 @@ xfs_icsb_disable_counter( } xfs_icsb_unlock_all_counters(mp); - - return 0; } STATIC void @@ -2195,39 +1791,33 @@ xfs_icsb_enable_counter( } void -xfs_icsb_sync_counters_flags( +xfs_icsb_sync_counters_locked( xfs_mount_t *mp, int flags) { xfs_icsb_cnts_t cnt; - int s; - - /* Pass 1: lock all counters */ - if ((flags & XFS_ICSB_SB_LOCKED) == 0) - s = XFS_SB_LOCK(mp); xfs_icsb_count(mp, &cnt, flags); - /* Step 3: update mp->m_sb fields */ if (!xfs_icsb_counter_disabled(mp, XFS_SBS_ICOUNT)) mp->m_sb.sb_icount = cnt.icsb_icount; if (!xfs_icsb_counter_disabled(mp, XFS_SBS_IFREE)) mp->m_sb.sb_ifree = cnt.icsb_ifree; if (!xfs_icsb_counter_disabled(mp, XFS_SBS_FDBLOCKS)) mp->m_sb.sb_fdblocks = cnt.icsb_fdblocks; - - if ((flags & XFS_ICSB_SB_LOCKED) == 0) - XFS_SB_UNLOCK(mp, s); } /* * Accurate update of per-cpu counters to incore superblock */ -STATIC void +void xfs_icsb_sync_counters( - xfs_mount_t *mp) + xfs_mount_t *mp, + int flags) { - xfs_icsb_sync_counters_flags(mp, 0); + spin_lock(&mp->m_sb_lock); + xfs_icsb_sync_counters_locked(mp, flags); + spin_unlock(&mp->m_sb_lock); } /* @@ -2250,20 +1840,15 @@ xfs_icsb_sync_counters( #define XFS_ICSB_FDBLK_CNTR_REENABLE(mp) \ (uint64_t)(512 + XFS_ALLOC_SET_ASIDE(mp)) STATIC void -xfs_icsb_balance_counter( +xfs_icsb_balance_counter_locked( xfs_mount_t *mp, xfs_sb_field_t field, - int flags, int min_per_cpu) { uint64_t count, resid; int weight = num_online_cpus(); - int s; uint64_t min = (uint64_t)min_per_cpu; - if (!(flags & XFS_ICSB_SB_LOCKED)) - s = XFS_SB_LOCK(mp); - /* disable counter and sync counter */ xfs_icsb_disable_counter(mp, field); @@ -2273,19 +1858,19 @@ xfs_icsb_balance_counter( count = mp->m_sb.sb_icount; resid = do_div(count, weight); if (count < max(min, XFS_ICSB_INO_CNTR_REENABLE)) - goto out; + return; break; case XFS_SBS_IFREE: count = mp->m_sb.sb_ifree; resid = do_div(count, weight); if (count < max(min, XFS_ICSB_INO_CNTR_REENABLE)) - goto out; + return; break; case XFS_SBS_FDBLOCKS: count = mp->m_sb.sb_fdblocks; resid = do_div(count, weight); if (count < max(min, XFS_ICSB_FDBLK_CNTR_REENABLE(mp))) - goto out; + return; break; default: BUG(); @@ -2294,9 +1879,17 @@ xfs_icsb_balance_counter( } xfs_icsb_enable_counter(mp, field, count, resid); -out: - if (!(flags & XFS_ICSB_SB_LOCKED)) - XFS_SB_UNLOCK(mp, s); +} + +STATIC void +xfs_icsb_balance_counter( + xfs_mount_t *mp, + xfs_sb_field_t fields, + int min_per_cpu) +{ + spin_lock(&mp->m_sb_lock); + xfs_icsb_balance_counter_locked(mp, fields, min_per_cpu); + spin_unlock(&mp->m_sb_lock); } int @@ -2308,12 +1901,12 @@ xfs_icsb_modify_counters( { xfs_icsb_cnts_t *icsbp; long long lcounter; /* long counter for 64 bit fields */ - int cpu, ret = 0, s; + int ret = 0; might_sleep(); again: - cpu = get_cpu(); - icsbp = (xfs_icsb_cnts_t *)per_cpu_ptr(mp->m_sb_cnts, cpu); + preempt_disable(); + icsbp = this_cpu_ptr(mp->m_sb_cnts); /* * if the counter is disabled, go to slow path @@ -2357,11 +1950,11 @@ again: break; } xfs_icsb_unlock_cntr(icsbp); - put_cpu(); + preempt_enable(); return 0; slow_path: - put_cpu(); + preempt_enable(); /* * serialise with a mutex so we don't burn lots of cpu on @@ -2386,15 +1979,15 @@ slow_path: * running atomically here, we know a rebalance cannot * be in progress. Hence we can go straight to operating * on the global superblock. We do not call xfs_mod_incore_sb() - * here even though we need to get the SB_LOCK. Doing so + * here even though we need to get the m_sb_lock. Doing so * will cause us to re-enter this function and deadlock. - * Hence we get the SB_LOCK ourselves and then call + * Hence we get the m_sb_lock ourselves and then call * xfs_mod_incore_sb_unlocked() as the unlocked path operates * directly on the global counters. */ - s = XFS_SB_LOCK(mp); + spin_lock(&mp->m_sb_lock); ret = xfs_mod_incore_sb_unlocked(mp, field, delta, rsvd); - XFS_SB_UNLOCK(mp, s); + spin_unlock(&mp->m_sb_lock); /* * Now that we've modified the global superblock, we @@ -2403,13 +1996,13 @@ slow_path: * we are done. */ if (ret != ENOSPC) - xfs_icsb_balance_counter(mp, field, 0, 0); + xfs_icsb_balance_counter(mp, field, 0); xfs_icsb_unlock(mp); return ret; balance_counter: xfs_icsb_unlock_cntr(icsbp); - put_cpu(); + preempt_enable(); /* * We may have multiple threads here if multiple per-cpu @@ -2427,7 +2020,7 @@ balance_counter: * will either succeed through the fast path or slow path without * another balance operation being required. */ - xfs_icsb_balance_counter(mp, field, 0, delta); + xfs_icsb_balance_counter(mp, field, delta); xfs_icsb_unlock(mp); goto again; } |