Merge tag 'for-6.1-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:

 - revert memory optimization for scrub blocks, this misses errors in
   2nd and following blocks

 - add exception for ENOMEM as reason for transaction abort to not print
   stack trace, syzbot has reported many

 - zoned fixes:
      - fix locking imbalance during scrub
      - initialize zones for seeding device
      - initialize zones for cloned device structures

 - when looking up device, change assertion to a real check as some of
   the search parameters can be passed by ioctl, reported by syzbot

 - fix error pointer check in self tests

* tag 'for-6.1-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: zoned: fix locking imbalance on scrub
  btrfs: zoned: initialize device's zone info for seeding
  btrfs: zoned: clone zoned device info when cloning a device
  Revert "btrfs: scrub: use larger block size for data extent scrub"
  btrfs: don't print stack trace when transaction is aborted due to ENOMEM
  btrfs: selftests: fix wrong error check in btrfs_free_dummy_root()
  btrfs: fix match incorrectly in dev_args_match_device

9 files changed, 109 insertions, 25 deletions

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index b39b339fbf96..a9543f01184c 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -114,6 +114,22 @@ noinline void btrfs_release_path(struct btrfs_path *p)
 }
 
 /*
+ * We want the transaction abort to print stack trace only for errors where the
+ * cause could be a bug, eg. due to ENOSPC, and not for common errors that are
+ * caused by external factors.
+ */
+bool __cold abort_should_print_stack(int errno)
+{
+	switch (errno) {
+	case -EIO:
+	case -EROFS:
+	case -ENOMEM:
+		return false;
+	}
+	return true;
+}
+
+/*
  * safely gets a reference on the root node of a tree.  A lock
  * is not taken, so a concurrent writer may put a different node
  * at the root of the tree.  See btrfs_lock_root_node for the
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index f677b49df8ae..9e6d48ff4597 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -3796,9 +3796,11 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
 			       const char *function,
 			       unsigned int line, int errno, bool first_hit);
 
+bool __cold abort_should_print_stack(int errno);
+
 /*
  * Call btrfs_abort_transaction as early as possible when an error condition is
- * detected, that way the exact line number is reported.
+ * detected, that way the exact stack trace is reported for some errors.
  */
 #define btrfs_abort_transaction(trans, errno)		\
 do {								\
@@ -3807,10 +3809,11 @@ do {								\
 	if (!test_and_set_bit(BTRFS_FS_STATE_TRANS_ABORTED,	\
 			&((trans)->fs_info->fs_state))) {	\
 		first = true;					\
-		if ((errno) != -EIO && (errno) != -EROFS) {		\
-			WARN(1, KERN_DEBUG				\
+		if (WARN(abort_should_print_stack(errno), 	\
+			KERN_DEBUG				\
 			"BTRFS: Transaction aborted (error %d)\n",	\
-			(errno));					\
+			(errno))) {					\
+			/* Stack trace printed. */			\
 		} else {						\
 			btrfs_debug((trans)->fs_info,			\
 				    "Transaction aborted (error %d)", \
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 4b28263c3d32..d99bf7c64611 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2551,7 +2551,9 @@ static int btrfs_read_roots(struct btrfs_fs_info *fs_info)
 		fs_info->dev_root = root;
 	}
 	/* Initialize fs_info for all devices in any case */
-	btrfs_init_devices_late(fs_info);
+	ret = btrfs_init_devices_late(fs_info);
+	if (ret)
+		goto out;
 
 	/*
 	 * This tree can share blocks with some other fs tree during relocation
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index f260c53829e5..196c4c6ed1ed 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -2672,17 +2672,11 @@ static int scrub_extent(struct scrub_ctx *sctx, struct map_lookup *map,
 	u8 csum[BTRFS_CSUM_SIZE];
 	u32 blocksize;
 
-	/*
-	 * Block size determines how many scrub_block will be allocated.  Here
-	 * we use BTRFS_STRIPE_LEN (64KiB) as default limit, so we won't
-	 * allocate too many scrub_block, while still won't cause too large
-	 * bios for large extents.
-	 */
 	if (flags & BTRFS_EXTENT_FLAG_DATA) {
 		if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK)
 			blocksize = map->stripe_len;
 		else
-			blocksize = BTRFS_STRIPE_LEN;
+			blocksize = sctx->fs_info->sectorsize;
 		spin_lock(&sctx->stat_lock);
 		sctx->stat.data_extents_scrubbed++;
 		sctx->stat.data_bytes_scrubbed += len;
@@ -3917,7 +3911,6 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 
 		if (sctx->is_dev_replace && btrfs_is_zoned(fs_info)) {
 			if (!test_bit(BLOCK_GROUP_FLAG_TO_COPY, &cache->runtime_flags)) {
-				spin_unlock(&cache->lock);
 				btrfs_put_block_group(cache);
 				goto skip;
 			}
diff --git a/fs/btrfs/tests/btrfs-tests.c b/fs/btrfs/tests/btrfs-tests.c
index 9c478fa256f6..d43cb5242fec 100644
--- a/fs/btrfs/tests/btrfs-tests.c
+++ b/fs/btrfs/tests/btrfs-tests.c
@@ -200,7 +200,7 @@ void btrfs_free_dummy_fs_info(struct btrfs_fs_info *fs_info)
 
 void btrfs_free_dummy_root(struct btrfs_root *root)
 {
-	if (!root)
+	if (IS_ERR_OR_NULL(root))
 		return;
 	/* Will be freed by btrfs_free_fs_roots */
 	if (WARN_ON(test_bit(BTRFS_ROOT_IN_RADIX, &root->state)))
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index a8d4bc6a1937..635f45f1a2ef 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1011,6 +1011,18 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 			rcu_assign_pointer(device->name, name);
 		}
 
+		if (orig_dev->zone_info) {
+			struct btrfs_zoned_device_info *zone_info;
+
+			zone_info = btrfs_clone_dev_zone_info(orig_dev);
+			if (!zone_info) {
+				btrfs_free_device(device);
+				ret = -ENOMEM;
+				goto error;
+			}
+			device->zone_info = zone_info;
+		}
+
 		list_add(&device->dev_list, &fs_devices->devices);
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
@@ -6918,18 +6930,18 @@ static bool dev_args_match_fs_devices(const struct btrfs_dev_lookup_args *args,
 static bool dev_args_match_device(const struct btrfs_dev_lookup_args *args,
 				  const struct btrfs_device *device)
 {
-	ASSERT((args->devid != (u64)-1) || args->missing);
+	if (args->missing) {
+		if (test_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state) &&
+		    !device->bdev)
+			return true;
+		return false;
+	}
 
-	if ((args->devid != (u64)-1) && device->devid != args->devid)
+	if (device->devid != args->devid)
 		return false;
 	if (args->uuid && memcmp(device->uuid, args->uuid, BTRFS_UUID_SIZE) != 0)
 		return false;
-	if (!args->missing)
-		return true;
-	if (test_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state) &&
-	    !device->bdev)
-		return true;
-	return false;
+	return true;
 }
 
 /*
@@ -7744,10 +7756,11 @@ error:
 	return ret;
 }
 
-void btrfs_init_devices_late(struct btrfs_fs_info *fs_info)
+int btrfs_init_devices_late(struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices, *seed_devs;
 	struct btrfs_device *device;
+	int ret = 0;
 
 	fs_devices->fs_info = fs_info;
 
@@ -7756,12 +7769,18 @@ void btrfs_init_devices_late(struct btrfs_fs_info *fs_info)
 		device->fs_info = fs_info;
 
 	list_for_each_entry(seed_devs, &fs_devices->seed_list, seed_list) {
-		list_for_each_entry(device, &seed_devs->devices, dev_list)
+		list_for_each_entry(device, &seed_devs->devices, dev_list) {
 			device->fs_info = fs_info;
+			ret = btrfs_get_dev_zone_info(device, false);
+			if (ret)
+				break;
+		}
 
 		seed_devs->fs_info = fs_info;
 	}
 	mutex_unlock(&fs_devices->device_list_mutex);
+
+	return ret;
 }
 
 static u64 btrfs_dev_stats_value(const struct extent_buffer *eb,
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index f8b668dc8bf8..099def5613b8 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -671,7 +671,7 @@ int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes,
 void btrfs_dev_stat_inc_and_print(struct btrfs_device *dev, int index);
 int btrfs_get_dev_stats(struct btrfs_fs_info *fs_info,
 			struct btrfs_ioctl_get_dev_stats *stats);
-void btrfs_init_devices_late(struct btrfs_fs_info *fs_info);
+int btrfs_init_devices_late(struct btrfs_fs_info *fs_info);
 int btrfs_init_dev_stats(struct btrfs_fs_info *fs_info);
 int btrfs_run_dev_stats(struct btrfs_trans_handle *trans);
 void btrfs_rm_dev_replace_remove_srcdev(struct btrfs_device *srcdev);
diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index e2d073b08a7d..1912abf6d020 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -639,6 +639,46 @@ void btrfs_destroy_dev_zone_info(struct btrfs_device *device)
 	device->zone_info = NULL;
 }
 
+struct btrfs_zoned_device_info *btrfs_clone_dev_zone_info(struct btrfs_device *orig_dev)
+{
+	struct btrfs_zoned_device_info *zone_info;
+
+	zone_info = kmemdup(orig_dev->zone_info, sizeof(*zone_info), GFP_KERNEL);
+	if (!zone_info)
+		return NULL;
+
+	zone_info->seq_zones = bitmap_zalloc(zone_info->nr_zones, GFP_KERNEL);
+	if (!zone_info->seq_zones)
+		goto out;
+
+	bitmap_copy(zone_info->seq_zones, orig_dev->zone_info->seq_zones,
+		    zone_info->nr_zones);
+
+	zone_info->empty_zones = bitmap_zalloc(zone_info->nr_zones, GFP_KERNEL);
+	if (!zone_info->empty_zones)
+		goto out;
+
+	bitmap_copy(zone_info->empty_zones, orig_dev->zone_info->empty_zones,
+		    zone_info->nr_zones);
+
+	zone_info->active_zones = bitmap_zalloc(zone_info->nr_zones, GFP_KERNEL);
+	if (!zone_info->active_zones)
+		goto out;
+
+	bitmap_copy(zone_info->active_zones, orig_dev->zone_info->active_zones,
+		    zone_info->nr_zones);
+	zone_info->zone_cache = NULL;
+
+	return zone_info;
+
+out:
+	bitmap_free(zone_info->seq_zones);
+	bitmap_free(zone_info->empty_zones);
+	bitmap_free(zone_info->active_zones);
+	kfree(zone_info);
+	return NULL;
+}
+
 int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 		       struct blk_zone *zone)
 {
diff --git a/fs/btrfs/zoned.h b/fs/btrfs/zoned.h
index e17462db3a84..8bd16d40b7c6 100644
--- a/fs/btrfs/zoned.h
+++ b/fs/btrfs/zoned.h
@@ -36,6 +36,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 int btrfs_get_dev_zone_info_all_devices(struct btrfs_fs_info *fs_info);
 int btrfs_get_dev_zone_info(struct btrfs_device *device, bool populate_cache);
 void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
+struct btrfs_zoned_device_info *btrfs_clone_dev_zone_info(struct btrfs_device *orig_dev);
 int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
 int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info);
 int btrfs_sb_log_location_bdev(struct block_device *bdev, int mirror, int rw,
@@ -103,6 +104,16 @@ static inline int btrfs_get_dev_zone_info(struct btrfs_device *device,
 
 static inline void btrfs_destroy_dev_zone_info(struct btrfs_device *device) { }
 
+/*
+ * In case the kernel is compiled without CONFIG_BLK_DEV_ZONED we'll never call
+ * into btrfs_clone_dev_zone_info() so it's safe to return NULL here.
+ */
+static inline struct btrfs_zoned_device_info *btrfs_clone_dev_zone_info(
+						 struct btrfs_device *orig_dev)
+{
+	return NULL;
+}
+
 static inline int btrfs_check_zoned_mode(const struct btrfs_fs_info *fs_info)
 {
 	if (!btrfs_is_zoned(fs_info))