From: Darrick J. Wong <darrick.wong@xxxxxxxxxx> Perform some quick sanity testing of the disk quota information. Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx> --- fs/xfs/Makefile | 1 fs/xfs/libxfs/xfs_fs.h | 5 + fs/xfs/scrub/common.c | 18 +++ fs/xfs/scrub/common.h | 2 fs/xfs/scrub/quota.c | 274 ++++++++++++++++++++++++++++++++++++++++++++++++ fs/xfs/xfs_trace.h | 5 + 6 files changed, 303 insertions(+), 2 deletions(-) create mode 100644 fs/xfs/scrub/quota.c diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index 86fc94c..010a90f 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -158,4 +158,5 @@ xfs-y += $(addprefix scrub/, \ ) xfs-$(CONFIG_XFS_RT) += scrub/rtbitmap.o +xfs-$(CONFIG_XFS_QUOTA) += scrub/quota.o endif diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h index 83121fc..444e286 100644 --- a/fs/xfs/libxfs/xfs_fs.h +++ b/fs/xfs/libxfs/xfs_fs.h @@ -503,7 +503,10 @@ struct xfs_scrub_metadata { #define XFS_SCRUB_TYPE_PARENT 19 /* parent pointers */ #define XFS_SCRUB_TYPE_RTBITMAP 20 /* realtime bitmap */ #define XFS_SCRUB_TYPE_RTSUM 21 /* realtime summary */ -#define XFS_SCRUB_TYPE_MAX 21 +#define XFS_SCRUB_TYPE_UQUOTA 22 /* user quotas */ +#define XFS_SCRUB_TYPE_GQUOTA 23 /* group quotas */ +#define XFS_SCRUB_TYPE_PQUOTA 24 /* project quotas */ +#define XFS_SCRUB_TYPE_MAX 24 /* i: repair this metadata */ #define XFS_SCRUB_FLAG_REPAIR (1 << 0) diff --git a/fs/xfs/scrub/common.c b/fs/xfs/scrub/common.c index 6e40fa6..62884a8 100644 --- a/fs/xfs/scrub/common.c +++ b/fs/xfs/scrub/common.c @@ -875,6 +875,24 @@ static const struct xfs_scrub_meta_fns meta_scrub_fns[] = { { NULL }, { NULL }, #endif +#ifdef CONFIG_XFS_QUOTA + { /* user quota */ + .setup = xfs_scrub_setup_quota, + .scrub = xfs_scrub_quota, + }, + { /* group quota */ + .setup = xfs_scrub_setup_quota, + .scrub = xfs_scrub_quota, + }, + { /* project quota */ + .setup = xfs_scrub_setup_quota, + .scrub = xfs_scrub_quota, + }, +#else + { NULL }, + { NULL }, + { NULL }, +#endif }; /* Dispatch metadata scrubbing. */ diff --git a/fs/xfs/scrub/common.h b/fs/xfs/scrub/common.h index 43a74f0..fcb3764 100644 --- a/fs/xfs/scrub/common.h +++ b/fs/xfs/scrub/common.h @@ -226,6 +226,7 @@ SETUP_FN(xfs_scrub_setup_xattr); SETUP_FN(xfs_scrub_setup_parent); SETUP_FN(xfs_scrub_setup_symlink); SETUP_FN(xfs_scrub_setup_rt); +SETUP_FN(xfs_scrub_setup_quota); #undef SETUP_FN /* Metadata scrubbers */ @@ -253,6 +254,7 @@ SCRUB_FN(xfs_scrub_parent); SCRUB_FN(xfs_scrub_symlink); SCRUB_FN(xfs_scrub_rtbitmap); SCRUB_FN(xfs_scrub_rtsummary); +SCRUB_FN(xfs_scrub_quota); #undef SCRUB_FN #endif /* __XFS_REPAIR_COMMON_H__ */ diff --git a/fs/xfs/scrub/quota.c b/fs/xfs/scrub/quota.c new file mode 100644 index 0000000..117b8b6 --- /dev/null +++ b/fs/xfs/scrub/quota.c @@ -0,0 +1,274 @@ +/* + * Copyright (C) 2017 Oracle. All Rights Reserved. + * + * Author: Darrick J. Wong <darrick.wong@xxxxxxxxxx> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version 2 + * of the License, or (at your option) any later version. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA. + */ +#include "xfs.h" +#include "xfs_fs.h" +#include "xfs_shared.h" +#include "xfs_format.h" +#include "xfs_trans_resv.h" +#include "xfs_mount.h" +#include "xfs_defer.h" +#include "xfs_btree.h" +#include "xfs_bit.h" +#include "xfs_log_format.h" +#include "xfs_trans.h" +#include "xfs_trace.h" +#include "xfs_sb.h" +#include "xfs_inode.h" +#include "xfs_inode_fork.h" +#include "xfs_bmap.h" +#include "xfs_quota.h" +#include "xfs_qm.h" +#include "xfs_dquot.h" +#include "xfs_dquot_item.h" +#include "scrub/common.h" + +/* Convert a scrub type code to a DQ flag, or return 0 if error. */ +static inline uint +xfs_scrub_quota_to_dqtype( + struct xfs_scrub_context *sc) +{ + switch (sc->sm->sm_type) { + case XFS_SCRUB_TYPE_UQUOTA: + return XFS_DQ_USER; + case XFS_SCRUB_TYPE_GQUOTA: + return XFS_DQ_GROUP; + case XFS_SCRUB_TYPE_PQUOTA: + return XFS_DQ_PROJ; + default: + return 0; + } +} + +/* Set us up to scrub a quota. */ +int +xfs_scrub_setup_quota( + struct xfs_scrub_context *sc, + struct xfs_inode *ip) +{ + uint dqtype; + + if (sc->sm->sm_agno || sc->sm->sm_ino || sc->sm->sm_gen) + return -EINVAL; + + dqtype = xfs_scrub_quota_to_dqtype(sc); + if (dqtype == 0) + return -EINVAL; + return 0; +} + +/* Quotas. */ + +#define XFS_SCRUB_QUOTA_CHECK(fs_ok) \ + XFS_SCRUB_DATA_CHECK(sc, XFS_DATA_FORK, id, tag, fs_ok) +#define XFS_SCRUB_QUOTA_WARN(fs_ok) \ + XFS_SCRUB_DATA_WARN(sc, XFS_DATA_FORK, id, tag, fs_ok) +#define XFS_SCRUB_QUOTA_GOTO(fs_ok, label) \ + XFS_SCRUB_DATA_GOTO(sc, XFS_DATA_FORK, id, tag, fs_ok, label) +#define XFS_SCRUB_QUOTA_OP_ERR(label) \ + XFS_SCRUB_FILE_OP_ERROR_GOTO(sc, XFS_DATA_FORK, id, tag, &error, label) +/* Scrub the fields in an individual quota item. */ +STATIC void +xfs_scrub_quota_item( + struct xfs_scrub_context *sc, + const char *tag, + uint dqtype, + struct xfs_dquot *dq, + xfs_dqid_t id) +{ + struct xfs_mount *mp = sc->mp; + struct xfs_disk_dquot *d = &dq->q_core; + unsigned long long bsoft; + unsigned long long isoft; + unsigned long long rsoft; + unsigned long long bhard; + unsigned long long ihard; + unsigned long long rhard; + unsigned long long bcount; + unsigned long long icount; + unsigned long long rcount; + xfs_ino_t inodes; + + /* Did we get the dquot we wanted? */ + XFS_SCRUB_QUOTA_CHECK(id <= be32_to_cpu(d->d_id)); + XFS_SCRUB_QUOTA_CHECK(dqtype == + (d->d_flags & XFS_DQ_ALLTYPES)); + + /* Check the limits. */ + bhard = be64_to_cpu(d->d_blk_hardlimit); + ihard = be64_to_cpu(d->d_ino_hardlimit); + rhard = be64_to_cpu(d->d_rtb_hardlimit); + + bsoft = be64_to_cpu(d->d_blk_softlimit); + isoft = be64_to_cpu(d->d_ino_softlimit); + rsoft = be64_to_cpu(d->d_rtb_softlimit); + + inodes = XFS_AGINO_TO_INO(mp, mp->m_sb.sb_agcount, 0); + + /* + * Warn if the limits are larger than the fs. Administrators + * can do this, though in production this seems suspect. + */ + XFS_SCRUB_QUOTA_WARN(bhard <= mp->m_sb.sb_dblocks); + XFS_SCRUB_QUOTA_WARN(ihard <= inodes); + XFS_SCRUB_QUOTA_WARN(rhard <= mp->m_sb.sb_rblocks); + + XFS_SCRUB_QUOTA_WARN(bsoft <= mp->m_sb.sb_dblocks); + XFS_SCRUB_QUOTA_WARN(isoft <= inodes); + XFS_SCRUB_QUOTA_WARN(rsoft <= mp->m_sb.sb_rblocks); + + /* Soft limit must be less than the hard limit. */ + XFS_SCRUB_QUOTA_CHECK(bsoft <= bhard); + XFS_SCRUB_QUOTA_CHECK(isoft <= ihard); + XFS_SCRUB_QUOTA_CHECK(rsoft <= rhard); + + /* Check the resource counts. */ + bcount = be64_to_cpu(d->d_bcount); + icount = be64_to_cpu(d->d_icount); + rcount = be64_to_cpu(d->d_rtbcount); + inodes = percpu_counter_sum(&mp->m_icount); + + /* + * Check that usage doesn't exceed physical limits. However, on + * a reflink filesystem we're allowed to exceed physical space + * if there are no quota limits. + */ + if (xfs_sb_version_hasreflink(&mp->m_sb)) + XFS_SCRUB_QUOTA_WARN(bcount <= mp->m_sb.sb_dblocks); + else + XFS_SCRUB_QUOTA_CHECK(bcount <= mp->m_sb.sb_dblocks); + XFS_SCRUB_QUOTA_CHECK(icount <= inodes); + XFS_SCRUB_QUOTA_CHECK(rcount <= mp->m_sb.sb_rblocks); + + /* + * We can violate the hard limits if the admin suddenly sets a + * lower limit than the actual usage. However, we flag it for + * admin review. + */ + XFS_SCRUB_QUOTA_WARN(id == 0 || bhard == 0 || bcount <= bhard); + XFS_SCRUB_QUOTA_WARN(id == 0 || ihard == 0 || icount <= ihard); + XFS_SCRUB_QUOTA_WARN(id == 0 || rhard == 0 || rcount <= rhard); +} + +/* Scrub all of a quota type's items. */ +int +xfs_scrub_quota( + struct xfs_scrub_context *sc) +{ + struct xfs_bmbt_irec irec = { 0 }; + struct xfs_mount *mp = sc->mp; + struct xfs_inode *ip; + const char *tag = NULL; + struct xfs_quotainfo *qi = mp->m_quotainfo; + struct xfs_dquot *dq; + xfs_fileoff_t max_dqid_off; + xfs_fileoff_t off = 0; + xfs_dqid_t id = 0; + uint dqtype; + int nimaps; + int error; + + if (!XFS_IS_QUOTA_RUNNING(mp) || !XFS_IS_QUOTA_ON(mp)) + return -ENOENT; + + dqtype = xfs_scrub_quota_to_dqtype(sc); + switch (dqtype) { + case XFS_DQ_USER: + tag = "usrquota"; + break; + case XFS_DQ_GROUP: + tag = "grpquota"; + break; + case XFS_DQ_PROJ: + tag = "prjquota"; + break; + default: + ASSERT(0); + } + + mutex_lock(&qi->qi_quotaofflock); + if (!xfs_this_quota_on(sc->mp, dqtype)) { + error = -ENOENT; + goto out; + } + + /* Attach to the quota inode and set sc->ip so that reporting works. */ + ip = xfs_quota_inode(sc->mp, dqtype); + sc->ip = ip; + + /* Look for problem extents. */ + xfs_ilock(ip, XFS_ILOCK_EXCL); + max_dqid_off = ((xfs_dqid_t)-1) / qi->qi_dqperchunk; + while (1) { + if (xfs_scrub_should_terminate(&error)) + break; + + off = irec.br_startoff + irec.br_blockcount; + nimaps = 1; + error = xfs_bmapi_read(ip, off, -1, &irec, &nimaps, + XFS_BMAPI_ENTIRE); + XFS_SCRUB_QUOTA_OP_ERR(out_unlock); + if (!nimaps) + break; + if (irec.br_startblock == HOLESTARTBLOCK) + continue; + + /* + * Unwritten extents or blocks mapped above the highest + * quota id shouldn't happen. + */ + XFS_SCRUB_QUOTA_GOTO(!isnullstartblock(irec.br_startblock), + next_extent); + XFS_SCRUB_QUOTA_GOTO(irec.br_startoff <= max_dqid_off, + next_extent); + XFS_SCRUB_QUOTA_GOTO(irec.br_startoff + irec.br_blockcount <= + max_dqid_off + 1, next_extent); +next_extent:; + } + xfs_iunlock(ip, XFS_ILOCK_EXCL); + + /* Check all the quota items. */ + while (id < ((xfs_dqid_t)-1ULL)) { + if (xfs_scrub_should_terminate(&error)) + break; + + error = xfs_qm_dqget(mp, NULL, id, dqtype, XFS_QMOPT_DQNEXT, + &dq); + if (error == -ENOENT) + break; + XFS_SCRUB_QUOTA_OP_ERR(out); + + xfs_scrub_quota_item(sc, tag, dqtype, dq, id); + + id = be32_to_cpu(dq->q_core.d_id) + 1; + xfs_qm_dqput(dq); + } + goto out; + +out_unlock: + xfs_iunlock(ip, XFS_ILOCK_EXCL); +out: + sc->ip = NULL; + mutex_unlock(&qi->qi_quotaofflock); + return error; +} +#undef XFS_SCRUB_QUOTA_OP_ERR +#undef XFS_SCRUB_QUOTA_GOTO +#undef XFS_SCRUB_QUOTA_WARN +#undef XFS_SCRUB_QUOTA_CHECK diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 1be7b00..9f71cb9 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -3332,7 +3332,10 @@ DEFINE_GETFSMAP_EVENT(xfs_getfsmap_mapping); { XFS_SCRUB_TYPE_XATTR, "xattr" }, \ { XFS_SCRUB_TYPE_SYMLINK, "symlink" }, \ { XFS_SCRUB_TYPE_RTBITMAP, "rtbitmap" }, \ - { XFS_SCRUB_TYPE_RTSUM, "rtsummary" } + { XFS_SCRUB_TYPE_RTSUM, "rtsummary" }, \ + { XFS_SCRUB_TYPE_UQUOTA, "usrquota" }, \ + { XFS_SCRUB_TYPE_GQUOTA, "grpquota" }, \ + { XFS_SCRUB_TYPE_PQUOTA, "prjquota" } DECLARE_EVENT_CLASS(xfs_scrub_class, TP_PROTO(struct xfs_inode *ip, struct xfs_scrub_metadata *sm, int error), -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html