From: Darrick J. Wong <djwong@xxxxxxxxxx>
Use the collected reference count information to rebuild the btree.
Signed-off-by: "Darrick J. Wong" <djwong@xxxxxxxxxx>
---
libxfs/libxfs_api_defs.h | 5 +
repair/Makefile | 1
repair/agbtree.c | 2
repair/phase5.c | 6 +
repair/phase6.c | 14 ++
repair/rmap.c | 22 ++++
repair/rmap.h | 7 +
repair/rtrefcount_repair.c | 257 ++++++++++++++++++++++++++++++++++++++++++++
8 files changed, 308 insertions(+), 6 deletions(-)
create mode 100644 repair/rtrefcount_repair.c
diff --git a/libxfs/libxfs_api_defs.h b/libxfs/libxfs_api_defs.h
index b5a39856bc1e80..66cbb34f05a48f 100644
--- a/libxfs/libxfs_api_defs.h
+++ b/libxfs/libxfs_api_defs.h
@@ -320,11 +320,16 @@
#define xfs_rtgroup_get libxfs_rtgroup_get
#define xfs_rtgroup_put libxfs_rtgroup_put
+#define xfs_rtrefcountbt_absolute_maxlevels libxfs_rtrefcountbt_absolute_maxlevels
#define xfs_rtrefcountbt_calc_reserves libxfs_rtrefcountbt_calc_reserves
+#define xfs_rtrefcountbt_calc_size libxfs_rtrefcountbt_calc_size
+#define xfs_rtrefcountbt_commit_staged_btree libxfs_rtrefcountbt_commit_staged_btree
+#define xfs_rtrefcountbt_create libxfs_rtrefcountbt_create
#define xfs_rtrefcountbt_droot_maxrecs libxfs_rtrefcountbt_droot_maxrecs
#define xfs_rtrefcountbt_init_cursor libxfs_rtrefcountbt_init_cursor
#define xfs_rtrefcountbt_maxlevels_ondisk libxfs_rtrefcountbt_maxlevels_ondisk
#define xfs_rtrefcountbt_maxrecs libxfs_rtrefcountbt_maxrecs
+#define xfs_rtrefcountbt_stage_cursor libxfs_rtrefcountbt_stage_cursor
#define xfs_rtrmapbt_calc_reserves libxfs_rtrmapbt_calc_reserves
#define xfs_rtrmapbt_calc_size libxfs_rtrmapbt_calc_size
diff --git a/repair/Makefile b/repair/Makefile
index 6f4ec3b3a9c4dc..ff5b1f5abedac0 100644
--- a/repair/Makefile
+++ b/repair/Makefile
@@ -73,6 +73,7 @@ CFILES = \
rcbag.c \
rmap.c \
rt.c \
+ rtrefcount_repair.c \
rtrmap_repair.c \
sb.c \
scan.c \
diff --git a/repair/agbtree.c b/repair/agbtree.c
index 01066130767cb6..983b645e1a35a3 100644
--- a/repair/agbtree.c
+++ b/repair/agbtree.c
@@ -730,7 +730,7 @@ init_refc_cursor(
/* Compute how many blocks we'll need. */
error = -libxfs_btree_bload_compute_geometry(btr->cur, &btr->bload,
- refcount_record_count(sc->mp, agno));
+ refcount_record_count(sc->mp, false, agno));
if (error)
do_error(
_("Unable to compute refcount btree geometry, error %d.\n"), error);
diff --git a/repair/phase5.c b/repair/phase5.c
index cacaf74dda3a60..4cf28d8ae1a250 100644
--- a/repair/phase5.c
+++ b/repair/phase5.c
@@ -705,7 +705,7 @@ are_packed_btrees_needed(
* If we don't have inode-based metadata, we can let the AG btrees
* pack as needed; there are no global space concerns here.
*/
- if (!xfs_has_rtrmapbt(mp))
+ if (!xfs_has_rtrmapbt(mp) && !xfs_has_rtreflink(mp))
return false;
while ((pag = xfs_perag_next(mp, pag))) {
@@ -716,8 +716,10 @@ are_packed_btrees_needed(
fdblocks += ag_fdblocks;
}
- while ((rtg = xfs_rtgroup_next(mp, rtg)))
+ while ((rtg = xfs_rtgroup_next(mp, rtg))) {
metadata_blocks += estimate_rtrmapbt_blocks(rtg);
+ metadata_blocks += estimate_rtrefcountbt_blocks(rtg);
+ }
/*
* If we think we'll have more metadata blocks than free space, then
diff --git a/repair/phase6.c b/repair/phase6.c
index 30ea19fda9fd87..4064a84b24509f 100644
--- a/repair/phase6.c
+++ b/repair/phase6.c
@@ -696,6 +696,15 @@ ensure_rtgroup_rmapbt(
populate_rtgroup_rmapbt(rtg, est_fdblocks);
}
+static void
+ensure_rtgroup_refcountbt(
+ struct xfs_rtgroup *rtg,
+ xfs_filblks_t est_fdblocks)
+{
+ if (ensure_rtgroup_file(rtg, XFS_RTGI_REFCOUNT))
+ populate_rtgroup_refcountbt(rtg, est_fdblocks);
+}
+
/* Initialize a root directory. */
static int
init_fs_root_dir(
@@ -3405,8 +3414,10 @@ reset_rt_metadir_inodes(
* maximally.
*/
if (!need_packed_btrees) {
- while ((rtg = xfs_rtgroup_next(mp, rtg)))
+ while ((rtg = xfs_rtgroup_next(mp, rtg))) {
metadata_blocks += estimate_rtrmapbt_blocks(rtg);
+ metadata_blocks += estimate_rtrefcountbt_blocks(rtg);
+ }
if (mp->m_sb.sb_fdblocks > metadata_blocks)
est_fdblocks = mp->m_sb.sb_fdblocks - metadata_blocks;
@@ -3427,6 +3438,7 @@ _(" - resetting contents of realtime bitmap and summary inodes\n"));
ensure_rtgroup_bitmap(rtg);
ensure_rtgroup_summary(rtg);
ensure_rtgroup_rmapbt(rtg, est_fdblocks);
+ ensure_rtgroup_refcountbt(rtg, est_fdblocks);
}
}
diff --git a/repair/rmap.c b/repair/rmap.c
index 638e5ea92278cb..97510dd875911a 100644
--- a/repair/rmap.c
+++ b/repair/rmap.c
@@ -1682,9 +1682,10 @@ _("Unable to fix reflink flag on inode %"PRIu64".\n"),
uint64_t
refcount_record_count(
struct xfs_mount *mp,
+ bool isrt,
xfs_agnumber_t agno)
{
- struct xfs_ag_rmap *x = rmaps_for_group(false, agno);
+ struct xfs_ag_rmap *x = rmaps_for_group(isrt, agno);
return slab_count(x->ar_refcount_items);
}
@@ -2081,3 +2082,22 @@ estimate_rtrmapbt_blocks(
nr_recs = xmbuf_bytes(x->ar_xmbtp) / sizeof(struct xfs_rmap_rec);
return libxfs_rtrmapbt_calc_size(mp, nr_recs);
}
+
+/* Estimate the size of the ondisk rtrefcountbt from the incore data. */
+xfs_filblks_t
+estimate_rtrefcountbt_blocks(
+ struct xfs_rtgroup *rtg)
+{
+ struct xfs_mount *mp = rtg_mount(rtg);
+ struct xfs_ag_rmap *x;
+
+ if (!rmap_needs_work(mp) || !xfs_has_rtreflink(mp))
+ return 0;
+
+ x = &rg_rmaps[rtg_rgno(rtg)];
+ if (!x->ar_refcount_items)
+ return 0;
+
+ return libxfs_rtrefcountbt_calc_size(mp,
+ slab_count(x->ar_refcount_items));
+}
diff --git a/repair/rmap.h b/repair/rmap.h
index 80e82a4ac4c008..1f234b8be32e72 100644
--- a/repair/rmap.h
+++ b/repair/rmap.h
@@ -39,7 +39,8 @@ extern void rmap_high_key_from_rec(struct xfs_rmap_irec *rec,
struct xfs_rmap_irec *key);
int compute_refcounts(struct xfs_mount *mp, bool isrt, xfs_agnumber_t agno);
-uint64_t refcount_record_count(struct xfs_mount *mp, xfs_agnumber_t agno);
+uint64_t refcount_record_count(struct xfs_mount *mp, bool isrt,
+ xfs_agnumber_t agno);
int init_refcount_cursor(bool isrt, xfs_agnumber_t agno,
struct xfs_slab_cursor **pcur);
extern void refcount_avoid_check(struct xfs_mount *mp);
@@ -64,4 +65,8 @@ void populate_rtgroup_rmapbt(struct xfs_rtgroup *rtg,
xfs_filblks_t est_fdblocks);
xfs_filblks_t estimate_rtrmapbt_blocks(struct xfs_rtgroup *rtg);
+int populate_rtgroup_refcountbt(struct xfs_rtgroup *rtg,
+ xfs_filblks_t est_fdblocks);
+xfs_filblks_t estimate_rtrefcountbt_blocks(struct xfs_rtgroup *rtg);
+
#endif /* RMAP_H_ */
diff --git a/repair/rtrefcount_repair.c b/repair/rtrefcount_repair.c
new file mode 100644
index 00000000000000..228d080e5a5dcb
--- /dev/null
+++ b/repair/rtrefcount_repair.c
@@ -0,0 +1,257 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2021-2025 Oracle. All Rights Reserved.
+ * Author: Darrick J. Wong <djwong@xxxxxxxxxx>
+ */
+#include <libxfs.h>
+#include "btree.h"
+#include "err_protos.h"
+#include "libxlog.h"
+#include "incore.h"
+#include "globals.h"
+#include "dinode.h"
+#include "slab.h"
+#include "rmap.h"
+#include "bulkload.h"
+
+/*
+ * Realtime Reference Count (RTREFCBT) Repair
+ * ==========================================
+ *
+ * Gather all the reference count records for the realtime device, reset the
+ * incore fork, then recreate the btree.
+ */
+struct xrep_rtrefc {
+ /* rtrefcbt slab cursor */
+ struct xfs_slab_cursor *slab_cursor;
+
+ /* New fork. */
+ struct bulkload new_fork_info;
+ struct xfs_btree_bload rtrefc_bload;
+
+ struct repair_ctx *sc;
+ struct xfs_rtgroup *rtg;
+
+ /* Estimated free space after building all rt btrees */
+ xfs_filblks_t est_fdblocks;
+};
+
+/* Retrieve rtrefc data for bulk load. */
+STATIC int
+xrep_rtrefc_get_records(
+ struct xfs_btree_cur *cur,
+ unsigned int idx,
+ struct xfs_btree_block *block,
+ unsigned int nr_wanted,
+ void *priv)
+{
+ struct xfs_refcount_irec *rec;
+ struct xrep_rtrefc *rc = priv;
+ union xfs_btree_rec *block_rec;
+ unsigned int loaded;
+
+ for (loaded = 0; loaded < nr_wanted; loaded++, idx++) {
+ rec = pop_slab_cursor(rc->slab_cursor);
+ memcpy(&cur->bc_rec.rc, rec, sizeof(struct xfs_refcount_irec));
+
+ block_rec = libxfs_btree_rec_addr(cur, idx, block);
+ cur->bc_ops->init_rec_from_cur(cur, block_rec);
+ }
+
+ return loaded;
+}
+
+/* Feed one of the new btree blocks to the bulk loader. */
+STATIC int
+xrep_rtrefc_claim_block(
+ struct xfs_btree_cur *cur,
+ union xfs_btree_ptr *ptr,
+ void *priv)
+{
+ struct xrep_rtrefc *rr = priv;
+
+ return bulkload_claim_block(cur, &rr->new_fork_info, ptr);
+}
+
+/* Figure out how much space we need to create the incore btree root block. */
+STATIC size_t
+xrep_rtrefc_iroot_size(
+ struct xfs_btree_cur *cur,
+ unsigned int level,
+ unsigned int nr_this_level,
+ void *priv)
+{
+ return xfs_rtrefcount_broot_space_calc(cur->bc_mp, level,
+ nr_this_level);
+}
+
+/* Reserve new btree blocks and bulk load all the rtrmap records. */
+STATIC int
+xrep_rtrefc_btree_load(
+ struct xrep_rtrefc *rr,
+ struct xfs_btree_cur *rtrmap_cur)
+{
+ struct repair_ctx *sc = rr->sc;
+ int error;
+
+ rr->rtrefc_bload.get_records = xrep_rtrefc_get_records;
+ rr->rtrefc_bload.claim_block = xrep_rtrefc_claim_block;
+ rr->rtrefc_bload.iroot_size = xrep_rtrefc_iroot_size;
+ bulkload_estimate_inode_slack(sc->mp, &rr->rtrefc_bload,
+ rr->est_fdblocks);
+
+ /* Compute how many blocks we'll need. */
+ error = -libxfs_btree_bload_compute_geometry(rtrmap_cur,
+ &rr->rtrefc_bload,
+ refcount_record_count(sc->mp, true, rtg_rgno(rr->rtg)));
+ if (error)
+ return error;
+
+ /*
+ * Guess how many blocks we're going to need to rebuild an entire
+ * rtrefcountbt from the number of extents we found, and pump up our
+ * transaction to have sufficient block reservation.
+ */
+ error = -libxfs_trans_reserve_more(sc->tp, rr->rtrefc_bload.nr_blocks,
+ 0);
+ if (error)
+ return error;
+
+ /*
+ * Reserve the space we'll need for the new btree. Drop the cursor
+ * while we do this because that can roll the transaction and cursors
+ * can't handle that.
+ */
+ error = bulkload_alloc_file_blocks(&rr->new_fork_info,
+ rr->rtrefc_bload.nr_blocks);
+ if (error)
+ return error;
+
+ /* Add all observed rtrmap records. */
+ error = init_refcount_cursor(true, rtg_rgno(rr->rtg), &rr->slab_cursor);
+ if (error)
+ return error;
+ error = -libxfs_btree_bload(rtrmap_cur, &rr->rtrefc_bload, rr);
+ free_slab_cursor(&rr->slab_cursor);
+ return error;
+}
+
+/* Update the inode counters. */
+STATIC int
+xrep_rtrefc_reset_counters(
+ struct xrep_rtrefc *rr)
+{
+ struct repair_ctx *sc = rr->sc;
+
+ /*
+ * Update the inode block counts to reflect the btree we just
+ * generated.
+ */
+ sc->ip->i_nblocks = rr->new_fork_info.ifake.if_blocks;
+ libxfs_trans_log_inode(sc->tp, sc->ip, XFS_ILOG_CORE);
+
+ /* Quotas don't exist so we're done. */
+ return 0;
+}
+
+/*
+ * Use the collected rmap information to stage a new rt refcount btree. If
+ * this is successful we'll return with the new btree root information logged
+ * to the repair transaction but not yet committed.
+ */
+static int
+xrep_rtrefc_build_new_tree(
+ struct xrep_rtrefc *rr)
+{
+ struct xfs_owner_info oinfo;
+ struct xfs_btree_cur *cur;
+ struct repair_ctx *sc = rr->sc;
+ struct xbtree_ifakeroot *ifake = &rr->new_fork_info.ifake;
+ int error;
+
+ /*
+ * Prepare to construct the new fork by initializing the new btree
+ * structure and creating a fake ifork in the ifakeroot structure.
+ */
+ libxfs_rmap_ino_bmbt_owner(&oinfo, sc->ip->i_ino, XFS_DATA_FORK);
+ bulkload_init_inode(&rr->new_fork_info, sc, XFS_DATA_FORK, &oinfo);
+ cur = libxfs_rtrefcountbt_init_cursor(NULL, rr->rtg);
+ libxfs_btree_stage_ifakeroot(cur, ifake);
+
+ /*
+ * Figure out the size and format of the new fork, then fill it with
+ * all the rtrmap records we've found. Join the inode to the
+ * transaction so that we can roll the transaction while holding the
+ * inode locked.
+ */
+ libxfs_trans_ijoin(sc->tp, sc->ip, 0);
+ ifake->if_fork->if_format = XFS_DINODE_FMT_META_BTREE;
+ error = xrep_rtrefc_btree_load(rr, cur);
+ if (error)
+ goto err_cur;
+
+ /*
+ * Install the new fork in the inode. After this point the old mapping
+ * data are no longer accessible and the new tree is live. We delete
+ * the cursor immediately after committing the staged root because the
+ * staged fork might be in extents format.
+ */
+ libxfs_rtrefcountbt_commit_staged_btree(cur, sc->tp);
+ libxfs_btree_del_cursor(cur, 0);
+
+ /* Reset the inode counters now that we've changed the fork. */
+ error = xrep_rtrefc_reset_counters(rr);
+ if (error)
+ goto err_newbt;
+
+ /* Dispose of any unused blocks and the accounting infomation. */
+ error = bulkload_commit(&rr->new_fork_info);
+ if (error)
+ return error;
+
+ return -libxfs_trans_roll_inode(&sc->tp, sc->ip);
+err_cur:
+ if (cur)
+ libxfs_btree_del_cursor(cur, error);
+err_newbt:
+ bulkload_cancel(&rr->new_fork_info);
+ return error;
+}
+
+/* Store the realtime reference counts in the rtrefcbt. */
+int
+populate_rtgroup_refcountbt(
+ struct xfs_rtgroup *rtg,
+ xfs_filblks_t est_fdblocks)
+{
+ struct xfs_mount *mp = rtg_mount(rtg);
+ struct xfs_inode *ip = rtg_refcount(rtg);
+ struct repair_ctx sc = {
+ .mp = mp,
+ .ip = ip,
+ };
+ struct xrep_rtrefc rr = {
+ .sc = &sc,
+ .rtg = rtg,
+ .est_fdblocks = est_fdblocks,
+ };
+ int error;
+
+ if (!xfs_has_rtreflink(mp))
+ return 0;
+
+ error = -libxfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0,
+ &sc.tp);
+ if (error)
+ return error;
+
+ error = xrep_rtrefc_build_new_tree(&rr);
+ if (error)
+ goto out_cancel;
+
+ return -libxfs_trans_commit(sc.tp);
+
+out_cancel:
+ libxfs_trans_cancel(sc.tp);
+ return error;
+}
