X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libbcachefs%2Frecovery.c;h=9127d0e3ca2f6a3fd44e076b42f01ee6f7736427;hb=f3f005c76eb5636542a8f5b137bd1904d57e8f86;hp=98f1454c23fb6c6307682082c27ecf072d392798;hpb=89b361f24a433a4a55e0032eca4e43045ea3f0d0;p=bcachefs-tools-debian

diff --git a/libbcachefs/recovery.c b/libbcachefs/recovery.c
index 98f1454..9127d0e 100644
--- a/libbcachefs/recovery.c
+++ b/libbcachefs/recovery.c
@@ -27,6 +27,7 @@
 #include "recovery.h"
 #include "replicas.h"
 #include "sb-clean.h"
+#include "sb-downgrade.h"
 #include "snapshot.h"
 #include "subvolume.h"
 #include "super-io.h"
@@ -145,12 +146,11 @@ static int bch2_journal_replay(struct bch_fs *c)
 {
 	struct journal_keys *keys = &c->journal_keys;
 	DARRAY(struct journal_key *) keys_sorted = { 0 };
-	struct journal_key **kp;
 	struct journal *j = &c->journal;
 	u64 start_seq	= c->journal_replay_seq_start;
 	u64 end_seq	= c->journal_replay_seq_start;
 	struct btree_trans *trans = bch2_trans_get(c);
-	int ret;
+	int ret = 0;
 
 	if (keys->nr) {
 		ret = bch2_journal_log_msg(c, "Starting journal replay (%zu keys in entries %llu-%llu)",
@@ -171,10 +171,12 @@ static int bch2_journal_replay(struct bch_fs *c)
 
 		struct journal_key *k = keys->d + i;
 
-		ret = commit_do(trans, NULL, NULL,
-				BCH_TRANS_COMMIT_no_enospc|
-				BCH_TRANS_COMMIT_journal_reclaim|
-				(!k->allocated ? BCH_TRANS_COMMIT_no_journal_res : 0),
+		/* Skip fastpath if we're low on space in the journal */
+		ret = c->journal.watermark ? -1 :
+			commit_do(trans, NULL, NULL,
+				  BCH_TRANS_COMMIT_no_enospc|
+				  BCH_TRANS_COMMIT_journal_reclaim|
+				  (!k->allocated ? BCH_TRANS_COMMIT_no_journal_res : 0),
 			     bch2_journal_replay_key(trans, k));
 		BUG_ON(!ret && !k->overwritten);
 		if (ret) {
@@ -278,7 +280,7 @@ static int journal_replay_entry_early(struct bch_fs *c,
 					le64_to_cpu(u->v);
 			break;
 		case BCH_FS_USAGE_inodes:
-			c->usage_base->nr_inodes = le64_to_cpu(u->v);
+			c->usage_base->b.nr_inodes = le64_to_cpu(u->v);
 			break;
 		case BCH_FS_USAGE_key_version:
 			atomic64_set(&c->key_version,
@@ -342,14 +344,11 @@ static int journal_replay_entry_early(struct bch_fs *c,
 static int journal_replay_early(struct bch_fs *c,
 				struct bch_sb_field_clean *clean)
 {
-	struct jset_entry *entry;
-	int ret;
-
 	if (clean) {
-		for (entry = clean->start;
+		for (struct jset_entry *entry = clean->start;
 		     entry != vstruct_end(&clean->field);
 		     entry = vstruct_next(entry)) {
-			ret = journal_replay_entry_early(c, entry);
+			int ret = journal_replay_entry_early(c, entry);
 			if (ret)
 				return ret;
 		}
@@ -364,7 +363,7 @@ static int journal_replay_early(struct bch_fs *c,
 				continue;
 
 			vstruct_for_each(&i->j, entry) {
-				ret = journal_replay_entry_early(c, entry);
+				int ret = journal_replay_entry_early(c, entry);
 				if (ret)
 					return ret;
 			}
@@ -460,8 +459,7 @@ static int bch2_initialize_subvolumes(struct bch_fs *c)
 	ret =   bch2_btree_insert(c, BTREE_ID_snapshot_trees,	&root_tree.k_i, NULL, 0) ?:
 		bch2_btree_insert(c, BTREE_ID_snapshots,	&root_snapshot.k_i, NULL, 0) ?:
 		bch2_btree_insert(c, BTREE_ID_subvolumes,	&root_volume.k_i, NULL, 0);
-	if (ret)
-		bch_err_fn(c, ret);
+	bch_err_fn(c, ret);
 	return ret;
 }
 
@@ -501,13 +499,12 @@ static int bch2_fs_upgrade_for_subvolumes(struct bch_fs *c)
 {
 	int ret = bch2_trans_do(c, NULL, NULL, BCH_TRANS_COMMIT_lazy_rw,
 				__bch2_fs_upgrade_for_subvolumes(trans));
-	if (ret)
-		bch_err_fn(c, ret);
+	bch_err_fn(c, ret);
 	return ret;
 }
 
 const char * const bch2_recovery_passes[] = {
-#define x(_fn, _when)	#_fn,
+#define x(_fn, ...)	#_fn,
 	BCH_RECOVERY_PASSES()
 #undef x
 	NULL
@@ -531,7 +528,8 @@ static int bch2_set_may_go_rw(struct bch_fs *c)
 	keys->gap = keys->nr;
 
 	set_bit(BCH_FS_may_go_rw, &c->flags);
-	if (keys->nr || c->opts.fsck)
+
+	if (keys->nr || c->opts.fsck || !c->sb.clean)
 		return bch2_fs_read_write_early(c);
 	return 0;
 }
@@ -542,18 +540,47 @@ struct recovery_pass_fn {
 };
 
 static struct recovery_pass_fn recovery_pass_fns[] = {
-#define x(_fn, _when)	{ .fn = bch2_##_fn, .when = _when },
+#define x(_fn, _id, _when)	{ .fn = bch2_##_fn, .when = _when },
 	BCH_RECOVERY_PASSES()
 #undef x
 };
 
-static void check_version_upgrade(struct bch_fs *c)
+u64 bch2_recovery_passes_to_stable(u64 v)
+{
+	static const u8 map[] = {
+#define x(n, id, ...)	[BCH_RECOVERY_PASS_##n] = BCH_RECOVERY_PASS_STABLE_##n,
+	BCH_RECOVERY_PASSES()
+#undef x
+	};
+
+	u64 ret = 0;
+	for (unsigned i = 0; i < ARRAY_SIZE(map); i++)
+		if (v & BIT_ULL(i))
+			ret |= BIT_ULL(map[i]);
+	return ret;
+}
+
+u64 bch2_recovery_passes_from_stable(u64 v)
+{
+	static const u8 map[] = {
+#define x(n, id, ...)	[BCH_RECOVERY_PASS_STABLE_##n] = BCH_RECOVERY_PASS_##n,
+	BCH_RECOVERY_PASSES()
+#undef x
+	};
+
+	u64 ret = 0;
+	for (unsigned i = 0; i < ARRAY_SIZE(map); i++)
+		if (v & BIT_ULL(i))
+			ret |= BIT_ULL(map[i]);
+	return ret;
+}
+
+static bool check_version_upgrade(struct bch_fs *c)
 {
 	unsigned latest_compatible = bch2_latest_compatible_version(c->sb.version);
 	unsigned latest_version	= bcachefs_metadata_version_current;
 	unsigned old_version = c->sb.version_upgrade_complete ?: c->sb.version;
 	unsigned new_version = 0;
-	u64 recovery_passes;
 
 	if (old_version < bcachefs_metadata_required_upgrade_below) {
 		if (c->opts.version_upgrade == BCH_VERSION_UPGRADE_incompatible ||
@@ -597,27 +624,26 @@ static void check_version_upgrade(struct bch_fs *c)
 		bch2_version_to_text(&buf, new_version);
 		prt_newline(&buf);
 
-		recovery_passes = bch2_upgrade_recovery_passes(c, old_version, new_version);
-		if (recovery_passes) {
-			if ((recovery_passes & RECOVERY_PASS_ALL_FSCK) == RECOVERY_PASS_ALL_FSCK)
-				prt_str(&buf, "fsck required");
-			else {
-				prt_str(&buf, "running recovery passes: ");
-				prt_bitflags(&buf, bch2_recovery_passes, recovery_passes);
-			}
+		struct bch_sb_field_ext *ext = bch2_sb_field_get(c->disk_sb.sb, ext);
+		__le64 passes = ext->recovery_passes_required[0];
+		bch2_sb_set_upgrade(c, old_version, new_version);
+		passes = ext->recovery_passes_required[0] & ~passes;
 
-			c->recovery_passes_explicit |= recovery_passes;
-			c->opts.fix_errors = FSCK_FIX_yes;
+		if (passes) {
+			prt_str(&buf, "  running recovery passes: ");
+			prt_bitflags(&buf, bch2_recovery_passes,
+				     bch2_recovery_passes_from_stable(le64_to_cpu(passes)));
 		}
 
 		bch_info(c, "%s", buf.buf);
 
-		mutex_lock(&c->sb_lock);
 		bch2_sb_upgrade(c, new_version);
-		mutex_unlock(&c->sb_lock);
 
 		printbuf_exit(&buf);
+		return true;
 	}
+
+	return false;
 }
 
 u64 bch2_fsck_recovery_passes(void)
@@ -632,7 +658,7 @@ u64 bch2_fsck_recovery_passes(void)
 
 static bool should_run_recovery_pass(struct bch_fs *c, enum bch_recovery_pass pass)
 {
-	struct recovery_pass_fn *p = recovery_pass_fns + c->curr_recovery_pass;
+	struct recovery_pass_fn *p = recovery_pass_fns + pass;
 
 	if (c->opts.norecovery && pass > BCH_RECOVERY_PASS_snapshots_read)
 		return false;
@@ -649,39 +675,62 @@ static bool should_run_recovery_pass(struct bch_fs *c, enum bch_recovery_pass pa
 
 static int bch2_run_recovery_pass(struct bch_fs *c, enum bch_recovery_pass pass)
 {
+	struct recovery_pass_fn *p = recovery_pass_fns + pass;
 	int ret;
 
-	c->curr_recovery_pass = pass;
+	if (!(p->when & PASS_SILENT))
+		bch2_print(c, KERN_INFO bch2_log_msg(c, "%s..."),
+			   bch2_recovery_passes[pass]);
+	ret = p->fn(c);
+	if (ret)
+		return ret;
+	if (!(p->when & PASS_SILENT))
+		bch2_print(c, KERN_CONT " done\n");
 
-	if (should_run_recovery_pass(c, pass)) {
-		struct recovery_pass_fn *p = recovery_pass_fns + pass;
+	return 0;
+}
 
-		if (!(p->when & PASS_SILENT))
-			printk(KERN_INFO bch2_log_msg(c, "%s..."),
-			       bch2_recovery_passes[pass]);
-		ret = p->fn(c);
-		if (ret)
-			return ret;
-		if (!(p->when & PASS_SILENT))
-			printk(KERN_CONT " done\n");
+static int bch2_run_recovery_passes(struct bch_fs *c)
+{
+	int ret = 0;
 
-		c->recovery_passes_complete |= BIT_ULL(pass);
+	while (c->curr_recovery_pass < ARRAY_SIZE(recovery_pass_fns)) {
+		if (should_run_recovery_pass(c, c->curr_recovery_pass)) {
+			unsigned pass = c->curr_recovery_pass;
+
+			ret = bch2_run_recovery_pass(c, c->curr_recovery_pass);
+			if (bch2_err_matches(ret, BCH_ERR_restart_recovery) ||
+			    (ret && c->curr_recovery_pass < pass))
+				continue;
+			if (ret)
+				break;
+
+			c->recovery_passes_complete |= BIT_ULL(c->curr_recovery_pass);
+		}
+		c->curr_recovery_pass++;
+		c->recovery_pass_done = max(c->recovery_pass_done, c->curr_recovery_pass);
 	}
 
-	return 0;
+	return ret;
 }
 
-static int bch2_run_recovery_passes(struct bch_fs *c)
+int bch2_run_online_recovery_passes(struct bch_fs *c)
 {
 	int ret = 0;
 
-	while (c->curr_recovery_pass < ARRAY_SIZE(recovery_pass_fns)) {
-		ret = bch2_run_recovery_pass(c, c->curr_recovery_pass);
-		if (bch2_err_matches(ret, BCH_ERR_restart_recovery))
+	for (unsigned i = 0; i < ARRAY_SIZE(recovery_pass_fns); i++) {
+		struct recovery_pass_fn *p = recovery_pass_fns + i;
+
+		if (!(p->when & PASS_ONLINE))
 			continue;
+
+		ret = bch2_run_recovery_pass(c, i);
+		if (bch2_err_matches(ret, BCH_ERR_restart_recovery)) {
+			i = c->curr_recovery_pass;
+			continue;
+		}
 		if (ret)
 			break;
-		c->curr_recovery_pass++;
 	}
 
 	return ret;
@@ -692,7 +741,6 @@ int bch2_fs_recovery(struct bch_fs *c)
 	struct bch_sb_field_clean *clean = NULL;
 	struct jset *last_journal_entry = NULL;
 	u64 last_seq = 0, blacklist_seq, journal_seq;
-	bool write_sb = false;
 	int ret = 0;
 
 	if (c->sb.clean) {
@@ -720,15 +768,76 @@ int bch2_fs_recovery(struct bch_fs *c)
 		goto err;
 	}
 
-	if (c->opts.fsck || !(c->opts.nochanges && c->opts.norecovery))
-		check_version_upgrade(c);
-
 	if (c->opts.fsck && c->opts.norecovery) {
 		bch_err(c, "cannot select both norecovery and fsck");
 		ret = -EINVAL;
 		goto err;
 	}
 
+	if (!(c->opts.nochanges && c->opts.norecovery)) {
+		mutex_lock(&c->sb_lock);
+		bool write_sb = false;
+
+		struct bch_sb_field_ext *ext =
+			bch2_sb_field_get_minsize(&c->disk_sb, ext, sizeof(*ext) / sizeof(u64));
+		if (!ext) {
+			ret = -BCH_ERR_ENOSPC_sb;
+			mutex_unlock(&c->sb_lock);
+			goto err;
+		}
+
+		if (BCH_SB_HAS_TOPOLOGY_ERRORS(c->disk_sb.sb)) {
+			ext->recovery_passes_required[0] |=
+				cpu_to_le64(bch2_recovery_passes_to_stable(BIT_ULL(BCH_RECOVERY_PASS_check_topology)));
+			write_sb = true;
+		}
+
+		u64 sb_passes = bch2_recovery_passes_from_stable(le64_to_cpu(ext->recovery_passes_required[0]));
+		if (sb_passes) {
+			struct printbuf buf = PRINTBUF;
+			prt_str(&buf, "superblock requires following recovery passes to be run:\n  ");
+			prt_bitflags(&buf, bch2_recovery_passes, sb_passes);
+			bch_info(c, "%s", buf.buf);
+			printbuf_exit(&buf);
+		}
+
+		if (bch2_check_version_downgrade(c)) {
+			struct printbuf buf = PRINTBUF;
+
+			prt_str(&buf, "Version downgrade required:\n");
+
+			__le64 passes = ext->recovery_passes_required[0];
+			bch2_sb_set_downgrade(c,
+					BCH_VERSION_MINOR(bcachefs_metadata_version_current),
+					BCH_VERSION_MINOR(c->sb.version));
+			passes = ext->recovery_passes_required[0] & ~passes;
+			if (passes) {
+				prt_str(&buf, "  running recovery passes: ");
+				prt_bitflags(&buf, bch2_recovery_passes,
+					     bch2_recovery_passes_from_stable(le64_to_cpu(passes)));
+			}
+
+			bch_info(c, "%s", buf.buf);
+			printbuf_exit(&buf);
+			write_sb = true;
+		}
+
+		if (check_version_upgrade(c))
+			write_sb = true;
+
+		if (write_sb)
+			bch2_write_super(c);
+
+		c->recovery_passes_explicit |= bch2_recovery_passes_from_stable(le64_to_cpu(ext->recovery_passes_required[0]));
+		mutex_unlock(&c->sb_lock);
+	}
+
+	if (c->opts.fsck && IS_ENABLED(CONFIG_BCACHEFS_DEBUG))
+		c->recovery_passes_explicit |= BIT_ULL(BCH_RECOVERY_PASS_check_topology);
+
+	if (c->opts.fsck)
+		set_bit(BCH_FS_fsck_running, &c->flags);
+
 	ret = bch2_blacklist_table_initialize(c);
 	if (ret) {
 		bch_err(c, "error initializing blacklist table");
@@ -865,15 +974,12 @@ use_clean:
 	if (ret)
 		goto err;
 
-	if (c->opts.fsck &&
-	    (IS_ENABLED(CONFIG_BCACHEFS_DEBUG) ||
-	     BCH_SB_HAS_TOPOLOGY_ERRORS(c->disk_sb.sb)))
-		c->recovery_passes_explicit |= BIT_ULL(BCH_RECOVERY_PASS_check_topology);
-
 	ret = bch2_run_recovery_passes(c);
 	if (ret)
 		goto err;
 
+	clear_bit(BCH_FS_fsck_running, &c->flags);
+
 	/* If we fixed errors, verify that fs is actually clean now: */
 	if (IS_ENABLED(CONFIG_BCACHEFS_DEBUG) &&
 	    test_bit(BCH_FS_errors_fixed, &c->flags) &&
@@ -908,16 +1014,30 @@ use_clean:
 	}
 
 	mutex_lock(&c->sb_lock);
-	if (BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb) != c->sb.version) {
-		SET_BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb, c->sb.version);
+	bool write_sb = false;
+
+	if (BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb) != le16_to_cpu(c->disk_sb.sb->version)) {
+		SET_BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb, le16_to_cpu(c->disk_sb.sb->version));
 		write_sb = true;
 	}
 
-	if (!test_bit(BCH_FS_error, &c->flags)) {
+	if (!test_bit(BCH_FS_error, &c->flags) &&
+	    !(c->disk_sb.sb->compat[0] & cpu_to_le64(1ULL << BCH_COMPAT_alloc_info))) {
 		c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_alloc_info);
 		write_sb = true;
 	}
 
+	if (!test_bit(BCH_FS_error, &c->flags)) {
+		struct bch_sb_field_ext *ext = bch2_sb_field_get(c->disk_sb.sb, ext);
+		if (ext &&
+		    (!bch2_is_zero(ext->recovery_passes_required, sizeof(ext->recovery_passes_required)) ||
+		     !bch2_is_zero(ext->errors_silent, sizeof(ext->errors_silent)))) {
+			memset(ext->recovery_passes_required, 0, sizeof(ext->recovery_passes_required));
+			memset(ext->errors_silent, 0, sizeof(ext->errors_silent));
+			write_sb = true;
+		}
+	}
+
 	if (c->opts.fsck &&
 	    !test_bit(BCH_FS_error, &c->flags) &&
 	    !test_bit(BCH_FS_errors_not_fixed, &c->flags)) {
@@ -954,7 +1074,6 @@ use_clean:
 
 	ret = 0;
 out:
-	set_bit(BCH_FS_fsck_done, &c->flags);
 	bch2_flush_fsck_errs(c);
 
 	if (!c->opts.keep_journal &&
@@ -962,13 +1081,14 @@ out:
 		bch2_journal_keys_put_initial(c);
 	kfree(clean);
 
-	if (!ret && test_bit(BCH_FS_need_delete_dead_snapshots, &c->flags)) {
+	if (!ret &&
+	    test_bit(BCH_FS_need_delete_dead_snapshots, &c->flags) &&
+	    !c->opts.nochanges) {
 		bch2_fs_read_write_early(c);
 		bch2_delete_dead_snapshots_async(c);
 	}
 
-	if (ret)
-		bch_err_fn(c, ret);
+	bch_err_fn(c, ret);
 	return ret;
 err:
 fsck_err:
@@ -981,8 +1101,6 @@ int bch2_fs_initialize(struct bch_fs *c)
 	struct bch_inode_unpacked root_inode, lostfound_inode;
 	struct bkey_inode_buf packed_inode;
 	struct qstr lostfound = QSTR("lost+found");
-	struct bch_dev *ca;
-	unsigned i;
 	int ret;
 
 	bch_notice(c, "initializing new filesystem");
@@ -991,7 +1109,7 @@ int bch2_fs_initialize(struct bch_fs *c)
 	c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_extents_above_btree_updates_done);
 	c->disk_sb.sb->compat[0] |= cpu_to_le64(1ULL << BCH_COMPAT_bformat_overflow_done);
 
-	bch2_sb_maybe_downgrade(c);
+	bch2_check_version_downgrade(c);
 
 	if (c->opts.version_upgrade != BCH_VERSION_UPGRADE_none) {
 		bch2_sb_upgrade(c, bcachefs_metadata_version_current);
@@ -1002,12 +1120,11 @@ int bch2_fs_initialize(struct bch_fs *c)
 
 	c->curr_recovery_pass = ARRAY_SIZE(recovery_pass_fns);
 	set_bit(BCH_FS_may_go_rw, &c->flags);
-	set_bit(BCH_FS_fsck_done, &c->flags);
 
-	for (i = 0; i < BTREE_ID_NR; i++)
+	for (unsigned i = 0; i < BTREE_ID_NR; i++)
 		bch2_btree_root_alloc(c, i);
 
-	for_each_member_device(ca, c, i)
+	for_each_member_device(c, ca)
 		bch2_dev_usage_init(ca);
 
 	ret = bch2_fs_journal_alloc(c);
@@ -1035,7 +1152,7 @@ int bch2_fs_initialize(struct bch_fs *c)
 	if (ret)
 		goto err;
 
-	for_each_online_member(ca, c, i)
+	for_each_online_member(c, ca)
 		ca->new_fs_bucket_idx = 0;
 
 	ret = bch2_fs_freespace_init(c);
@@ -1059,10 +1176,9 @@ int bch2_fs_initialize(struct bch_fs *c)
 	packed_inode.inode.k.p.snapshot = U32_MAX;
 
 	ret = bch2_btree_insert(c, BTREE_ID_inodes, &packed_inode.inode.k_i, NULL, 0);
-	if (ret) {
-		bch_err_msg(c, ret, "creating root directory");
+	bch_err_msg(c, ret, "creating root directory");
+	if (ret)
 		goto err;
-	}
 
 	bch2_inode_init_early(c, &lostfound_inode);
 
@@ -1073,10 +1189,11 @@ int bch2_fs_initialize(struct bch_fs *c)
 				  &lostfound,
 				  0, 0, S_IFDIR|0700, 0,
 				  NULL, NULL, (subvol_inum) { 0 }, 0));
-	if (ret) {
-		bch_err_msg(c, ret, "creating lost+found");
+	bch_err_msg(c, ret, "creating lost+found");
+	if (ret)
 		goto err;
-	}
+
+	c->recovery_pass_done = ARRAY_SIZE(recovery_pass_fns) - 1;
 
 	if (enabled_qtypes(c)) {
 		ret = bch2_fs_quota_read(c);
@@ -1085,10 +1202,9 @@ int bch2_fs_initialize(struct bch_fs *c)
 	}
 
 	ret = bch2_journal_flush(&c->journal);
-	if (ret) {
-		bch_err_msg(c, ret, "writing first journal entry");
+	bch_err_msg(c, ret, "writing first journal entry");
+	if (ret)
 		goto err;
-	}
 
 	mutex_lock(&c->sb_lock);
 	SET_BCH_SB_INITIALIZED(c->disk_sb.sb, true);
@@ -1099,6 +1215,6 @@ int bch2_fs_initialize(struct bch_fs *c)
 
 	return 0;
 err:
-	bch_err_fn(ca, ret);
+	bch_err_fn(c, ret);
 	return ret;
 }