aboutsummaryrefslogtreecommitdiff
path: root/fs/xfs/xfs_dquot.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/xfs/xfs_dquot.c')
-rw-r--r--fs/xfs/xfs_dquot.c281
1 files changed, 176 insertions, 105 deletions
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index bf27fcca484..3ee0cd43edc 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -17,26 +17,29 @@
*/
#include "xfs.h"
#include "xfs_fs.h"
+#include "xfs_format.h"
+#include "xfs_log_format.h"
+#include "xfs_shared.h"
+#include "xfs_trans_resv.h"
#include "xfs_bit.h"
-#include "xfs_log.h"
-#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
-#include "xfs_alloc.h"
-#include "xfs_quota.h"
#include "xfs_mount.h"
-#include "xfs_bmap_btree.h"
#include "xfs_inode.h"
#include "xfs_bmap.h"
-#include "xfs_rtalloc.h"
+#include "xfs_bmap_util.h"
+#include "xfs_alloc.h"
+#include "xfs_quota.h"
#include "xfs_error.h"
-#include "xfs_itable.h"
-#include "xfs_attr.h"
+#include "xfs_trans.h"
#include "xfs_buf_item.h"
#include "xfs_trans_space.h"
#include "xfs_trans_priv.h"
#include "xfs_qm.h"
+#include "xfs_cksum.h"
#include "xfs_trace.h"
+#include "xfs_log.h"
+#include "xfs_bmap_btree.h"
/*
* Lock order:
@@ -61,7 +64,8 @@ int xfs_dqerror_mod = 33;
struct kmem_zone *xfs_qm_dqtrxzone;
static struct kmem_zone *xfs_qm_dqzone;
-static struct lock_class_key xfs_dquot_other_class;
+static struct lock_class_key xfs_dquot_group_class;
+static struct lock_class_key xfs_dquot_project_class;
/*
* This is called to free all the memory associated with a dquot
@@ -85,17 +89,23 @@ xfs_qm_dqdestroy(
*/
void
xfs_qm_adjust_dqlimits(
- xfs_mount_t *mp,
- xfs_disk_dquot_t *d)
+ struct xfs_mount *mp,
+ struct xfs_dquot *dq)
{
- xfs_quotainfo_t *q = mp->m_quotainfo;
+ struct xfs_quotainfo *q = mp->m_quotainfo;
+ struct xfs_disk_dquot *d = &dq->q_core;
+ int prealloc = 0;
ASSERT(d->d_id);
- if (q->qi_bsoftlimit && !d->d_blk_softlimit)
+ if (q->qi_bsoftlimit && !d->d_blk_softlimit) {
d->d_blk_softlimit = cpu_to_be64(q->qi_bsoftlimit);
- if (q->qi_bhardlimit && !d->d_blk_hardlimit)
+ prealloc = 1;
+ }
+ if (q->qi_bhardlimit && !d->d_blk_hardlimit) {
d->d_blk_hardlimit = cpu_to_be64(q->qi_bhardlimit);
+ prealloc = 1;
+ }
if (q->qi_isoftlimit && !d->d_ino_softlimit)
d->d_ino_softlimit = cpu_to_be64(q->qi_isoftlimit);
if (q->qi_ihardlimit && !d->d_ino_hardlimit)
@@ -104,6 +114,9 @@ xfs_qm_adjust_dqlimits(
d->d_rtb_softlimit = cpu_to_be64(q->qi_rtbsoftlimit);
if (q->qi_rtbhardlimit && !d->d_rtb_hardlimit)
d->d_rtb_hardlimit = cpu_to_be64(q->qi_rtbhardlimit);
+
+ if (prealloc)
+ xfs_dquot_set_prealloc_limits(dq);
}
/*
@@ -239,6 +252,11 @@ xfs_qm_init_dquot_blk(
d->dd_diskdq.d_version = XFS_DQUOT_VERSION;
d->dd_diskdq.d_id = cpu_to_be32(curid);
d->dd_diskdq.d_flags = type;
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ uuid_copy(&d->dd_uuid, &mp->m_sb.sb_uuid);
+ xfs_update_cksum((char *)d, sizeof(struct xfs_dqblk),
+ XFS_DQUOT_CRC_OFF);
+ }
}
xfs_trans_dquot_buf(tp, bp,
@@ -248,7 +266,31 @@ xfs_qm_init_dquot_blk(
xfs_trans_log_buf(tp, bp, 0, BBTOB(q->qi_dqchunklen) - 1);
}
+/*
+ * Initialize the dynamic speculative preallocation thresholds. The lo/hi
+ * watermarks correspond to the soft and hard limits by default. If a soft limit
+ * is not specified, we use 95% of the hard limit.
+ */
+void
+xfs_dquot_set_prealloc_limits(struct xfs_dquot *dqp)
+{
+ __uint64_t space;
+
+ dqp->q_prealloc_hi_wmark = be64_to_cpu(dqp->q_core.d_blk_hardlimit);
+ dqp->q_prealloc_lo_wmark = be64_to_cpu(dqp->q_core.d_blk_softlimit);
+ if (!dqp->q_prealloc_lo_wmark) {
+ dqp->q_prealloc_lo_wmark = dqp->q_prealloc_hi_wmark;
+ do_div(dqp->q_prealloc_lo_wmark, 100);
+ dqp->q_prealloc_lo_wmark *= 95;
+ }
+
+ space = dqp->q_prealloc_hi_wmark;
+ do_div(space, 100);
+ dqp->q_low_space[XFS_QLOWSP_1_PCNT] = space;
+ dqp->q_low_space[XFS_QLOWSP_3_PCNT] = space * 3;
+ dqp->q_low_space[XFS_QLOWSP_5_PCNT] = space * 5;
+}
/*
* Allocate a block and fill it with dquots.
@@ -311,10 +353,11 @@ xfs_qm_dqalloc(
dqp->q_blkno,
mp->m_quotainfo->qi_dqchunklen,
0);
-
- error = xfs_buf_geterror(bp);
- if (error)
+ if (!bp) {
+ error = ENOMEM;
goto error1;
+ }
+ bp->b_ops = &xfs_dquot_buf_ops;
/*
* Make a chunk of dquots out of this buffer and log
@@ -360,6 +403,52 @@ xfs_qm_dqalloc(
return (error);
}
+STATIC int
+xfs_qm_dqrepair(
+ struct xfs_mount *mp,
+ struct xfs_trans *tp,
+ struct xfs_dquot *dqp,
+ xfs_dqid_t firstid,
+ struct xfs_buf **bpp)
+{
+ int error;
+ struct xfs_disk_dquot *ddq;
+ struct xfs_dqblk *d;
+ int i;
+
+ /*
+ * Read the buffer without verification so we get the corrupted
+ * buffer returned to us. make sure we verify it on write, though.
+ */
+ error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, dqp->q_blkno,
+ mp->m_quotainfo->qi_dqchunklen,
+ 0, bpp, NULL);
+
+ if (error) {
+ ASSERT(*bpp == NULL);
+ return XFS_ERROR(error);
+ }
+ (*bpp)->b_ops = &xfs_dquot_buf_ops;
+
+ ASSERT(xfs_buf_islocked(*bpp));
+ d = (struct xfs_dqblk *)(*bpp)->b_addr;
+
+ /* Do the actual repair of dquots in this buffer */
+ for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++) {
+ ddq = &d[i].dd_diskdq;
+ error = xfs_dqcheck(mp, ddq, firstid + i,
+ dqp->dq_flags & XFS_DQ_ALLTYPES,
+ XFS_QMOPT_DQREPAIR, "xfs_qm_dqrepair");
+ if (error) {
+ /* repair failed, we're screwed */
+ xfs_trans_brelse(tp, *bpp);
+ return XFS_ERROR(EIO);
+ }
+ }
+
+ return 0;
+}
+
/*
* Maps a dquot to the buffer containing its on-disk version.
* This returns a ptr to the buffer containing the on-disk dquot
@@ -373,24 +462,24 @@ xfs_qm_dqtobp(
xfs_buf_t **O_bpp,
uint flags)
{
- xfs_bmbt_irec_t map;
- int nmaps = 1, error;
- xfs_buf_t *bp;
- xfs_inode_t *quotip = XFS_DQ_TO_QIP(dqp);
- xfs_mount_t *mp = dqp->q_mount;
- xfs_disk_dquot_t *ddq;
- xfs_dqid_t id = be32_to_cpu(dqp->q_core.d_id);
- xfs_trans_t *tp = (tpp ? *tpp : NULL);
+ struct xfs_bmbt_irec map;
+ int nmaps = 1, error;
+ struct xfs_buf *bp;
+ struct xfs_inode *quotip = xfs_dq_to_quota_inode(dqp);
+ 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;
dqp->q_fileoffset = (xfs_fileoff_t)id / mp->m_quotainfo->qi_dqperchunk;
- xfs_ilock(quotip, XFS_ILOCK_SHARED);
+ 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.
*/
- xfs_iunlock(quotip, XFS_ILOCK_SHARED);
+ xfs_iunlock(quotip, lock_mode);
return ESRCH;
}
@@ -400,7 +489,7 @@ xfs_qm_dqtobp(
error = xfs_bmapi_read(quotip, dqp->q_fileoffset,
XFS_DQUOT_CLUSTER_SIZE_FSB, &map, &nmaps, 0);
- xfs_iunlock(quotip, XFS_ILOCK_SHARED);
+ xfs_iunlock(quotip, lock_mode);
if (error)
return error;
@@ -439,33 +528,24 @@ xfs_qm_dqtobp(
error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp,
dqp->q_blkno,
mp->m_quotainfo->qi_dqchunklen,
- 0, &bp);
- if (error || !bp)
- return XFS_ERROR(error);
- }
-
- ASSERT(xfs_buf_islocked(bp));
+ 0, &bp, &xfs_dquot_buf_ops);
- /*
- * calculate the location of the dquot inside the buffer.
- */
- ddq = bp->b_addr + dqp->q_bufoffset;
+ if (error == EFSCORRUPTED && (flags & XFS_QMOPT_DQREPAIR)) {
+ xfs_dqid_t firstid = (xfs_dqid_t)map.br_startoff *
+ mp->m_quotainfo->qi_dqperchunk;
+ ASSERT(bp == NULL);
+ error = xfs_qm_dqrepair(mp, tp, dqp, firstid, &bp);
+ }
- /*
- * A simple sanity check in case we got a corrupted dquot...
- */
- error = xfs_qm_dqcheck(mp, ddq, id, dqp->dq_flags & XFS_DQ_ALLTYPES,
- flags & (XFS_QMOPT_DQREPAIR|XFS_QMOPT_DOWARN),
- "dqtobp");
- if (error) {
- if (!(flags & XFS_QMOPT_DQREPAIR)) {
- xfs_trans_brelse(tp, bp);
- return XFS_ERROR(EIO);
+ if (error) {
+ ASSERT(bp == NULL);
+ return XFS_ERROR(error);
}
}
+ ASSERT(xfs_buf_islocked(bp));
*O_bpp = bp;
- *O_ddpp = ddq;
+ *O_ddpp = bp->b_addr + dqp->q_bufoffset;
return (0);
}
@@ -514,8 +594,20 @@ xfs_qm_dqread(
* Make sure group quotas have a different lock class than user
* quotas.
*/
- if (!(type & XFS_DQ_USER))
- lockdep_set_class(&dqp->q_qlock, &xfs_dquot_other_class);
+ switch (type) {
+ case XFS_DQ_USER:
+ /* uses the default lock class */
+ break;
+ case XFS_DQ_GROUP:
+ lockdep_set_class(&dqp->q_qlock, &xfs_dquot_group_class);
+ break;
+ case XFS_DQ_PROJ:
+ lockdep_set_class(&dqp->q_qlock, &xfs_dquot_project_class);
+ break;
+ default:
+ ASSERT(0);
+ break;
+ }
XFS_STATS_INC(xs_qm_dquot);
@@ -523,16 +615,8 @@ xfs_qm_dqread(
if (flags & XFS_QMOPT_DQALLOC) {
tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC);
- error = xfs_trans_reserve(tp, XFS_QM_DQALLOC_SPACE_RES(mp),
- XFS_WRITE_LOG_RES(mp) +
- /*
- * Round the chunklen up to the next multiple
- * of 128 (buf log item chunk size)).
- */
- BBTOB(mp->m_quotainfo->qi_dqchunklen) - 1 + 128,
- 0,
- XFS_TRANS_PERM_LOG_RES,
- XFS_WRITE_LOG_COUNT);
+ error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_dqalloc,
+ XFS_QM_DQALLOC_SPACE_RES(mp), 0);
if (error)
goto error1;
cancelflags = XFS_TRANS_RELEASE_LOG_RES;
@@ -566,6 +650,9 @@ xfs_qm_dqread(
dqp->q_res_icount = be64_to_cpu(ddqp->d_icount);
dqp->q_res_rtbcount = be64_to_cpu(ddqp->d_rtbcount);
+ /* initialize the dquot speculative prealloc thresholds */
+ xfs_dquot_set_prealloc_limits(dqp);
+
/* Mark the buf so that this will stay incore a little longer */
xfs_buf_set_ref(bp, XFS_DQUOT_REF);
@@ -620,7 +707,7 @@ xfs_qm_dqget(
xfs_dquot_t **O_dqpp) /* OUT : locked incore dquot */
{
struct xfs_quotainfo *qi = mp->m_quotainfo;
- struct radix_tree_root *tree = XFS_DQUOT_TREE(qi, type);
+ struct radix_tree_root *tree = xfs_dquot_tree(qi, type);
struct xfs_dquot *dqp;
int error;
@@ -745,43 +832,6 @@ restart:
return (0);
}
-
-STATIC void
-xfs_qm_dqput_final(
- struct xfs_dquot *dqp)
-{
- struct xfs_quotainfo *qi = dqp->q_mount->m_quotainfo;
- struct xfs_dquot *gdqp;
-
- trace_xfs_dqput_free(dqp);
-
- mutex_lock(&qi->qi_lru_lock);
- if (list_empty(&dqp->q_lru)) {
- list_add_tail(&dqp->q_lru, &qi->qi_lru_list);
- qi->qi_lru_count++;
- XFS_STATS_INC(xs_qm_dquot_unused);
- }
- mutex_unlock(&qi->qi_lru_lock);
-
- /*
- * If we just added a udquot to the freelist, then we want to release
- * the gdquot reference that it (probably) has. Otherwise it'll keep
- * the gdquot from getting reclaimed.
- */
- gdqp = dqp->q_gdquot;
- if (gdqp) {
- xfs_dqlock(gdqp);
- dqp->q_gdquot = NULL;
- }
- xfs_dqunlock(dqp);
-
- /*
- * If we had a group quota hint, release it now.
- */
- if (gdqp)
- xfs_qm_dqput(gdqp);
-}
-
/*
* Release a reference to the dquot (decrement ref-count) and unlock it.
*
@@ -797,10 +847,14 @@ xfs_qm_dqput(
trace_xfs_dqput(dqp);
- if (--dqp->q_nrefs > 0)
- xfs_dqunlock(dqp);
- else
- xfs_qm_dqput_final(dqp);
+ if (--dqp->q_nrefs == 0) {
+ struct xfs_quotainfo *qi = dqp->q_mount->m_quotainfo;
+ trace_xfs_dqput_free(dqp);
+
+ if (list_lru_add(&qi->qi_lru, &dqp->q_lru))
+ XFS_STATS_INC(xs_qm_dquot_unused);
+ }
+ xfs_dqunlock(dqp);
}
/*
@@ -920,7 +974,7 @@ xfs_qm_dqflush(
* Get the buffer containing the on-disk dquot
*/
error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dqp->q_blkno,
- mp->m_quotainfo->qi_dqchunklen, 0, &bp);
+ mp->m_quotainfo->qi_dqchunklen, 0, &bp, NULL);
if (error)
goto out_unlock;
@@ -932,7 +986,7 @@ xfs_qm_dqflush(
/*
* A simple sanity check in case we got a corrupted dquot..
*/
- error = xfs_qm_dqcheck(mp, &dqp->q_core, be32_to_cpu(ddqp->d_id), 0,
+ error = xfs_dqcheck(mp, &dqp->q_core, be32_to_cpu(ddqp->d_id), 0,
XFS_QMOPT_DOWARN, "dqflush (incore copy)");
if (error) {
xfs_buf_relse(bp);
@@ -953,6 +1007,23 @@ xfs_qm_dqflush(
&dqp->q_logitem.qli_item.li_lsn);
/*
+ * copy the lsn into the on-disk dquot now while we have the in memory
+ * dquot here. This can't be done later in the write verifier as we
+ * can't get access to the log item at that point in time.
+ *
+ * We also calculate the CRC here so that the on-disk dquot in the
+ * buffer always has a valid CRC. This ensures there is no possibility
+ * of a dquot without an up-to-date CRC getting to disk.
+ */
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dqblk *dqb = (struct xfs_dqblk *)ddqp;
+
+ dqb->dd_lsn = cpu_to_be64(dqp->q_logitem.qli_item.li_lsn);
+ xfs_update_cksum((char *)dqb, sizeof(struct xfs_dqblk),
+ XFS_DQUOT_CRC_OFF);
+ }
+
+ /*
* Attach an iodone routine so that we can remove this dquot from the
* AIL and release the flush lock once the dquot is synced to disk.
*/
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-11-10 06:49:01 +0000