diff options
Diffstat (limited to 'fs/xfs')
-rw-r--r-- | fs/xfs/kmem.h | 10 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_attr.c | 4 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_bmap.c | 21 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_btree.c | 6 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_dir2_data.c | 4 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_ialloc.c | 2 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_quota_defs.h | 2 | ||||
-rw-r--r-- | fs/xfs/libxfs/xfs_refcount.c | 4 | ||||
-rw-r--r-- | fs/xfs/xfs_attr_list.c | 2 | ||||
-rw-r--r-- | fs/xfs/xfs_dquot.c | 14 | ||||
-rw-r--r-- | fs/xfs/xfs_inode.c | 12 | ||||
-rw-r--r-- | fs/xfs/xfs_log.c | 11 | ||||
-rw-r--r-- | fs/xfs/xfs_log_cil.c | 1 | ||||
-rw-r--r-- | fs/xfs/xfs_mount.c | 12 | ||||
-rw-r--r-- | fs/xfs/xfs_qm.c | 3 | ||||
-rw-r--r-- | fs/xfs/xfs_reflink.c | 4 | ||||
-rw-r--r-- | fs/xfs/xfs_symlink.c | 2 |
17 files changed, 82 insertions, 32 deletions
diff --git a/fs/xfs/kmem.h b/fs/xfs/kmem.h index d6ea520162b2..4d85992d75b2 100644 --- a/fs/xfs/kmem.h +++ b/fs/xfs/kmem.h @@ -54,6 +54,16 @@ kmem_flags_convert(xfs_km_flags_t flags) lflags &= ~__GFP_FS; } + /* + * Default page/slab allocator behavior is to retry for ever + * for small allocations. We can override this behavior by using + * __GFP_RETRY_MAYFAIL which will tell the allocator to retry as long + * as it is feasible but rather fail than retry forever for all + * request sizes. + */ + if (flags & KM_MAYFAIL) + lflags |= __GFP_RETRY_MAYFAIL; + if (flags & KM_ZERO) lflags |= __GFP_ZERO; diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c index ef8a1c75a467..de7b9bd30bec 100644 --- a/fs/xfs/libxfs/xfs_attr.c +++ b/fs/xfs/libxfs/xfs_attr.c @@ -114,12 +114,14 @@ xfs_inode_hasattr( * Overall external interface routines. *========================================================================*/ -/* Retrieve an extended attribute and its value. Must have iolock. */ +/* Retrieve an extended attribute and its value. Must have ilock. */ int xfs_attr_get_ilocked( struct xfs_inode *ip, struct xfs_da_args *args) { + ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + if (!xfs_inode_hasattr(ip)) return -ENOATTR; else if (ip->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index 0a9880777c9c..c09c16b1ad3b 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -5435,6 +5435,7 @@ __xfs_bunmapi( xfs_fsblock_t sum; xfs_filblks_t len = *rlen; /* length to unmap in file */ xfs_fileoff_t max_len; + xfs_agnumber_t prev_agno = NULLAGNUMBER, agno; trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_); @@ -5534,6 +5535,17 @@ __xfs_bunmapi( */ del = got; wasdel = isnullstartblock(del.br_startblock); + + /* + * Make sure we don't touch multiple AGF headers out of order + * in a single transaction, as that could cause AB-BA deadlocks. + */ + if (!wasdel) { + agno = XFS_FSB_TO_AGNO(mp, del.br_startblock); + if (prev_agno != NULLAGNUMBER && prev_agno > agno) + break; + prev_agno = agno; + } if (got.br_startoff < start) { del.br_startoff = start; del.br_blockcount -= start - got.br_startoff; @@ -6499,6 +6511,15 @@ xfs_bmap_finish_one( xfs_fsblock_t firstfsb; int error = 0; + /* + * firstfsb is tied to the transaction lifetime and is used to + * ensure correct AG locking order and schedule work item + * continuations. XFS_BUI_MAX_FAST_EXTENTS (== 1) restricts us + * to only making one bmap call per transaction, so it should + * be safe to have it as a local variable here. + */ + firstfsb = NULLFSBLOCK; + trace_xfs_bmap_deferred(tp->t_mountp, XFS_FSB_TO_AGNO(tp->t_mountp, startblock), type, XFS_FSB_TO_AGBNO(tp->t_mountp, startblock), diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c index 4da85fff69ad..e0bcc4a59efd 100644 --- a/fs/xfs/libxfs/xfs_btree.c +++ b/fs/xfs/libxfs/xfs_btree.c @@ -728,7 +728,8 @@ xfs_btree_firstrec( * Get the block pointer for this level. */ block = xfs_btree_get_block(cur, level, &bp); - xfs_btree_check_block(cur, block, level, bp); + if (xfs_btree_check_block(cur, block, level, bp)) + return 0; /* * It's empty, there is no such record. */ @@ -757,7 +758,8 @@ xfs_btree_lastrec( * Get the block pointer for this level. */ block = xfs_btree_get_block(cur, level, &bp); - xfs_btree_check_block(cur, block, level, bp); + if (xfs_btree_check_block(cur, block, level, bp)) + return 0; /* * It's empty, there is no such record. */ diff --git a/fs/xfs/libxfs/xfs_dir2_data.c b/fs/xfs/libxfs/xfs_dir2_data.c index d478065b9544..8727a43115ef 100644 --- a/fs/xfs/libxfs/xfs_dir2_data.c +++ b/fs/xfs/libxfs/xfs_dir2_data.c @@ -136,6 +136,8 @@ __xfs_dir3_data_check( */ if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) { XFS_WANT_CORRUPTED_RETURN(mp, lastfree == 0); + XFS_WANT_CORRUPTED_RETURN(mp, endp >= + p + be16_to_cpu(dup->length)); XFS_WANT_CORRUPTED_RETURN(mp, be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)) == (char *)dup - (char *)hdr); @@ -164,6 +166,8 @@ __xfs_dir3_data_check( XFS_WANT_CORRUPTED_RETURN(mp, dep->namelen != 0); XFS_WANT_CORRUPTED_RETURN(mp, !xfs_dir_ino_validate(mp, be64_to_cpu(dep->inumber))); + XFS_WANT_CORRUPTED_RETURN(mp, endp >= + p + ops->data_entsize(dep->namelen)); XFS_WANT_CORRUPTED_RETURN(mp, be16_to_cpu(*ops->data_entry_tag_p(dep)) == (char *)dep - (char *)hdr); diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c index ffd5a15d1bb6..abf5beaae907 100644 --- a/fs/xfs/libxfs/xfs_ialloc.c +++ b/fs/xfs/libxfs/xfs_ialloc.c @@ -1246,13 +1246,13 @@ xfs_dialloc_ag_inobt( /* 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; + rec = trec; goto alloc_inode; } diff --git a/fs/xfs/libxfs/xfs_quota_defs.h b/fs/xfs/libxfs/xfs_quota_defs.h index 2834574cb6e7..d69c772271cb 100644 --- a/fs/xfs/libxfs/xfs_quota_defs.h +++ b/fs/xfs/libxfs/xfs_quota_defs.h @@ -136,8 +136,6 @@ typedef uint16_t xfs_qwarncnt_t; */ #define XFS_QMOPT_INHERIT 0x1000000 -#define XFS_QMOPT_NOLOCK 0x2000000 /* don't ilock during dqget */ - /* * flags to xfs_trans_mod_dquot. */ diff --git a/fs/xfs/libxfs/xfs_refcount.c b/fs/xfs/libxfs/xfs_refcount.c index 900ea231f9a3..45b1c3b4e047 100644 --- a/fs/xfs/libxfs/xfs_refcount.c +++ b/fs/xfs/libxfs/xfs_refcount.c @@ -1638,6 +1638,10 @@ xfs_refcount_recover_cow_leftovers( error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp); if (error) goto out_trans; + if (!agbp) { + error = -ENOMEM; + goto out_trans; + } cur = xfs_refcountbt_init_cursor(mp, tp, agbp, agno, NULL); /* Find all the leftover CoW staging extents. */ diff --git a/fs/xfs/xfs_attr_list.c b/fs/xfs/xfs_attr_list.c index 545eca508d42..7740c8a5e736 100644 --- a/fs/xfs/xfs_attr_list.c +++ b/fs/xfs/xfs_attr_list.c @@ -463,6 +463,8 @@ xfs_attr_list_int_ilocked( { struct xfs_inode *dp = context->dp; + ASSERT(xfs_isilocked(dp, XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + /* * Decide on what work routines to call based on the inode size. */ diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index f89f7b5241e6..fd2ef8c2c9a7 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -472,23 +472,18 @@ xfs_qm_dqtobp( struct xfs_mount *mp = dqp->q_mount; xfs_dqid_t id = be32_to_cpu(dqp->q_core.d_id); struct xfs_trans *tp = (tpp ? *tpp : NULL); - uint lock_mode = 0; + uint lock_mode; quotip = xfs_quota_inode(dqp->q_mount, dqp->dq_flags); dqp->q_fileoffset = (xfs_fileoff_t)id / mp->m_quotainfo->qi_dqperchunk; - ASSERT(!(flags & XFS_QMOPT_NOLOCK) || - xfs_isilocked(quotip, XFS_ILOCK_SHARED) || - xfs_isilocked(quotip, XFS_ILOCK_EXCL)); - if (!(flags & XFS_QMOPT_NOLOCK)) - lock_mode = xfs_ilock_data_map_shared(quotip); + lock_mode = xfs_ilock_data_map_shared(quotip); if (!xfs_this_quota_on(dqp->q_mount, dqp->dq_flags)) { /* * Return if this type of quotas is turned off while we * didn't have the quota inode lock. */ - if (lock_mode) - xfs_iunlock(quotip, lock_mode); + xfs_iunlock(quotip, lock_mode); return -ESRCH; } @@ -498,8 +493,7 @@ xfs_qm_dqtobp( error = xfs_bmapi_read(quotip, dqp->q_fileoffset, XFS_DQUOT_CLUSTER_SIZE_FSB, &map, &nmaps, 0); - if (lock_mode) - xfs_iunlock(quotip, lock_mode); + xfs_iunlock(quotip, lock_mode); if (error) return error; diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index ceef77c0416a..ff48f0096810 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -874,7 +874,6 @@ xfs_ialloc( case S_IFREG: case S_IFDIR: if (pip && (pip->i_d.di_flags & XFS_DIFLAG_ANY)) { - uint64_t di_flags2 = 0; uint di_flags = 0; if (S_ISDIR(mode)) { @@ -911,20 +910,23 @@ xfs_ialloc( di_flags |= XFS_DIFLAG_NODEFRAG; if (pip->i_d.di_flags & XFS_DIFLAG_FILESTREAM) di_flags |= XFS_DIFLAG_FILESTREAM; - if (pip->i_d.di_flags2 & XFS_DIFLAG2_DAX) - di_flags2 |= XFS_DIFLAG2_DAX; ip->i_d.di_flags |= di_flags; - ip->i_d.di_flags2 |= di_flags2; } if (pip && (pip->i_d.di_flags2 & XFS_DIFLAG2_ANY) && pip->i_d.di_version == 3 && ip->i_d.di_version == 3) { + uint64_t di_flags2 = 0; + if (pip->i_d.di_flags2 & XFS_DIFLAG2_COWEXTSIZE) { - ip->i_d.di_flags2 |= XFS_DIFLAG2_COWEXTSIZE; + di_flags2 |= XFS_DIFLAG2_COWEXTSIZE; ip->i_d.di_cowextsize = pip->i_d.di_cowextsize; } + if (pip->i_d.di_flags2 & XFS_DIFLAG2_DAX) + di_flags2 |= XFS_DIFLAG2_DAX; + + ip->i_d.di_flags2 |= di_flags2; } /* FALLTHROUGH */ case S_IFLNK: diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 0053bcf2b10a..4ebd0bafc914 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -749,9 +749,20 @@ xfs_log_mount_finish( return 0; } + /* + * During the second phase of log recovery, we need iget and + * iput to behave like they do for an active filesystem. + * xfs_fs_drop_inode needs to be able to prevent the deletion + * of inodes before we're done replaying log items on those + * inodes. Turn it off immediately after recovery finishes + * so that we don't leak the quota inodes if subsequent mount + * activities fail. + */ + mp->m_super->s_flags |= MS_ACTIVE; error = xlog_recover_finish(mp->m_log); if (!error) xfs_log_work_queue(mp); + mp->m_super->s_flags &= ~MS_ACTIVE; return error; } diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index fbe72b134bef..43aa42a3a5d3 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -539,6 +539,7 @@ xlog_discard_endio( INIT_WORK(&ctx->discard_endio_work, xlog_discard_endio_work); queue_work(xfs_discard_wq, &ctx->discard_endio_work); + bio_put(bio); } static void diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 40d4e8b4e193..ea7d4b4e50d0 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -945,15 +945,6 @@ xfs_mountfs( } /* - * During the second phase of log recovery, we need iget and - * iput to behave like they do for an active filesystem. - * xfs_fs_drop_inode needs to be able to prevent the deletion - * of inodes before we're done replaying log items on those - * inodes. - */ - mp->m_super->s_flags |= MS_ACTIVE; - - /* * Finish recovering the file system. This part needed to be delayed * until after the root and real-time bitmap inodes were consistently * read in. @@ -1028,12 +1019,13 @@ xfs_mountfs( out_quota: xfs_qm_unmount_quotas(mp); out_rtunmount: - mp->m_super->s_flags &= ~MS_ACTIVE; xfs_rtunmount_inodes(mp); out_rele_rip: IRELE(rip); cancel_delayed_work_sync(&mp->m_reclaim_work); xfs_reclaim_inodes(mp, SYNC_WAIT); + /* Clean out dquots that might be in memory after quotacheck. */ + xfs_qm_unmount(mp); out_log_dealloc: mp->m_flags |= XFS_MOUNT_UNMOUNTING; xfs_log_mount_cancel(mp); diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 6ce948c436d5..15751dc2a27d 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -111,6 +111,9 @@ restart: skipped = 0; break; } + /* we're done if id overflows back to zero */ + if (!next_index) + break; } if (skipped) { diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c index ab2270a87196..f45fbf0db9bb 100644 --- a/fs/xfs/xfs_reflink.c +++ b/fs/xfs/xfs_reflink.c @@ -170,6 +170,8 @@ xfs_reflink_find_shared( error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp); if (error) return error; + if (!agbp) + return -ENOMEM; cur = xfs_refcountbt_init_cursor(mp, tp, agbp, agno, NULL); @@ -329,7 +331,7 @@ xfs_reflink_convert_cow_extent( xfs_filblks_t count_fsb, struct xfs_defer_ops *dfops) { - xfs_fsblock_t first_block; + xfs_fsblock_t first_block = NULLFSBLOCK; int nimaps = 1; if (imap->br_state == XFS_EXT_NORM) diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c index 12cd9cf7de41..23a50d7aa46a 100644 --- a/fs/xfs/xfs_symlink.c +++ b/fs/xfs/xfs_symlink.c @@ -61,6 +61,8 @@ xfs_readlink_bmap_ilocked( int fsblocks = 0; int offset; + ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + fsblocks = xfs_symlink_blocks(mp, pathlen); error = xfs_bmapi_read(ip, 0, fsblocks, mval, &nmaps, 0); if (error) |