diff options
Diffstat (limited to 'fs/xfs/xfs_ialloc.c')
| -rw-r--r-- | fs/xfs/xfs_ialloc.c | 2545 |
1 files changed, 1650 insertions, 895 deletions
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index db9d5fa600a..5960e5593fe 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -17,96 +17,337 @@ */ #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_inode.h" #include "xfs_btree.h" #include "xfs_ialloc.h" +#include "xfs_ialloc_btree.h" #include "xfs_alloc.h" #include "xfs_rtalloc.h" #include "xfs_error.h" #include "xfs_bmap.h" +#include "xfs_cksum.h" +#include "xfs_trans.h" +#include "xfs_buf_item.h" +#include "xfs_icreate_item.h" +#include "xfs_icache.h" +#include "xfs_dinode.h" +#include "xfs_trace.h" + /* - * Log specified fields for the inode given by bp and off. + * Allocation group level functions. */ -STATIC void -xfs_ialloc_log_di( - xfs_trans_t *tp, /* transaction pointer */ - xfs_buf_t *bp, /* inode buffer */ - int off, /* index of inode in buffer */ - int fields) /* bitmask of fields to log */ +static inline int +xfs_ialloc_cluster_alignment( + xfs_alloc_arg_t *args) { - int first; /* first byte number */ - int ioffset; /* off in bytes */ - int last; /* last byte number */ - xfs_mount_t *mp; /* mount point structure */ - static const short offsets[] = { /* field offsets */ - /* keep in sync with bits */ - offsetof(xfs_dinode_core_t, di_magic), - offsetof(xfs_dinode_core_t, di_mode), - offsetof(xfs_dinode_core_t, di_version), - offsetof(xfs_dinode_core_t, di_format), - offsetof(xfs_dinode_core_t, di_onlink), - offsetof(xfs_dinode_core_t, di_uid), - offsetof(xfs_dinode_core_t, di_gid), - offsetof(xfs_dinode_core_t, di_nlink), - offsetof(xfs_dinode_core_t, di_projid), - offsetof(xfs_dinode_core_t, di_pad), - offsetof(xfs_dinode_core_t, di_atime), - offsetof(xfs_dinode_core_t, di_mtime), - offsetof(xfs_dinode_core_t, di_ctime), - offsetof(xfs_dinode_core_t, di_size), - offsetof(xfs_dinode_core_t, di_nblocks), - offsetof(xfs_dinode_core_t, di_extsize), - offsetof(xfs_dinode_core_t, di_nextents), - offsetof(xfs_dinode_core_t, di_anextents), - offsetof(xfs_dinode_core_t, di_forkoff), - offsetof(xfs_dinode_core_t, di_aformat), - offsetof(xfs_dinode_core_t, di_dmevmask), - offsetof(xfs_dinode_core_t, di_dmstate), - offsetof(xfs_dinode_core_t, di_flags), - offsetof(xfs_dinode_core_t, di_gen), - offsetof(xfs_dinode_t, di_next_unlinked), - offsetof(xfs_dinode_t, di_u), - offsetof(xfs_dinode_t, di_a), - sizeof(xfs_dinode_t) - }; + if (xfs_sb_version_hasalign(&args->mp->m_sb) && + args->mp->m_sb.sb_inoalignmt >= + XFS_B_TO_FSBT(args->mp, args->mp->m_inode_cluster_size)) + return args->mp->m_sb.sb_inoalignmt; + return 1; +} +/* + * Lookup a record by ino in the btree given by cur. + */ +int /* error */ +xfs_inobt_lookup( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t ino, /* starting inode of chunk */ + xfs_lookup_t dir, /* <=, >=, == */ + int *stat) /* success/failure */ +{ + cur->bc_rec.i.ir_startino = ino; + cur->bc_rec.i.ir_freecount = 0; + cur->bc_rec.i.ir_free = 0; + return xfs_btree_lookup(cur, dir, stat); +} + +/* + * Update the record referred to by cur to the value given. + * This either works (return 0) or gets an EFSCORRUPTED error. + */ +STATIC int /* error */ +xfs_inobt_update( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_inobt_rec_incore_t *irec) /* btree record */ +{ + union xfs_btree_rec rec; + + rec.inobt.ir_startino = cpu_to_be32(irec->ir_startino); + rec.inobt.ir_freecount = cpu_to_be32(irec->ir_freecount); + rec.inobt.ir_free = cpu_to_be64(irec->ir_free); + return xfs_btree_update(cur, &rec); +} + +/* + * Get the data from the pointed-to record. + */ +int /* error */ +xfs_inobt_get_rec( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_inobt_rec_incore_t *irec, /* btree record */ + int *stat) /* output: success/failure */ +{ + union xfs_btree_rec *rec; + int error; + + error = xfs_btree_get_rec(cur, &rec, stat); + if (!error && *stat == 1) { + irec->ir_startino = be32_to_cpu(rec->inobt.ir_startino); + irec->ir_freecount = be32_to_cpu(rec->inobt.ir_freecount); + irec->ir_free = be64_to_cpu(rec->inobt.ir_free); + } + return error; +} + +/* + * Insert a single inobt record. Cursor must already point to desired location. + */ +STATIC int +xfs_inobt_insert_rec( + struct xfs_btree_cur *cur, + __int32_t freecount, + xfs_inofree_t free, + int *stat) +{ + cur->bc_rec.i.ir_freecount = freecount; + cur->bc_rec.i.ir_free = free; + return xfs_btree_insert(cur, stat); +} + +/* + * Insert records describing a newly allocated inode chunk into the inobt. + */ +STATIC int +xfs_inobt_insert( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct xfs_buf *agbp, + xfs_agino_t newino, + xfs_agino_t newlen, + xfs_btnum_t btnum) +{ + struct xfs_btree_cur *cur; + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + xfs_agino_t thisino; + int i; + int error; + + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, btnum); + + for (thisino = newino; + thisino < newino + newlen; + thisino += XFS_INODES_PER_CHUNK) { + error = xfs_inobt_lookup(cur, thisino, XFS_LOOKUP_EQ, &i); + if (error) { + xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + return error; + } + ASSERT(i == 0); + + error = xfs_inobt_insert_rec(cur, XFS_INODES_PER_CHUNK, + XFS_INOBT_ALL_FREE, &i); + if (error) { + xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + return error; + } + ASSERT(i == 1); + } + + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + + return 0; +} + +/* + * Verify that the number of free inodes in the AGI is correct. + */ +#ifdef DEBUG +STATIC int +xfs_check_agi_freecount( + struct xfs_btree_cur *cur, + struct xfs_agi *agi) +{ + if (cur->bc_nlevels == 1) { + xfs_inobt_rec_incore_t rec; + int freecount = 0; + int error; + int i; + + error = xfs_inobt_lookup(cur, 0, XFS_LOOKUP_GE, &i); + if (error) + return error; + + do { + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) + return error; + + if (i) { + freecount += rec.ir_freecount; + error = xfs_btree_increment(cur, 0, &i); + if (error) + return error; + } + } while (i == 1); + + if (!XFS_FORCED_SHUTDOWN(cur->bc_mp)) + ASSERT(freecount == be32_to_cpu(agi->agi_freecount)); + } + return 0; +} +#else +#define xfs_check_agi_freecount(cur, agi) 0 +#endif + +/* + * Initialise a new set of inodes. When called without a transaction context + * (e.g. from recovery) we initiate a delayed write of the inode buffers rather + * than logging them (which in a transaction context puts them into the AIL + * for writeback rather than the xfsbufd queue). + */ +int +xfs_ialloc_inode_init( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct list_head *buffer_list, + xfs_agnumber_t agno, + xfs_agblock_t agbno, + xfs_agblock_t length, + unsigned int gen) +{ + struct xfs_buf *fbuf; + struct xfs_dinode *free; + int nbufs, blks_per_cluster, inodes_per_cluster; + int version; + int i, j; + xfs_daddr_t d; + xfs_ino_t ino = 0; - ASSERT(offsetof(xfs_dinode_t, di_core) == 0); - ASSERT((fields & (XFS_DI_U|XFS_DI_A)) == 0); - mp = tp->t_mountp; /* - * Get the inode-relative first and last bytes for these fields + * Loop over the new block(s), filling in the inodes. For small block + * sizes, manipulate the inodes in buffers which are multiples of the + * blocks size. */ - xfs_btree_offsets(fields, offsets, XFS_DI_NUM_BITS, &first, &last); + blks_per_cluster = xfs_icluster_size_fsb(mp); + inodes_per_cluster = blks_per_cluster << mp->m_sb.sb_inopblog; + nbufs = length / blks_per_cluster; + /* - * Convert to buffer offsets and log it. + * Figure out what version number to use in the inodes we create. If + * the superblock version has caught up to the one that supports the new + * inode format, then use the new inode version. Otherwise use the old + * version so that old kernels will continue to be able to use the file + * system. + * + * For v3 inodes, we also need to write the inode number into the inode, + * so calculate the first inode number of the chunk here as + * XFS_OFFBNO_TO_AGINO() only works within a filesystem block, not + * across multiple filesystem blocks (such as a cluster) and so cannot + * be used in the cluster buffer loop below. + * + * Further, because we are writing the inode directly into the buffer + * and calculating a CRC on the entire inode, we have ot log the entire + * inode so that the entire range the CRC covers is present in the log. + * That means for v3 inode we log the entire buffer rather than just the + * inode cores. */ - ioffset = off << mp->m_sb.sb_inodelog; - first += ioffset; - last += ioffset; - xfs_trans_log_buf(tp, bp, first, last); -} + if (xfs_sb_version_hascrc(&mp->m_sb)) { + version = 3; + ino = XFS_AGINO_TO_INO(mp, agno, + XFS_OFFBNO_TO_AGINO(mp, agbno, 0)); -/* - * Allocation group level functions. - */ + /* + * log the initialisation that is about to take place as an + * logical operation. This means the transaction does not + * need to log the physical changes to the inode buffers as log + * recovery will know what initialisation is actually needed. + * Hence we only need to log the buffers as "ordered" buffers so + * they track in the AIL as if they were physically logged. + */ + if (tp) + xfs_icreate_log(tp, agno, agbno, mp->m_ialloc_inos, + mp->m_sb.sb_inodesize, length, gen); + } else + version = 2; + + for (j = 0; j < nbufs; j++) { + /* + * Get the block. + */ + d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster)); + fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d, + mp->m_bsize * blks_per_cluster, + XBF_UNMAPPED); + if (!fbuf) + return ENOMEM; + + /* Initialize the inode buffers and log them appropriately. */ + fbuf->b_ops = &xfs_inode_buf_ops; + xfs_buf_zero(fbuf, 0, BBTOB(fbuf->b_length)); + for (i = 0; i < inodes_per_cluster; i++) { + int ioffset = i << mp->m_sb.sb_inodelog; + uint isize = xfs_dinode_size(version); + + free = xfs_make_iptr(mp, fbuf, i); + free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC); + free->di_version = version; + free->di_gen = cpu_to_be32(gen); + free->di_next_unlinked = cpu_to_be32(NULLAGINO); + + if (version == 3) { + free->di_ino = cpu_to_be64(ino); + ino++; + uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid); + xfs_dinode_calc_crc(mp, free); + } else if (tp) { + /* just log the inode core */ + xfs_trans_log_buf(tp, fbuf, ioffset, + ioffset + isize - 1); + } + } + + if (tp) { + /* + * Mark the buffer as an inode allocation buffer so it + * sticks in AIL at the point of this allocation + * transaction. This ensures the they are on disk before + * the tail of the log can be moved past this + * transaction (i.e. by preventing relogging from moving + * it forward in the log). + */ + xfs_trans_inode_alloc_buf(tp, fbuf); + if (version == 3) { + /* + * Mark the buffer as ordered so that they are + * not physically logged in the transaction but + * still tracked in the AIL as part of the + * transaction and pin the log appropriately. + */ + xfs_trans_ordered_buf(tp, fbuf); + xfs_trans_log_buf(tp, fbuf, 0, + BBTOB(fbuf->b_length) - 1); + } + } else { + fbuf->b_flags |= XBF_DONE; + xfs_buf_delwri_queue(fbuf, buffer_list); + xfs_buf_relse(fbuf); + } + } + return 0; +} /* * Allocate new inodes in the allocation group specified by agbp. @@ -120,24 +361,15 @@ xfs_ialloc_ag_alloc( { xfs_agi_t *agi; /* allocation group header */ xfs_alloc_arg_t args; /* allocation argument structure */ - int blks_per_cluster; /* fs blocks per inode cluster */ - xfs_btree_cur_t *cur; /* inode btree cursor */ - xfs_daddr_t d; /* disk addr of buffer */ xfs_agnumber_t agno; int error; - xfs_buf_t *fbuf; /* new free inodes' buffer */ - xfs_dinode_t *free; /* new free inode structure */ - int i; /* inode counter */ - int j; /* block counter */ - int nbufs; /* num bufs of new inodes */ xfs_agino_t newino; /* new first inode's number */ xfs_agino_t newlen; /* new number of inodes */ - int ninodes; /* num inodes per buf */ - xfs_agino_t thisino; /* current inode number, for loop */ - int version; /* inode version number to use */ int isaligned = 0; /* inode allocation at stripe unit */ /* boundary */ + struct xfs_perag *pag; + memset(&args, 0, sizeof(args)); args.tp = tp; args.mp = tp->t_mountp; @@ -145,35 +377,59 @@ xfs_ialloc_ag_alloc( * Locking will ensure that we don't have two callers in here * at one time. */ - newlen = XFS_IALLOC_INODES(args.mp); + newlen = args.mp->m_ialloc_inos; if (args.mp->m_maxicount && args.mp->m_sb.sb_icount + newlen > args.mp->m_maxicount) return XFS_ERROR(ENOSPC); - args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp); + args.minlen = args.maxlen = args.mp->m_ialloc_blks; /* * First try to allocate inodes contiguous with the last-allocated * chunk of inodes. If the filesystem is striped, this will fill * an entire stripe unit with inodes. - */ + */ agi = XFS_BUF_TO_AGI(agbp); newino = be32_to_cpu(agi->agi_newino); + agno = be32_to_cpu(agi->agi_seqno); args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) + - XFS_IALLOC_BLOCKS(args.mp); + args.mp->m_ialloc_blks; if (likely(newino != NULLAGINO && (args.agbno < be32_to_cpu(agi->agi_length)))) { - args.fsbno = XFS_AGB_TO_FSB(args.mp, - be32_to_cpu(agi->agi_seqno), args.agbno); + args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); args.type = XFS_ALLOCTYPE_THIS_BNO; - args.mod = args.total = args.wasdel = args.isfl = - args.userdata = args.minalignslop = 0; args.prod = 1; - args.alignment = 1; + /* - * Allow space for the inode btree to split. + * We need to take into account alignment here to ensure that + * we don't modify the free list if we fail to have an exact + * block. If we don't have an exact match, and every oher + * attempt allocation attempt fails, we'll end up cancelling + * a dirty transaction and shutting down. + * + * For an exact allocation, alignment must be 1, + * however we need to take cluster alignment into account when + * fixing up the freelist. Use the minalignslop field to + * indicate that extra blocks might be required for alignment, + * but not to use them in the actual exact allocation. */ - args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1; + args.alignment = 1; + args.minalignslop = xfs_ialloc_cluster_alignment(&args) - 1; + + /* Allow space for the inode btree to split. */ + args.minleft = args.mp->m_in_maxlevels - 1; if ((error = xfs_alloc_vextent(&args))) return error; + + /* + * This request might have dirtied the transaction if the AG can + * satisfy the request, but the exact block was not available. + * If the allocation did fail, subsequent requests will relax + * the exact agbno requirement and increase the alignment + * instead. It is critical that the total size of the request + * (len + alignment + slop) does not increase from this point + * on, so reset minalignslop to ensure it is not included in + * subsequent requests. + */ + args.minalignslop = 0; } else args.fsbno = NULLFSBLOCK; @@ -191,32 +447,24 @@ xfs_ialloc_ag_alloc( ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN)); args.alignment = args.mp->m_dalign; isaligned = 1; - } else if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) && - args.mp->m_sb.sb_inoalignmt >= - XFS_B_TO_FSBT(args.mp, - XFS_INODE_CLUSTER_SIZE(args.mp))) - args.alignment = args.mp->m_sb.sb_inoalignmt; - else - args.alignment = 1; + } else + args.alignment = xfs_ialloc_cluster_alignment(&args); /* * Need to figure out where to allocate the inode blocks. * Ideally they should be spaced out through the a.g. * For now, just allocate blocks up front. */ args.agbno = be32_to_cpu(agi->agi_root); - args.fsbno = XFS_AGB_TO_FSB(args.mp, - be32_to_cpu(agi->agi_seqno), args.agbno); + args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); /* * Allocate a fixed-size extent of inodes. */ args.type = XFS_ALLOCTYPE_NEAR_BNO; - args.mod = args.total = args.wasdel = args.isfl = - args.userdata = args.minalignslop = 0; args.prod = 1; /* * Allow space for the inode btree to split. */ - args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1; + args.minleft = args.mp->m_in_maxlevels - 1; if ((error = xfs_alloc_vextent(&args))) return error; } @@ -228,14 +476,8 @@ xfs_ialloc_ag_alloc( if (isaligned && args.fsbno == NULLFSBLOCK) { args.type = XFS_ALLOCTYPE_NEAR_BNO; args.agbno = be32_to_cpu(agi->agi_root); - args.fsbno = XFS_AGB_TO_FSB(args.mp, - be32_to_cpu(agi->agi_seqno), args.agbno); - if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) && - args.mp->m_sb.sb_inoalignmt >= - XFS_B_TO_FSBT(args.mp, XFS_INODE_CLUSTER_SIZE(args.mp))) - args.alignment = args.mp->m_sb.sb_inoalignmt; - else - args.alignment = 1; + args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); + args.alignment = xfs_ialloc_cluster_alignment(&args); if ((error = xfs_alloc_vextent(&args))) return error; } @@ -245,90 +487,46 @@ xfs_ialloc_ag_alloc( return 0; } ASSERT(args.len == args.minlen); + /* - * Convert the results. - */ - newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0); - /* - * Loop over the new block(s), filling in the inodes. - * For small block sizes, manipulate the inodes in buffers - * which are multiples of the blocks size. + * Stamp and write the inode buffers. + * + * Seed the new inode cluster with a random generation number. This + * prevents short-term reuse of generation numbers if a chunk is + * freed and then immediately reallocated. We use random numbers + * rather than a linear progression to prevent the next generation + * number from being easily guessable. */ - if (args.mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(args.mp)) { - blks_per_cluster = 1; - nbufs = (int)args.len; - ninodes = args.mp->m_sb.sb_inopblock; - } else { - blks_per_cluster = XFS_INODE_CLUSTER_SIZE(args.mp) / - args.mp->m_sb.sb_blocksize; - nbufs = (int)args.len / blks_per_cluster; - ninodes = blks_per_cluster * args.mp->m_sb.sb_inopblock; - } + error = xfs_ialloc_inode_init(args.mp, tp, NULL, agno, args.agbno, + args.len, prandom_u32()); + + if (error) + return error; /* - * Figure out what version number to use in the inodes we create. - * If the superblock version has caught up to the one that supports - * the new inode format, then use the new inode version. Otherwise - * use the old version so that old kernels will continue to be - * able to use the file system. + * Convert the results. */ - if (XFS_SB_VERSION_HASNLINK(&args.mp->m_sb)) - version = XFS_DINODE_VERSION_2; - else - version = XFS_DINODE_VERSION_1; - - for (j = 0; j < nbufs; j++) { - /* - * Get the block. - */ - d = XFS_AGB_TO_DADDR(args.mp, be32_to_cpu(agi->agi_seqno), - args.agbno + (j * blks_per_cluster)); - fbuf = xfs_trans_get_buf(tp, args.mp->m_ddev_targp, d, - args.mp->m_bsize * blks_per_cluster, - XFS_BUF_LOCK); - ASSERT(fbuf); - ASSERT(!XFS_BUF_GETERROR(fbuf)); - /* - * Set initial values for the inodes in this buffer. - */ - xfs_biozero(fbuf, 0, ninodes << args.mp->m_sb.sb_inodelog); - for (i = 0; i < ninodes; i++) { - free = XFS_MAKE_IPTR(args.mp, fbuf, i); - free->di_core.di_magic = cpu_to_be16(XFS_DINODE_MAGIC); - free->di_core.di_version = version; - free->di_next_unlinked = cpu_to_be32(NULLAGINO); - xfs_ialloc_log_di(tp, fbuf, i, - XFS_DI_CORE_BITS | XFS_DI_NEXT_UNLINKED); - } - xfs_trans_inode_alloc_buf(tp, fbuf); - } + newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0); be32_add_cpu(&agi->agi_count, newlen); be32_add_cpu(&agi->agi_freecount, newlen); - agno = be32_to_cpu(agi->agi_seqno); - down_read(&args.mp->m_peraglock); - args.mp->m_perag[agno].pagi_freecount += newlen; - up_read(&args.mp->m_peraglock); + pag = xfs_perag_get(args.mp, agno); + pag->pagi_freecount += newlen; + xfs_perag_put(pag); agi->agi_newino = cpu_to_be32(newino); + /* - * Insert records describing the new inode chunk into the btree. + * Insert records describing the new inode chunk into the btrees. */ - cur = xfs_btree_init_cursor(args.mp, tp, agbp, agno, - XFS_BTNUM_INO, (xfs_inode_t *)0, 0); - for (thisino = newino; - thisino < newino + newlen; - thisino += XFS_INODES_PER_CHUNK) { - if ((error = xfs_inobt_lookup_eq(cur, thisino, - XFS_INODES_PER_CHUNK, XFS_INOBT_ALL_FREE, &i))) { - xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); - return error; - } - ASSERT(i == 0); - if ((error = xfs_inobt_insert(cur, &i))) { - xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + error = xfs_inobt_insert(args.mp, tp, agbp, newino, newlen, + XFS_BTNUM_INO); + if (error) + return error; + + if (xfs_sb_version_hasfinobt(&args.mp->m_sb)) { + error = xfs_inobt_insert(args.mp, tp, agbp, newino, newlen, + XFS_BTNUM_FINO); + if (error) return error; - } - ASSERT(i == 1); } - xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); /* * Log allocation group header fields */ @@ -343,7 +541,7 @@ xfs_ialloc_ag_alloc( return 0; } -STATIC_INLINE xfs_agnumber_t +STATIC xfs_agnumber_t xfs_ialloc_next_ag( xfs_mount_t *mp) { @@ -351,7 +549,7 @@ xfs_ialloc_next_ag( spin_lock(&mp->m_agirotor_lock); agno = mp->m_agirotor; - if (++mp->m_agirotor == mp->m_maxagi) + if (++mp->m_agirotor >= mp->m_maxagi) mp->m_agirotor = 0; spin_unlock(&mp->m_agirotor_lock); @@ -360,16 +558,15 @@ xfs_ialloc_next_ag( /* * Select an allocation group to look for a free inode in, based on the parent - * inode and then mode. Return the allocation group buffer. + * inode and the mode. Return the allocation group buffer. */ -STATIC xfs_buf_t * /* allocation group buffer */ +STATIC xfs_agnumber_t xfs_ialloc_ag_select( xfs_trans_t *tp, /* transaction pointer */ xfs_ino_t parent, /* parent directory inode number */ - mode_t mode, /* bits set to indicate file type */ + umode_t mode, /* bits set to indicate file type */ int okalloc) /* ok to allocate more space */ { - xfs_buf_t *agbp; /* allocation group header buffer */ xfs_agnumber_t agcount; /* number of ag's in the filesystem */ xfs_agnumber_t agno; /* current ag number */ int flags; /* alloc buffer locking flags */ @@ -379,6 +576,7 @@ xfs_ialloc_ag_select( int needspace; /* file mode implies space allocated */ xfs_perag_t *pag; /* per allocation group data */ xfs_agnumber_t pagno; /* parent (starting) ag number */ + int error; /* * Files of these types need at least one block if length > 0 @@ -394,7 +592,9 @@ xfs_ialloc_ag_select( if (pagno >= agcount) pagno = 0; } + ASSERT(pagno < agcount); + /* * Loop through allocation groups, looking for one with a little * free space in it. Note we don't look for free inodes, exactly. @@ -404,637 +604,876 @@ xfs_ialloc_ag_select( */ agno = pagno; flags = XFS_ALLOC_FLAG_TRYLOCK; - down_read(&mp->m_peraglock); for (;;) { - pag = &mp->m_perag[agno]; + pag = xfs_perag_get(mp, agno); + if (!pag->pagi_inodeok) { + xfs_ialloc_next_ag(mp); + goto nextag; + } + if (!pag->pagi_init) { - if (xfs_ialloc_read_agi(mp, tp, agno, &agbp)) { - agbp = NULL; + error = xfs_ialloc_pagi_init(mp, tp, agno); + if (error) goto nextag; - } - } else - agbp = NULL; + } - if (!pag->pagi_inodeok) { - xfs_ialloc_next_ag(mp); - goto unlock_nextag; + if (pag->pagi_freecount) { + xfs_perag_put(pag); + return agno; } - /* - * Is there enough free space for the file plus a block - * of inodes (if we need to allocate some)? - */ - ineed = pag->pagi_freecount ? 0 : XFS_IALLOC_BLOCKS(mp); - if (ineed && !pag->pagf_init) { - if (agbp == NULL && - xfs_ialloc_read_agi(mp, tp, agno, &agbp)) { - agbp = NULL; + if (!okalloc) + goto nextag; + + if (!pag->pagf_init) { + error = xfs_alloc_pagf_init(mp, tp, agno, flags); + if (error) goto nextag; - } - (void)xfs_alloc_pagf_init(mp, tp, agno, flags); } - if (!ineed || pag->pagf_init) { - if (ineed && !(longest = pag->pagf_longest)) - longest = pag->pagf_flcount > 0; - if (!ineed || - (pag->pagf_freeblks >= needspace + ineed && - longest >= ineed && - okalloc)) { - if (agbp == NULL && - xfs_ialloc_read_agi(mp, tp, agno, &agbp)) { - agbp = NULL; - goto nextag; - } - up_read(&mp->m_peraglock); - return agbp; - } + + /* + * Is there enough free space for the file plus a block of + * inodes? (if we need to allocate some)? + */ + ineed = mp->m_ialloc_blks; + longest = pag->pagf_longest; + if (!longest) + longest = pag->pagf_flcount > 0; + + if (pag->pagf_freeblks >= needspace + ineed && + longest >= ineed) { + xfs_perag_put(pag); + return agno; } -unlock_nextag: - if (agbp) - xfs_trans_brelse(tp, agbp); nextag: + xfs_perag_put(pag); /* * No point in iterating over the rest, if we're shutting * down. */ - if (XFS_FORCED_SHUTDOWN(mp)) { - up_read(&mp->m_peraglock); - return NULL; - } + if (XFS_FORCED_SHUTDOWN(mp)) + return NULLAGNUMBER; agno++; if (agno >= agcount) agno = 0; if (agno == pagno) { - if (flags == 0) { - up_read(&mp->m_peraglock); - return NULL; - } + if (flags == 0) + return NULLAGNUMBER; flags = 0; } } } /* - * Visible inode allocation functions. - */ - -/* - * Allocate an inode on disk. - * Mode is used to tell whether the new inode will need space, and whether - * it is a directory. - * - * The arguments IO_agbp and alloc_done are defined to work within - * the constraint of one allocation per transaction. - * xfs_dialloc() is designed to be called twice if it has to do an - * allocation to make more free inodes. On the first call, - * IO_agbp should be set to NULL. If an inode is available, - * i.e., xfs_dialloc() did not need to do an allocation, an inode - * number is returned. In this case, IO_agbp would be set to the - * current ag_buf and alloc_done set to false. - * If an allocation needed to be done, xfs_dialloc would return - * the current ag_buf in IO_agbp and set alloc_done to true. - * The caller should then commit the current transaction, allocate a new - * transaction, and call xfs_dialloc() again, passing in the previous - * value of IO_agbp. IO_agbp should be held across the transactions. - * Since the agbp is locked across the two calls, the second call is - * guaranteed to have a free inode available. - * - * Once we successfully pick an inode its number is returned and the - * on-disk data structures are updated. The inode itself is not read - * in, since doing so would break ordering constraints with xfs_reclaim. + * Try to retrieve the next record to the left/right from the current one. */ -int -xfs_dialloc( - xfs_trans_t *tp, /* transaction pointer */ - xfs_ino_t parent, /* parent inode (directory) */ - mode_t mode, /* mode bits for new inode */ - int okalloc, /* ok to allocate more space */ - xfs_buf_t **IO_agbp, /* in/out ag header's buffer */ - boolean_t *alloc_done, /* true if we needed to replenish - inode freelist */ - xfs_ino_t *inop) /* inode number allocated */ +STATIC int +xfs_ialloc_next_rec( + struct xfs_btree_cur *cur, + xfs_inobt_rec_incore_t *rec, + int *done, + int left) { - xfs_agnumber_t agcount; /* number of allocation groups */ - xfs_buf_t *agbp; /* allocation group header's buffer */ - xfs_agnumber_t agno; /* allocation group number */ - xfs_agi_t *agi; /* allocation group header structure */ - xfs_btree_cur_t *cur; /* inode allocation btree cursor */ - int error; /* error return value */ - int i; /* result code */ - int ialloced; /* inode allocation status */ - int noroom = 0; /* no space for inode blk allocation */ - xfs_ino_t ino; /* fs-relative inode to be returned */ - /* REFERENCED */ - int j; /* result code */ - xfs_mount_t *mp; /* file system mount structure */ - int offset; /* index of inode in chunk */ - xfs_agino_t pagino; /* parent's a.g. relative inode # */ - xfs_agnumber_t pagno; /* parent's allocation group number */ - xfs_inobt_rec_incore_t rec; /* inode allocation record */ - xfs_agnumber_t tagno; /* testing allocation group number */ - xfs_btree_cur_t *tcur; /* temp cursor */ - xfs_inobt_rec_incore_t trec; /* temp inode allocation record */ - - - if (*IO_agbp == NULL) { - /* - * We do not have an agbp, so select an initial allocation - * group for inode allocation. - */ - agbp = xfs_ialloc_ag_select(tp, parent, mode, okalloc); - /* - * Couldn't find an allocation group satisfying the - * criteria, give up. - */ - if (!agbp) { - *inop = NULLFSINO; - return 0; - } - agi = XFS_BUF_TO_AGI(agbp); - ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC); - } else { - /* - * Continue where we left off before. In this case, we - * know that the allocation group has free inodes. - */ - agbp = *IO_agbp; - agi = XFS_BUF_TO_AGI(agbp); - ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC); - ASSERT(be32_to_cpu(agi->agi_freecount) > 0); - } - mp = tp->t_mountp; - agcount = mp->m_sb.sb_agcount; - agno = be32_to_cpu(agi->agi_seqno); - tagno = agno; - pagno = XFS_INO_TO_AGNO(mp, parent); - pagino = XFS_INO_TO_AGINO(mp, parent); + int error; + int i; - /* - * If we have already hit the ceiling of inode blocks then clear - * okalloc so we scan all available agi structures for a free - * inode. - */ + if (left) + error = xfs_btree_decrement(cur, 0, &i); + else + error = xfs_btree_increment(cur, 0, &i); - if (mp->m_maxicount && - mp->m_sb.sb_icount + XFS_IALLOC_INODES(mp) > mp->m_maxicount) { - noroom = 1; - okalloc = 0; + if (error) + return error; + *done = !i; + if (i) { + error = xfs_inobt_get_rec(cur, rec, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); } - /* - * Loop until we find an allocation group that either has free inodes - * or in which we can allocate some inodes. Iterate through the - * allocation groups upward, wrapping at the end. - */ - *alloc_done = B_FALSE; - while (!agi->agi_freecount) { - /* - * Don't do anything if we're not supposed to allocate - * any blocks, just go on to the next ag. - */ - if (okalloc) { - /* - * Try to allocate some new inodes in the allocation - * group. - */ - if ((error = xfs_ialloc_ag_alloc(tp, agbp, &ialloced))) { - xfs_trans_brelse(tp, agbp); - if (error == ENOSPC) { - *inop = NULLFSINO; - return 0; - } else - return error; - } - if (ialloced) { - /* - * We successfully allocated some inodes, return - * the current context to the caller so that it - * can commit the current transaction and call - * us again where we left off. - */ - ASSERT(be32_to_cpu(agi->agi_freecount) > 0); - *alloc_done = B_TRUE; - *IO_agbp = agbp; - *inop = NULLFSINO; - return 0; - } - } - /* - * If it failed, give up on this ag. - */ - xfs_trans_brelse(tp, agbp); - /* - * Go on to the next ag: get its ag header. - */ -nextag: - if (++tagno == agcount) - tagno = 0; - if (tagno == agno) { - *inop = NULLFSINO; - return noroom ? ENOSPC : 0; - } - down_read(&mp->m_peraglock); - if (mp->m_perag[tagno].pagi_inodeok == 0) { - up_read(&mp->m_peraglock); - goto nextag; - } - error = xfs_ialloc_read_agi(mp, tp, tagno, &agbp); - up_read(&mp->m_peraglock); + return 0; +} + +STATIC int +xfs_ialloc_get_rec( + struct xfs_btree_cur *cur, + xfs_agino_t agino, + xfs_inobt_rec_incore_t *rec, + int *done) +{ + int error; + int i; + + error = xfs_inobt_lookup(cur, agino, XFS_LOOKUP_EQ, &i); + if (error) + return error; + *done = !i; + if (i) { + error = xfs_inobt_get_rec(cur, rec, &i); if (error) - goto nextag; - agi = XFS_BUF_TO_AGI(agbp); - ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC); + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); } - /* - * Here with an allocation group that has a free inode. - * Reset agno since we may have chosen a new ag in the - * loop above. - */ - agno = tagno; - *IO_agbp = NULL; - cur = xfs_btree_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno), - XFS_BTNUM_INO, (xfs_inode_t *)0, 0); + + return 0; +} + +/* + * Allocate an inode using the inobt-only algorithm. + */ +STATIC int +xfs_dialloc_ag_inobt( + struct xfs_trans *tp, + struct xfs_buf *agbp, + xfs_ino_t parent, + xfs_ino_t *inop) +{ + struct xfs_mount *mp = tp->t_mountp; + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + xfs_agnumber_t pagno = XFS_INO_TO_AGNO(mp, parent); + xfs_agino_t pagino = XFS_INO_TO_AGINO(mp, parent); + struct xfs_perag *pag; + struct xfs_btree_cur *cur, *tcur; + struct xfs_inobt_rec_incore rec, trec; + xfs_ino_t ino; + int error; + int offset; + int i, j; + + pag = xfs_perag_get(mp, agno); + + ASSERT(pag->pagi_init); + ASSERT(pag->pagi_inodeok); + ASSERT(pag->pagi_freecount > 0); + + restart_pagno: + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO); /* * If pagino is 0 (this is the root inode allocation) use newino. * This must work because we've just allocated some. */ if (!pagino) pagino = be32_to_cpu(agi->agi_newino); -#ifdef DEBUG - if (cur->bc_nlevels == 1) { - int freecount = 0; - if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - do { - if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) - goto error0; - } while (i == 1); + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error0; - ASSERT(freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); - } -#endif /* - * If in the same a.g. as the parent, try to get near the parent. + * If in the same AG as the parent, try to get near the parent. */ if (pagno == agno) { - if ((error = xfs_inobt_lookup_le(cur, pagino, 0, 0, &i))) + int doneleft; /* done, to the left */ + int doneright; /* done, to the right */ + int searchdistance = 10; + + error = xfs_inobt_lookup(cur, pagino, XFS_LOOKUP_LE, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + error = xfs_inobt_get_rec(cur, &rec, &j); + if (error) goto error0; - if (i != 0 && - (error = xfs_inobt_get_rec(cur, &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, &j)) == 0 && - j == 1 && - rec.ir_freecount > 0) { + XFS_WANT_CORRUPTED_GOTO(j == 1, error0); + + if (rec.ir_freecount > 0) { /* * Found a free inode in the same chunk - * as parent, done. + * as the parent, done. */ + goto alloc_inode; } + + + /* + * In the same AG as parent, but parent's chunk is full. + */ + + /* duplicate the cursor, search left & right simultaneously */ + error = xfs_btree_dup_cursor(cur, &tcur); + if (error) + goto error0; + /* - * In the same a.g. as parent, but parent's chunk is full. + * Skip to last blocks looked up if same parent inode. */ - else { - int doneleft; /* done, to the left */ - int doneright; /* done, to the right */ + if (pagino != NULLAGINO && + pag->pagl_pagino == pagino && + pag->pagl_leftrec != NULLAGINO && + pag->pagl_rightrec != NULLAGINO) { + error = xfs_ialloc_get_rec(tcur, pag->pagl_leftrec, + &trec, &doneleft); + if (error) + goto error1; + error = xfs_ialloc_get_rec(cur, pag->pagl_rightrec, + &rec, &doneright); if (error) - goto error0; - ASSERT(i == 1); - ASSERT(j == 1); - /* - * Duplicate the cursor, search left & right - * simultaneously. - */ - if ((error = xfs_btree_dup_cursor(cur, &tcur))) - goto error0; - /* - * Search left with tcur, back up 1 record. - */ - if ((error = xfs_inobt_decrement(tcur, 0, &i))) goto error1; - doneleft = !i; - if (!doneleft) { - if ((error = xfs_inobt_get_rec(tcur, - &trec.ir_startino, - &trec.ir_freecount, - &trec.ir_free, &i))) - goto error1; - XFS_WANT_CORRUPTED_GOTO(i == 1, error1); - } - /* - * Search right with cur, go forward 1 record. - */ - if ((error = xfs_inobt_increment(cur, 0, &i))) + } else { + /* search left with tcur, back up 1 record */ + error = xfs_ialloc_next_rec(tcur, &trec, &doneleft, 1); + if (error) goto error1; - doneright = !i; - if (!doneright) { - if ((error = xfs_inobt_get_rec(cur, - &rec.ir_startino, - &rec.ir_freecount, - &rec.ir_free, &i))) - goto error1; - XFS_WANT_CORRUPTED_GOTO(i == 1, error1); - } - /* - * Loop until we find the closest inode chunk - * with a free one. - */ - while (!doneleft || !doneright) { - int useleft; /* using left inode - chunk this time */ + /* search right with cur, go forward 1 record. */ + error = xfs_ialloc_next_rec(cur, &rec, &doneright, 0); + if (error) + goto error1; + } + + /* + * Loop until we find an inode chunk with a free inode. + */ + while (!doneleft || !doneright) { + int useleft; /* using left inode chunk this time */ + + if (!--searchdistance) { /* - * Figure out which block is closer, - * if both are valid. - */ - if (!doneleft && !doneright) - useleft = - pagino - - (trec.ir_startino + - XFS_INODES_PER_CHUNK - 1) < - rec.ir_startino - pagino; - else - useleft = !doneleft; - /* - * If checking the left, does it have - * free inodes? - */ - if (useleft && trec.ir_freecount) { - /* - * Yes, set it up as the chunk to use. - */ - rec = trec; - xfs_btree_del_cursor(cur, - XFS_BTREE_NOERROR); - cur = tcur; - break; - } - /* - * If checking the right, does it have - * free inodes? - */ - if (!useleft && rec.ir_freecount) { - /* - * Yes, it's already set up. - */ - xfs_btree_del_cursor(tcur, - XFS_BTREE_NOERROR); - break; - } - /* - * If used the left, get another one - * further left. - */ - if (useleft) { - if ((error = xfs_inobt_decrement(tcur, 0, - &i))) - goto error1; - doneleft = !i; - if (!doneleft) { - if ((error = xfs_inobt_get_rec( - tcur, - &trec.ir_startino, - &trec.ir_freecount, - &trec.ir_free, &i))) - goto error1; - XFS_WANT_CORRUPTED_GOTO(i == 1, - error1); - } - } - /* - * If used the right, get another one - * further right. + * Not in range - save last search + * location and allocate a new inode */ - else { - if ((error = xfs_inobt_increment(cur, 0, - &i))) - goto error1; - doneright = !i; - if (!doneright) { - if ((error = xfs_inobt_get_rec( - cur, - &rec.ir_startino, - &rec.ir_freecount, - &rec.ir_free, &i))) - goto error1; - XFS_WANT_CORRUPTED_GOTO(i == 1, - error1); - } - } + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + pag->pagl_leftrec = trec.ir_startino; + pag->pagl_rightrec = rec.ir_startino; + pag->pagl_pagino = pagino; + goto newino; + } + + /* figure out the closer block if both are valid. */ + if (!doneleft && !doneright) { + useleft = pagino - + (trec.ir_startino + XFS_INODES_PER_CHUNK - 1) < + rec.ir_startino - pagino; + } else { + useleft = !doneleft; + } + + /* free inodes to the left? */ + if (useleft && trec.ir_freecount) { + rec = trec; + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + cur = tcur; + + pag->pagl_leftrec = trec.ir_startino; + pag->pagl_rightrec = rec.ir_startino; + pag->pagl_pagino = pagino; + goto alloc_inode; + } + + /* free inodes to the right? */ + if (!useleft && rec.ir_freecount) { + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + + pag->pagl_leftrec = trec.ir_startino; + pag->pagl_rightrec = rec.ir_startino; + pag->pagl_pagino = pagino; + goto alloc_inode; + } + + /* get next record to check */ + if (useleft) { + error = xfs_ialloc_next_rec(tcur, &trec, + &doneleft, 1); + } else { + error = xfs_ialloc_next_rec(cur, &rec, + &doneright, 0); } - ASSERT(!doneleft || !doneright); + if (error) + goto error1; } + + /* + * We've reached the end of the btree. because + * we are only searching a small chunk of the + * btree each search, there is obviously free + * inodes closer to the parent inode than we + * are now. restart the search again. + */ + pag->pagl_pagino = NULLAGINO; + pag->pagl_leftrec = NULLAGINO; + pag->pagl_rightrec = NULLAGINO; + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + goto restart_pagno; } + /* - * In a different a.g. from the parent. + * In a different AG from the parent. * See if the most recently allocated block has any free. */ - else if (be32_to_cpu(agi->agi_newino) != NULLAGINO) { - if ((error = xfs_inobt_lookup_eq(cur, - be32_to_cpu(agi->agi_newino), 0, 0, &i))) +newino: + if (agi->agi_newino != cpu_to_be32(NULLAGINO)) { + error = xfs_inobt_lookup(cur, be32_to_cpu(agi->agi_newino), + XFS_LOOKUP_EQ, &i); + if (error) goto error0; - if (i == 1 && - (error = xfs_inobt_get_rec(cur, &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, &j)) == 0 && - j == 1 && - rec.ir_freecount > 0) { - /* - * The last chunk allocated in the group still has - * a free inode. - */ - } - /* - * None left in the last group, search the whole a.g. - */ - else { + + if (i == 1) { + error = xfs_inobt_get_rec(cur, &rec, &j); if (error) goto error0; - if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i))) - goto error0; - ASSERT(i == 1); - for (;;) { - if ((error = xfs_inobt_get_rec(cur, - &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, - &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if (rec.ir_freecount > 0) - break; - if ((error = xfs_inobt_increment(cur, 0, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + if (j == 1 && rec.ir_freecount > 0) { + /* + * The last chunk allocated in the group + * still has a free inode. + */ + goto alloc_inode; } } } - offset = XFS_IALLOC_FIND_FREE(&rec.ir_free); + + /* + * None left in the last group, search the whole AG + */ + error = xfs_inobt_lookup(cur, 0, XFS_LOOKUP_GE, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + for (;;) { + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + if (rec.ir_freecount > 0) + break; + error = xfs_btree_increment(cur, 0, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + } + +alloc_inode: + offset = xfs_lowbit64(rec.ir_free); ASSERT(offset >= 0); ASSERT(offset < XFS_INODES_PER_CHUNK); ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) % XFS_INODES_PER_CHUNK) == 0); ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino + offset); - XFS_INOBT_CLR_FREE(&rec, offset); + rec.ir_free &= ~XFS_INOBT_MASK(offset); rec.ir_freecount--; - if ((error = xfs_inobt_update(cur, rec.ir_startino, rec.ir_freecount, - rec.ir_free))) + error = xfs_inobt_update(cur, &rec); + if (error) goto error0; be32_add_cpu(&agi->agi_freecount, -1); xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT); - down_read(&mp->m_peraglock); - mp->m_perag[tagno].pagi_freecount--; - up_read(&mp->m_peraglock); -#ifdef DEBUG - if (cur->bc_nlevels == 1) { - int freecount = 0; + pag->pagi_freecount--; + + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error0; - if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i))) - goto error0; - do { - if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) - goto error0; - } while (i == 1); - ASSERT(freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); - } -#endif xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -1); + xfs_perag_put(pag); *inop = ino; return 0; error1: xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); error0: xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + xfs_perag_put(pag); return error; } /* - * Free disk inode. Carefully avoids touching the incore inode, all - * manipulations incore are the caller's responsibility. - * The on-disk inode is not changed by this operation, only the - * btree (free inode mask) is changed. + * Use the free inode btree to allocate an inode based on distance from the + * parent. Note that the provided cursor may be deleted and replaced. */ -int -xfs_difree( - xfs_trans_t *tp, /* transaction pointer */ - xfs_ino_t inode, /* inode to be freed */ - xfs_bmap_free_t *flist, /* extents to free */ - int *delete, /* set if inode cluster was deleted */ - xfs_ino_t *first_ino) /* first inode in deleted cluster */ +STATIC int +xfs_dialloc_ag_finobt_near( + xfs_agino_t pagino, + struct xfs_btree_cur **ocur, + struct xfs_inobt_rec_incore *rec) { - /* REFERENCED */ - xfs_agblock_t agbno; /* block number containing inode */ - xfs_buf_t *agbp; /* buffer containing allocation group header */ - xfs_agino_t agino; /* inode number relative to allocation group */ - xfs_agnumber_t agno; /* allocation group number */ - xfs_agi_t *agi; /* allocation group header */ - xfs_btree_cur_t *cur; /* inode btree cursor */ - int error; /* error return value */ - int i; /* result code */ - int ilen; /* inodes in an inode cluster */ - xfs_mount_t *mp; /* mount structure for filesystem */ - int off; /* offset of inode in inode chunk */ - xfs_inobt_rec_incore_t rec; /* btree record */ + struct xfs_btree_cur *lcur = *ocur; /* left search cursor */ + struct xfs_btree_cur *rcur; /* right search cursor */ + struct xfs_inobt_rec_incore rrec; + int error; + int i, j; - mp = tp->t_mountp; + error = xfs_inobt_lookup(lcur, pagino, XFS_LOOKUP_LE, &i); + if (error) + return error; + + if (i == 1) { + error = xfs_inobt_get_rec(lcur, rec, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + + /* + * See if we've landed in the parent inode record. The finobt + * only tracks chunks with at least one free inode, so record + * existence is enough. + */ + if (pagino >= rec->ir_startino && + pagino < (rec->ir_startino + XFS_INODES_PER_CHUNK)) + return 0; + } + + error = xfs_btree_dup_cursor(lcur, &rcur); + if (error) + return error; + + error = xfs_inobt_lookup(rcur, pagino, XFS_LOOKUP_GE, &j); + if (error) + goto error_rcur; + if (j == 1) { + error = xfs_inobt_get_rec(rcur, &rrec, &j); + if (error) + goto error_rcur; + XFS_WANT_CORRUPTED_GOTO(j == 1, error_rcur); + } + + XFS_WANT_CORRUPTED_GOTO(i == 1 || j == 1, error_rcur); + if (i == 1 && j == 1) { + /* + * Both the left and right records are valid. Choose the closer + * inode chunk to the target. + */ + if ((pagino - rec->ir_startino + XFS_INODES_PER_CHUNK - 1) > + (rrec.ir_startino - pagino)) { + *rec = rrec; + xfs_btree_del_cursor(lcur, XFS_BTREE_NOERROR); + *ocur = rcur; + } else { + xfs_btree_del_cursor(rcur, XFS_BTREE_NOERROR); + } + } else if (j == 1) { + /* only the right record is valid */ + *rec = rrec; + xfs_btree_del_cursor(lcur, XFS_BTREE_NOERROR); + *ocur = rcur; + } else if (i == 1) { + /* only the left record is valid */ + xfs_btree_del_cursor(rcur, XFS_BTREE_NOERROR); + } + + return 0; + +error_rcur: + xfs_btree_del_cursor(rcur, XFS_BTREE_ERROR); + return error; +} + +/* + * Use the free inode btree to find a free inode based on a newino hint. If + * the hint is NULL, find the first free inode in the AG. + */ +STATIC int +xfs_dialloc_ag_finobt_newino( + struct xfs_agi *agi, + struct xfs_btree_cur *cur, + struct xfs_inobt_rec_incore *rec) +{ + int error; + int i; + + if (agi->agi_newino != cpu_to_be32(NULLAGINO)) { + error = xfs_inobt_lookup(cur, agi->agi_newino, XFS_LOOKUP_EQ, + &i); + if (error) + return error; + if (i == 1) { + error = xfs_inobt_get_rec(cur, rec, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + + return 0; + } + } /* - * Break up inode number into its components. + * Find the first inode available in the AG. */ - agno = XFS_INO_TO_AGNO(mp, inode); - if (agno >= mp->m_sb.sb_agcount) { - cmn_err(CE_WARN, - "xfs_difree: agno >= mp->m_sb.sb_agcount (%d >= %d) on %s. Returning EINVAL.", - agno, mp->m_sb.sb_agcount, mp->m_fsname); - ASSERT(0); - return XFS_ERROR(EINVAL); + error = xfs_inobt_lookup(cur, 0, XFS_LOOKUP_GE, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + + error = xfs_inobt_get_rec(cur, rec, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + + return 0; +} + +/* + * Update the inobt based on a modification made to the finobt. Also ensure that + * the records from both trees are equivalent post-modification. + */ +STATIC int +xfs_dialloc_ag_update_inobt( + struct xfs_btree_cur *cur, /* inobt cursor */ + struct xfs_inobt_rec_incore *frec, /* finobt record */ + int offset) /* inode offset */ +{ + struct xfs_inobt_rec_incore rec; + int error; + int i; + + error = xfs_inobt_lookup(cur, frec->ir_startino, XFS_LOOKUP_EQ, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) + return error; + XFS_WANT_CORRUPTED_RETURN(i == 1); + ASSERT((XFS_AGINO_TO_OFFSET(cur->bc_mp, rec.ir_startino) % + XFS_INODES_PER_CHUNK) == 0); + + rec.ir_free &= ~XFS_INOBT_MASK(offset); + rec.ir_freecount--; + + XFS_WANT_CORRUPTED_RETURN((rec.ir_free == frec->ir_free) && + (rec.ir_freecount == frec->ir_freecount)); + + error = xfs_inobt_update(cur, &rec); + if (error) + return error; + + return 0; +} + +/* + * Allocate an inode using the free inode btree, if available. Otherwise, fall + * back to the inobt search algorithm. + * + * The caller selected an AG for us, and made sure that free inodes are + * available. + */ +STATIC int +xfs_dialloc_ag( + struct xfs_trans *tp, + struct xfs_buf *agbp, + xfs_ino_t parent, + xfs_ino_t *inop) +{ + struct xfs_mount *mp = tp->t_mountp; + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + xfs_agnumber_t pagno = XFS_INO_TO_AGNO(mp, parent); + xfs_agino_t pagino = XFS_INO_TO_AGINO(mp, parent); + struct xfs_perag *pag; + struct xfs_btree_cur *cur; /* finobt cursor */ + struct xfs_btree_cur *icur; /* inobt cursor */ + struct xfs_inobt_rec_incore rec; + xfs_ino_t ino; + int error; + int offset; + int i; + + if (!xfs_sb_version_hasfinobt(&mp->m_sb)) + return xfs_dialloc_ag_inobt(tp, agbp, parent, inop); + + pag = xfs_perag_get(mp, agno); + + /* + * If pagino is 0 (this is the root inode allocation) use newino. + * This must work because we've just allocated some. + */ + if (!pagino) + pagino = be32_to_cpu(agi->agi_newino); + + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_FINO); + + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error_cur; + + /* + * The search algorithm depends on whether we're in the same AG as the + * parent. If so, find the closest available inode to the parent. If + * not, consider the agi hint or find the first free inode in the AG. + */ + if (agno == pagno) + error = xfs_dialloc_ag_finobt_near(pagino, &cur, &rec); + else + error = xfs_dialloc_ag_finobt_newino(agi, cur, &rec); + if (error) + goto error_cur; + + offset = xfs_lowbit64(rec.ir_free); + ASSERT(offset >= 0); + ASSERT(offset < XFS_INODES_PER_CHUNK); + ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) % + XFS_INODES_PER_CHUNK) == 0); + ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino + offset); + + /* + * Modify or remove the finobt record. + */ + rec.ir_free &= ~XFS_INOBT_MASK(offset); + rec.ir_freecount--; + if (rec.ir_freecount) + error = xfs_inobt_update(cur, &rec); + else + error = xfs_btree_delete(cur, &i); + if (error) + goto error_cur; + + /* + * The finobt has now been updated appropriately. We haven't updated the + * agi and superblock yet, so we can create an inobt cursor and validate + * the original freecount. If all is well, make the equivalent update to + * the inobt using the finobt record and offset information. + */ + icur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO); + + error = xfs_check_agi_freecount(icur, agi); + if (error) + goto error_icur; + + error = xfs_dialloc_ag_update_inobt(icur, &rec, offset); + if (error) + goto error_icur; + + /* + * Both trees have now been updated. We must update the perag and + * superblock before we can check the freecount for each btree. + */ + be32_add_cpu(&agi->agi_freecount, -1); + xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT); + pag->pagi_freecount--; + + xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -1); + + error = xfs_check_agi_freecount(icur, agi); + if (error) + goto error_icur; + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error_icur; + + xfs_btree_del_cursor(icur, XFS_BTREE_NOERROR); + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + xfs_perag_put(pag); + *inop = ino; + return 0; + +error_icur: + xfs_btree_del_cursor(icur, XFS_BTREE_ERROR); +error_cur: + xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + xfs_perag_put(pag); + return error; +} + +/* + * Allocate an inode on disk. + * + * Mode is used to tell whether the new inode will need space, and whether it + * is a directory. + * + * This function is designed to be called twice if it has to do an allocation + * to make more free inodes. On the first call, *IO_agbp should be set to NULL. + * If an inode is available without having to performn an allocation, an inode + * number is returned. In this case, *IO_agbp is set to NULL. If an allocation + * needs to be done, xfs_dialloc returns the current AGI buffer in *IO_agbp. + * The caller should then commit the current transaction, allocate a + * new transaction, and call xfs_dialloc() again, passing in the previous value + * of *IO_agbp. IO_agbp should be held across the transactions. Since the AGI + * buffer is locked across the two calls, the second call is guaranteed to have + * a free inode available. + * + * Once we successfully pick an inode its number is returned and the on-disk + * data structures are updated. The inode itself is not read in, since doing so + * would break ordering constraints with xfs_reclaim. + */ +int +xfs_dialloc( + struct xfs_trans *tp, + xfs_ino_t parent, + umode_t mode, + int okalloc, + struct xfs_buf **IO_agbp, + xfs_ino_t *inop) +{ + struct xfs_mount *mp = tp->t_mountp; + struct xfs_buf *agbp; + xfs_agnumber_t agno; + int error; + int ialloced; + int noroom = 0; + xfs_agnumber_t start_agno; + struct xfs_perag *pag; + + if (*IO_agbp) { + /* + * If the caller passes in a pointer to the AGI buffer, + * continue where we left off before. In this case, we + * know that the allocation group has free inodes. + */ + agbp = *IO_agbp; + goto out_alloc; } - agino = XFS_INO_TO_AGINO(mp, inode); - if (inode != XFS_AGINO_TO_INO(mp, agno, agino)) { - cmn_err(CE_WARN, - "xfs_difree: inode != XFS_AGINO_TO_INO() " - "(%llu != %llu) on %s. Returning EINVAL.", - (unsigned long long)inode, - (unsigned long long)XFS_AGINO_TO_INO(mp, agno, agino), - mp->m_fsname); - ASSERT(0); - return XFS_ERROR(EINVAL); + + /* + * We do not have an agbp, so select an initial allocation + * group for inode allocation. + */ + start_agno = xfs_ialloc_ag_select(tp, parent, mode, okalloc); + if (start_agno == NULLAGNUMBER) { + *inop = NULLFSINO; + return 0; } - agbno = XFS_AGINO_TO_AGBNO(mp, agino); - if (agbno >= mp->m_sb.sb_agblocks) { - cmn_err(CE_WARN, - "xfs_difree: agbno >= mp->m_sb.sb_agblocks (%d >= %d) on %s. Returning EINVAL.", - agbno, mp->m_sb.sb_agblocks, mp->m_fsname); - ASSERT(0); - return XFS_ERROR(EINVAL); + + /* + * If we have already hit the ceiling of inode blocks then clear + * okalloc so we scan all available agi structures for a free + * inode. + */ + if (mp->m_maxicount && + mp->m_sb.sb_icount + mp->m_ialloc_inos > mp->m_maxicount) { + noroom = 1; + okalloc = 0; } + /* - * Get the allocation group header. + * Loop until we find an allocation group that either has free inodes + * or in which we can allocate some inodes. Iterate through the + * allocation groups upward, wrapping at the end. */ - down_read(&mp->m_peraglock); - error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); - up_read(&mp->m_peraglock); - if (error) { - cmn_err(CE_WARN, - "xfs_difree: xfs_ialloc_read_agi() returned an error %d on %s. Returning error.", - error, mp->m_fsname); - return error; + agno = start_agno; + for (;;) { + pag = xfs_perag_get(mp, agno); + if (!pag->pagi_inodeok) { + xfs_ialloc_next_ag(mp); + goto nextag; + } + + if (!pag->pagi_init) { + error = xfs_ialloc_pagi_init(mp, tp, agno); + if (error) + goto out_error; + } + + /* + * Do a first racy fast path check if this AG is usable. + */ + if (!pag->pagi_freecount && !okalloc) + goto nextag; + + /* + * Then read in the AGI buffer and recheck with the AGI buffer + * lock held. + */ + error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); + if (error) + goto out_error; + + if (pag->pagi_freecount) { + xfs_perag_put(pag); + goto out_alloc; + } + + if (!okalloc) + goto nextag_relse_buffer; + + + error = xfs_ialloc_ag_alloc(tp, agbp, &ialloced); + if (error) { + xfs_trans_brelse(tp, agbp); + + if (error != ENOSPC) + goto out_error; + + xfs_perag_put(pag); + *inop = NULLFSINO; + return 0; + } + + if (ialloced) { + /* + * We successfully allocated some inodes, return + * the current context to the caller so that it + * can commit the current transaction and call + * us again where we left off. + */ + ASSERT(pag->pagi_freecount > 0); + xfs_perag_put(pag); + + *IO_agbp = agbp; + *inop = NULLFSINO; + return 0; + } + +nextag_relse_buffer: + xfs_trans_brelse(tp, agbp); +nextag: + xfs_perag_put(pag); + if (++agno == mp->m_sb.sb_agcount) + agno = 0; + if (agno == start_agno) { + *inop = NULLFSINO; + return noroom ? ENOSPC : 0; + } } - agi = XFS_BUF_TO_AGI(agbp); - ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC); - ASSERT(agbno < be32_to_cpu(agi->agi_length)); + +out_alloc: + *IO_agbp = NULL; + return xfs_dialloc_ag(tp, agbp, parent, inop); +out_error: + xfs_perag_put(pag); + return XFS_ERROR(error); +} + +STATIC int +xfs_difree_inobt( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct xfs_buf *agbp, + xfs_agino_t agino, + struct xfs_bmap_free *flist, + int *deleted, + xfs_ino_t *first_ino, + struct xfs_inobt_rec_incore *orec) +{ + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + struct xfs_perag *pag; + struct xfs_btree_cur *cur; + struct xfs_inobt_rec_incore rec; + int ilen; + int error; + int i; + int off; + + ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC)); + ASSERT(XFS_AGINO_TO_AGBNO(mp, agino) < be32_to_cpu(agi->agi_length)); + /* * Initialize the cursor. */ - cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); -#ifdef DEBUG - if (cur->bc_nlevels == 1) { - int freecount = 0; + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO); + + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error0; - if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i))) - goto error0; - do { - if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino, - &rec.ir_freecount, &rec.ir_free, &i))) - goto error0; - if (i) { - freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) - goto error0; - } - } while (i == 1); - ASSERT(freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); - } -#endif /* * Look for the entry describing this inode. */ - if ((error = xfs_inobt_lookup_le(cur, agino, 0, 0, &i))) { - cmn_err(CE_WARN, - "xfs_difree: xfs_inobt_lookup_le returned() an error %d on %s. Returning error.", - error, mp->m_fsname); + if ((error = xfs_inobt_lookup(cur, agino, XFS_LOOKUP_LE, &i))) { + xfs_warn(mp, "%s: xfs_inobt_lookup() returned error %d.", + __func__, error); goto error0; } XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino, &rec.ir_freecount, - &rec.ir_free, &i))) { - cmn_err(CE_WARN, - "xfs_difree: xfs_inobt_get_rec() returned an error %d on %s. Returning error.", - error, mp->m_fsname); + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) { + xfs_warn(mp, "%s: xfs_inobt_get_rec() returned error %d.", + __func__, error); goto error0; } XFS_WANT_CORRUPTED_GOTO(i == 1, error0); @@ -1043,20 +1482,20 @@ xfs_difree( */ off = agino - rec.ir_startino; ASSERT(off >= 0 && off < XFS_INODES_PER_CHUNK); - ASSERT(!XFS_INOBT_IS_FREE(&rec, off)); + ASSERT(!(rec.ir_free & XFS_INOBT_MASK(off))); /* * Mark the inode free & increment the count. */ - XFS_INOBT_SET_FREE(&rec, off); + rec.ir_free |= XFS_INOBT_MASK(off); rec.ir_freecount++; /* * When an inode cluster is free, it becomes eligible for removal */ if (!(mp->m_flags & XFS_MOUNT_IKEEP) && - (rec.ir_freecount == XFS_IALLOC_INODES(mp))) { + (rec.ir_freecount == mp->m_ialloc_inos)) { - *delete = 1; + *deleted = 1; *first_ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino); /* @@ -1064,67 +1503,51 @@ xfs_difree( * AGI and Superblock inode counts, and mark the disk space * to be freed when the transaction is committed. */ - ilen = XFS_IALLOC_INODES(mp); + ilen = mp->m_ialloc_inos; be32_add_cpu(&agi->agi_count, -ilen); be32_add_cpu(&agi->agi_freecount, -(ilen - 1)); xfs_ialloc_log_agi(tp, agbp, XFS_AGI_COUNT | XFS_AGI_FREECOUNT); - down_read(&mp->m_peraglock); - mp->m_perag[agno].pagi_freecount -= ilen - 1; - up_read(&mp->m_peraglock); + pag = xfs_perag_get(mp, agno); + pag->pagi_freecount -= ilen - 1; + xfs_perag_put(pag); xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, -ilen); xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -(ilen - 1)); - if ((error = xfs_inobt_delete(cur, &i))) { - cmn_err(CE_WARN, "xfs_difree: xfs_inobt_delete returned an error %d on %s.\n", - error, mp->m_fsname); + if ((error = xfs_btree_delete(cur, &i))) { + xfs_warn(mp, "%s: xfs_btree_delete returned error %d.", + __func__, error); goto error0; } - xfs_bmap_add_free(XFS_AGB_TO_FSB(mp, - agno, XFS_INO_TO_AGBNO(mp,rec.ir_startino)), - XFS_IALLOC_BLOCKS(mp), flist, mp); + xfs_bmap_add_free(XFS_AGB_TO_FSB(mp, agno, + XFS_AGINO_TO_AGBNO(mp, rec.ir_startino)), + mp->m_ialloc_blks, flist, mp); } else { - *delete = 0; + *deleted = 0; - if ((error = xfs_inobt_update(cur, rec.ir_startino, rec.ir_freecount, rec.ir_free))) { - cmn_err(CE_WARN, - "xfs_difree: xfs_inobt_update() returned an error %d on %s. Returning error.", - error, mp->m_fsname); + error = xfs_inobt_update(cur, &rec); + if (error) { + xfs_warn(mp, "%s: xfs_inobt_update returned error %d.", + __func__, error); goto error0; } + /* * Change the inode free counts and log the ag/sb changes. */ be32_add_cpu(&agi->agi_freecount, 1); xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT); - down_read(&mp->m_peraglock); - mp->m_perag[agno].pagi_freecount++; - up_read(&mp->m_peraglock); + pag = xfs_perag_get(mp, agno); + pag->pagi_freecount++; + xfs_perag_put(pag); xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, 1); } -#ifdef DEBUG - if (cur->bc_nlevels == 1) { - int freecount = 0; + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error0; - if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i))) - goto error0; - do { - if ((error = xfs_inobt_get_rec(cur, - &rec.ir_startino, - &rec.ir_freecount, - &rec.ir_free, &i))) - goto error0; - if (i) { - freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) - goto error0; - } - } while (i == 1); - ASSERT(freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); - } -#endif + *orec = rec; xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); return 0; @@ -1134,36 +1557,264 @@ error0: } /* - * Return the location of the inode in bno/off, for mapping it into a buffer. + * Free an inode in the free inode btree. + */ +STATIC int +xfs_difree_finobt( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct xfs_buf *agbp, + xfs_agino_t agino, + struct xfs_inobt_rec_incore *ibtrec) /* inobt record */ +{ + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + struct xfs_btree_cur *cur; + struct xfs_inobt_rec_incore rec; + int offset = agino - ibtrec->ir_startino; + int error; + int i; + + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_FINO); + + error = xfs_inobt_lookup(cur, ibtrec->ir_startino, XFS_LOOKUP_EQ, &i); + if (error) + goto error; + if (i == 0) { + /* + * If the record does not exist in the finobt, we must have just + * freed an inode in a previously fully allocated chunk. If not, + * something is out of sync. + */ + XFS_WANT_CORRUPTED_GOTO(ibtrec->ir_freecount == 1, error); + + error = xfs_inobt_insert_rec(cur, ibtrec->ir_freecount, + ibtrec->ir_free, &i); + if (error) + goto error; + ASSERT(i == 1); + + goto out; + } + + /* + * Read and update the existing record. We could just copy the ibtrec + * across here, but that would defeat the purpose of having redundant + * metadata. By making the modifications independently, we can catch + * corruptions that we wouldn't see if we just copied from one record + * to another. + */ + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) + goto error; + XFS_WANT_CORRUPTED_GOTO(i == 1, error); + + rec.ir_free |= XFS_INOBT_MASK(offset); + rec.ir_freecount++; + + XFS_WANT_CORRUPTED_GOTO((rec.ir_free == ibtrec->ir_free) && + (rec.ir_freecount == ibtrec->ir_freecount), + error); + + /* + * The content of inobt records should always match between the inobt + * and finobt. The lifecycle of records in the finobt is different from + * the inobt in that the finobt only tracks records with at least one + * free inode. Hence, if all of the inodes are free and we aren't + * keeping inode chunks permanently on disk, remove the record. + * Otherwise, update the record with the new information. + */ + if (rec.ir_freecount == mp->m_ialloc_inos && + !(mp->m_flags & XFS_MOUNT_IKEEP)) { + error = xfs_btree_delete(cur, &i); + if (error) + goto error; + ASSERT(i == 1); + } else { + error = xfs_inobt_update(cur, &rec); + if (error) + goto error; + } + +out: + error = xfs_check_agi_freecount(cur, agi); + if (error) + goto error; + + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + return 0; + +error: + xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + return error; +} + +/* + * Free disk inode. Carefully avoids touching the incore inode, all + * manipulations incore are the caller's responsibility. + * The on-disk inode is not changed by this operation, only the + * btree (free inode mask) is changed. */ -/*ARGSUSED*/ int -xfs_dilocate( - xfs_mount_t *mp, /* file system mount structure */ - xfs_trans_t *tp, /* transaction pointer */ +xfs_difree( + struct xfs_trans *tp, /* transaction pointer */ + xfs_ino_t inode, /* inode to be freed */ + struct xfs_bmap_free *flist, /* extents to free */ + int *deleted,/* set if inode cluster was deleted */ + xfs_ino_t *first_ino)/* first inode in deleted cluster */ +{ + /* REFERENCED */ + xfs_agblock_t agbno; /* block number containing inode */ + struct xfs_buf *agbp; /* buffer for allocation group header */ + xfs_agino_t agino; /* allocation group inode number */ + xfs_agnumber_t agno; /* allocation group number */ + int error; /* error return value */ + struct xfs_mount *mp; /* mount structure for filesystem */ + struct xfs_inobt_rec_incore rec;/* btree record */ + + mp = tp->t_mountp; + + /* + * Break up inode number into its components. + */ + agno = XFS_INO_TO_AGNO(mp, inode); + if (agno >= mp->m_sb.sb_agcount) { + xfs_warn(mp, "%s: agno >= mp->m_sb.sb_agcount (%d >= %d).", + __func__, agno, mp->m_sb.sb_agcount); + ASSERT(0); + return XFS_ERROR(EINVAL); + } + agino = XFS_INO_TO_AGINO(mp, inode); + if (inode != XFS_AGINO_TO_INO(mp, agno, agino)) { + xfs_warn(mp, "%s: inode != XFS_AGINO_TO_INO() (%llu != %llu).", + __func__, (unsigned long long)inode, + (unsigned long long)XFS_AGINO_TO_INO(mp, agno, agino)); + ASSERT(0); + return XFS_ERROR(EINVAL); + } + agbno = XFS_AGINO_TO_AGBNO(mp, agino); + if (agbno >= mp->m_sb.sb_agblocks) { + xfs_warn(mp, "%s: agbno >= mp->m_sb.sb_agblocks (%d >= %d).", + __func__, agbno, mp->m_sb.sb_agblocks); + ASSERT(0); + return XFS_ERROR(EINVAL); + } + /* + * Get the allocation group header. + */ + error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); + if (error) { + xfs_warn(mp, "%s: xfs_ialloc_read_agi() returned error %d.", + __func__, error); + return error; + } + + /* + * Fix up the inode allocation btree. + */ + error = xfs_difree_inobt(mp, tp, agbp, agino, flist, deleted, first_ino, + &rec); + if (error) + goto error0; + + /* + * Fix up the free inode btree. + */ + if (xfs_sb_version_hasfinobt(&mp->m_sb)) { + error = xfs_difree_finobt(mp, tp, agbp, agino, &rec); + if (error) + goto error0; + } + + return 0; + +error0: + return error; +} + +STATIC int +xfs_imap_lookup( + struct xfs_mount *mp, + struct xfs_trans *tp, + xfs_agnumber_t agno, + xfs_agino_t agino, + xfs_agblock_t agbno, + xfs_agblock_t *chunk_agbno, + xfs_agblock_t *offset_agbno, + int flags) +{ + struct xfs_inobt_rec_incore rec; + struct xfs_btree_cur *cur; + struct xfs_buf *agbp; + int error; + int i; + + error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); + if (error) { + xfs_alert(mp, + "%s: xfs_ialloc_read_agi() returned error %d, agno %d", + __func__, error, agno); + return error; + } + + /* + * Lookup the inode record for the given agino. If the record cannot be + * found, then it's an invalid inode number and we should abort. Once + * we have a record, we need to ensure it contains the inode number + * we are looking up. + */ + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO); + error = xfs_inobt_lookup(cur, agino, XFS_LOOKUP_LE, &i); + if (!error) { + if (i) + error = xfs_inobt_get_rec(cur, &rec, &i); + if (!error && i == 0) + error = EINVAL; + } + + xfs_trans_brelse(tp, agbp); + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + if (error) + return error; + + /* check that the returned record contains the required inode */ + if (rec.ir_startino > agino || + rec.ir_startino + mp->m_ialloc_inos <= agino) + return EINVAL; + + /* for untrusted inodes check it is allocated first */ + if ((flags & XFS_IGET_UNTRUSTED) && + (rec.ir_free & XFS_INOBT_MASK(agino - rec.ir_startino))) + return EINVAL; + + *chunk_agbno = XFS_AGINO_TO_AGBNO(mp, rec.ir_startino); + *offset_agbno = agbno - *chunk_agbno; + return 0; +} + +/* + * Return the location of the inode in imap, for mapping it into a buffer. + */ +int +xfs_imap( + xfs_mount_t *mp, /* file system mount structure */ + xfs_trans_t *tp, /* transaction pointer */ xfs_ino_t ino, /* inode to locate */ - xfs_fsblock_t *bno, /* output: block containing inode */ - int *len, /* output: num blocks in inode cluster */ - int *off, /* output: index in block of inode */ - uint flags) /* flags concerning inode lookup */ + struct xfs_imap *imap, /* location map structure */ + uint flags) /* flags for inode btree lookup */ { xfs_agblock_t agbno; /* block number of inode in the alloc group */ - xfs_buf_t *agbp; /* agi buffer */ xfs_agino_t agino; /* inode number within alloc group */ xfs_agnumber_t agno; /* allocation group number */ int blks_per_cluster; /* num blocks per inode cluster */ xfs_agblock_t chunk_agbno; /* first block in inode chunk */ - xfs_agino_t chunk_agino; /* first agino in inode chunk */ - __int32_t chunk_cnt; /* count of free inodes in chunk */ - xfs_inofree_t chunk_free; /* mask of free inodes in chunk */ xfs_agblock_t cluster_agbno; /* first block in inode cluster */ - xfs_btree_cur_t *cur; /* inode btree cursor */ int error; /* error code */ - int i; /* temp state */ int offset; /* index of inode in its buffer */ - int offset_agbno; /* blks from chunk start to inode */ + xfs_agblock_t offset_agbno; /* blks from chunk start to inode */ ASSERT(ino != NULLFSINO); + /* * Split up the inode number into its parts. */ @@ -1173,112 +1824,107 @@ xfs_dilocate( if (agno >= mp->m_sb.sb_agcount || agbno >= mp->m_sb.sb_agblocks || ino != XFS_AGINO_TO_INO(mp, agno, agino)) { #ifdef DEBUG - /* no diagnostics for bulkstat, ino comes from userspace */ - if (flags & XFS_IMAP_BULKSTAT) + /* + * Don't output diagnostic information for untrusted inodes + * as they can be invalid without implying corruption. + */ + if (flags & XFS_IGET_UNTRUSTED) return XFS_ERROR(EINVAL); if (agno >= mp->m_sb.sb_agcount) { - xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: agno (%d) >= " - "mp->m_sb.sb_agcount (%d)", - agno, mp->m_sb.sb_agcount); + xfs_alert(mp, + "%s: agno (%d) >= mp->m_sb.sb_agcount (%d)", + __func__, agno, mp->m_sb.sb_agcount); } if (agbno >= mp->m_sb.sb_agblocks) { - xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: agbno (0x%llx) >= " - "mp->m_sb.sb_agblocks (0x%lx)", - (unsigned long long) agbno, - (unsigned long) mp->m_sb.sb_agblocks); + xfs_alert(mp, + "%s: agbno (0x%llx) >= mp->m_sb.sb_agblocks (0x%lx)", + __func__, (unsigned long long)agbno, + (unsigned long)mp->m_sb.sb_agblocks); } if (ino != XFS_AGINO_TO_INO(mp, agno, agino)) { - xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: ino (0x%llx) != " - "XFS_AGINO_TO_INO(mp, agno, agino) " - "(0x%llx)", - ino, XFS_AGINO_TO_INO(mp, agno, agino)); + xfs_alert(mp, + "%s: ino (0x%llx) != XFS_AGINO_TO_INO() (0x%llx)", + __func__, ino, + XFS_AGINO_TO_INO(mp, agno, agino)); } xfs_stack_trace(); #endif /* DEBUG */ return XFS_ERROR(EINVAL); } - if ((mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) || - !(flags & XFS_IMAP_LOOKUP)) { - offset = XFS_INO_TO_OFFSET(mp, ino); - ASSERT(offset < mp->m_sb.sb_inopblock); - *bno = XFS_AGB_TO_FSB(mp, agno, agbno); - *off = offset; - *len = 1; - return 0; + + blks_per_cluster = xfs_icluster_size_fsb(mp); + + /* + * For bulkstat and handle lookups, we have an untrusted inode number + * that we have to verify is valid. We cannot do this just by reading + * the inode buffer as it may have been unlinked and removed leaving + * inodes in stale state on disk. Hence we have to do a btree lookup + * in all cases where an untrusted inode number is passed. + */ + if (flags & XFS_IGET_UNTRUSTED) { + error = xfs_imap_lookup(mp, tp, agno, agino, agbno, + &chunk_agbno, &offset_agbno, flags); + if (error) + return error; + goto out_map; } - blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_blocklog; - if (*bno != NULLFSBLOCK) { + + /* + * If the inode cluster size is the same as the blocksize or + * smaller we get to the buffer by simple arithmetics. + */ + if (blks_per_cluster == 1) { offset = XFS_INO_TO_OFFSET(mp, ino); ASSERT(offset < mp->m_sb.sb_inopblock); - cluster_agbno = XFS_FSB_TO_AGBNO(mp, *bno); - *off = ((agbno - cluster_agbno) * mp->m_sb.sb_inopblock) + - offset; - *len = blks_per_cluster; + + imap->im_blkno = XFS_AGB_TO_DADDR(mp, agno, agbno); + imap->im_len = XFS_FSB_TO_BB(mp, 1); + imap->im_boffset = (ushort)(offset << mp->m_sb.sb_inodelog); return 0; } + + /* + * If the inode chunks are aligned then use simple maths to + * find the location. Otherwise we have to do a btree + * lookup to find the location. + */ if (mp->m_inoalign_mask) { offset_agbno = agbno & mp->m_inoalign_mask; chunk_agbno = agbno - offset_agbno; } else { - down_read(&mp->m_peraglock); - error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); - up_read(&mp->m_peraglock); - if (error) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " - "xfs_ialloc_read_agi() returned " - "error %d, agno %d", - error, agno); -#endif /* DEBUG */ - return error; - } - cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); - if ((error = xfs_inobt_lookup_le(cur, agino, 0, 0, &i))) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " - "xfs_inobt_lookup_le() failed"); -#endif /* DEBUG */ - goto error0; - } - if ((error = xfs_inobt_get_rec(cur, &chunk_agino, &chunk_cnt, - &chunk_free, &i))) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " - "xfs_inobt_get_rec() failed"); -#endif /* DEBUG */ - goto error0; - } - if (i == 0) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " - "xfs_inobt_get_rec() failed"); -#endif /* DEBUG */ - error = XFS_ERROR(EINVAL); - } - xfs_trans_brelse(tp, agbp); - xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + error = xfs_imap_lookup(mp, tp, agno, agino, agbno, + &chunk_agbno, &offset_agbno, flags); if (error) return error; - chunk_agbno = XFS_AGINO_TO_AGBNO(mp, chunk_agino); - offset_agbno = agbno - chunk_agbno; } + +out_map: ASSERT(agbno >= chunk_agbno); cluster_agbno = chunk_agbno + ((offset_agbno / blks_per_cluster) * blks_per_cluster); offset = ((agbno - cluster_agbno) * mp->m_sb.sb_inopblock) + XFS_INO_TO_OFFSET(mp, ino); - *bno = XFS_AGB_TO_FSB(mp, agno, cluster_agbno); - *off = offset; - *len = blks_per_cluster; + + imap->im_blkno = XFS_AGB_TO_DADDR(mp, agno, cluster_agbno); + imap->im_len = XFS_FSB_TO_BB(mp, blks_per_cluster); + imap->im_boffset = (ushort)(offset << mp->m_sb.sb_inodelog); + + /* + * If the inode number maps to a block outside the bounds + * of the file system then return NULL rather than calling + * read_buf and panicing when we get an error from the + * driver. + */ + if ((imap->im_blkno + imap->im_len) > + XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)) { + xfs_alert(mp, + "%s: (im_blkno (0x%llx) + im_len (0x%llx)) > sb_dblocks (0x%llx)", + __func__, (unsigned long long) imap->im_blkno, + (unsigned long long) imap->im_len, + XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)); + return XFS_ERROR(EINVAL); + } return 0; -error0: - xfs_trans_brelse(tp, agbp); - xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); - return error; } /* @@ -1305,7 +1951,16 @@ xfs_ialloc_compute_maxlevels( } /* - * Log specified fields for the ag hdr (inode section) + * Log specified fields for the ag hdr (inode section). The growth of the agi + * structure over time requires that we interpret the buffer as two logical + * regions delineated by the end of the unlinked list. This is due to the size + * of the hash table and its location in the middle of the agi. + * + * For example, a request to log a field before agi_unlinked and a field after + * agi_unlinked could cause us to log the entire hash table and use an excessive + * amount of log space. To avoid this behavior, log the region up through + * agi_unlinked in one call and the region after agi_unlinked through the end of + * the structure in another. */ void xfs_ialloc_log_agi( @@ -1328,88 +1983,188 @@ xfs_ialloc_log_agi( offsetof(xfs_agi_t, agi_newino), offsetof(xfs_agi_t, agi_dirino), offsetof(xfs_agi_t, agi_unlinked), + offsetof(xfs_agi_t, agi_free_root), + offsetof(xfs_agi_t, agi_free_level), sizeof(xfs_agi_t) }; #ifdef DEBUG xfs_agi_t *agi; /* allocation group header */ agi = XFS_BUF_TO_AGI(bp); - ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC); + ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC)); +#endif + + xfs_trans_buf_set_type(tp, bp, XFS_BLFT_AGI_BUF); + + /* + * Compute byte offsets for the first and last fields in the first + * region and log the agi buffer. This only logs up through + * agi_unlinked. + */ + if (fields & XFS_AGI_ALL_BITS_R1) { + xfs_btree_offsets(fields, offsets, XFS_AGI_NUM_BITS_R1, + &first, &last); + xfs_trans_log_buf(tp, bp, first, last); + } + + /* + * Mask off the bits in the first region and calculate the first and + * last field offsets for any bits in the second region. + */ + fields &= ~XFS_AGI_ALL_BITS_R1; + if (fields) { + xfs_btree_offsets(fields, offsets, XFS_AGI_NUM_BITS_R2, + &first, &last); + xfs_trans_log_buf(tp, bp, first, last); + } +} + +#ifdef DEBUG +STATIC void +xfs_check_agi_unlinked( + struct xfs_agi *agi) +{ + int i; + + for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) + ASSERT(agi->agi_unlinked[i]); +} +#else +#define xfs_check_agi_unlinked(agi) #endif + +static bool +xfs_agi_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_agi *agi = XFS_BUF_TO_AGI(bp); + + if (xfs_sb_version_hascrc(&mp->m_sb) && + !uuid_equal(&agi->agi_uuid, &mp->m_sb.sb_uuid)) + return false; /* - * Compute byte offsets for the first and last fields. + * Validate the magic number of the agi block. */ - xfs_btree_offsets(fields, offsets, XFS_AGI_NUM_BITS, &first, &last); + if (agi->agi_magicnum != cpu_to_be32(XFS_AGI_MAGIC)) + return false; + if (!XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum))) + return false; + /* - * Log the allocation group inode header buffer. + * during growfs operations, the perag is not fully initialised, + * so we can't use it for any useful checking. growfs ensures we can't + * use it by using uncached buffers that don't have the perag attached + * so we can detect and avoid this problem. */ - xfs_trans_log_buf(tp, bp, first, last); + if (bp->b_pag && be32_to_cpu(agi->agi_seqno) != bp->b_pag->pag_agno) + return false; + + xfs_check_agi_unlinked(agi); + return true; +} + +static void +xfs_agi_read_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + + if (xfs_sb_version_hascrc(&mp->m_sb) && + !xfs_buf_verify_cksum(bp, XFS_AGI_CRC_OFF)) + xfs_buf_ioerror(bp, EFSBADCRC); + else if (XFS_TEST_ERROR(!xfs_agi_verify(bp), mp, + XFS_ERRTAG_IALLOC_READ_AGI, + XFS_RANDOM_IALLOC_READ_AGI)) + xfs_buf_ioerror(bp, EFSCORRUPTED); + + if (bp->b_error) + xfs_verifier_error(bp); } +static void +xfs_agi_write_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_buf_log_item *bip = bp->b_fspriv; + + if (!xfs_agi_verify(bp)) { + xfs_buf_ioerror(bp, EFSCORRUPTED); + xfs_verifier_error(bp); + return; + } + + if (!xfs_sb_version_hascrc(&mp->m_sb)) + return; + + if (bip) + XFS_BUF_TO_AGI(bp)->agi_lsn = cpu_to_be64(bip->bli_item.li_lsn); + xfs_buf_update_cksum(bp, XFS_AGI_CRC_OFF); +} + +const struct xfs_buf_ops xfs_agi_buf_ops = { + .verify_read = xfs_agi_read_verify, + .verify_write = xfs_agi_write_verify, +}; + /* * Read in the allocation group header (inode allocation section) */ int -xfs_ialloc_read_agi( - xfs_mount_t *mp, /* file system mount structure */ - xfs_trans_t *tp, /* transaction pointer */ - xfs_agnumber_t agno, /* allocation group number */ - xfs_buf_t **bpp) /* allocation group hdr buf */ +xfs_read_agi( + struct xfs_mount *mp, /* file system mount structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + struct xfs_buf **bpp) /* allocation group hdr buf */ { - xfs_agi_t *agi; /* allocation group header */ - int agi_ok; /* agi is consistent */ - xfs_buf_t *bp; /* allocation group hdr buf */ - xfs_perag_t *pag; /* per allocation group data */ - int error; + int error; + + trace_xfs_read_agi(mp, agno); ASSERT(agno != NULLAGNUMBER); - error = xfs_trans_read_buf( - mp, tp, mp->m_ddev_targp, + error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0, &bp); + XFS_FSS_TO_BB(mp, 1), 0, bpp, &xfs_agi_buf_ops); if (error) return error; - ASSERT(bp && !XFS_BUF_GETERROR(bp)); - /* - * Validate the magic number of the agi block. - */ - agi = XFS_BUF_TO_AGI(bp); - agi_ok = - be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC && - XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)); - if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI, - XFS_RANDOM_IALLOC_READ_AGI))) { - XFS_CORRUPTION_ERROR("xfs_ialloc_read_agi", XFS_ERRLEVEL_LOW, - mp, agi); - xfs_trans_brelse(tp, bp); - return XFS_ERROR(EFSCORRUPTED); - } - pag = &mp->m_perag[agno]; + xfs_buf_set_ref(*bpp, XFS_AGI_REF); + return 0; +} + +int +xfs_ialloc_read_agi( + struct xfs_mount *mp, /* file system mount structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + struct xfs_buf **bpp) /* allocation group hdr buf */ +{ + struct xfs_agi *agi; /* allocation group header */ + struct xfs_perag *pag; /* per allocation group data */ + int error; + + trace_xfs_ialloc_read_agi(mp, agno); + + error = xfs_read_agi(mp, tp, agno, bpp); + if (error) + return error; + + agi = XFS_BUF_TO_AGI(*bpp); + pag = xfs_perag_get(mp, agno); if (!pag->pagi_init) { pag->pagi_freecount = be32_to_cpu(agi->agi_freecount); pag->pagi_count = be32_to_cpu(agi->agi_count); pag->pagi_init = 1; - } else { - /* - * It's possible for these to be out of sync if - * we are in the middle of a forced shutdown. - */ - ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); } -#ifdef DEBUG - { - int i; - - for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) - ASSERT(agi->agi_unlinked[i]); - } -#endif - - XFS_BUF_SET_VTYPE_REF(bp, B_FS_AGI, XFS_AGI_REF); - *bpp = bp; + /* + * It's possible for these to be out of sync if + * we are in the middle of a forced shutdown. + */ + ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) || + XFS_FORCED_SHUTDOWN(mp)); + xfs_perag_put(pag); return 0; } |