Merge tag 'for-6.0-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "A few more fixes to zoned mode and one regression fix for chunk limit:

    - Zoned mode fixes:
        - fix how wait/wake up is done when finishing zone
        - fix zone append limit in emulated mode
        - fix mount on devices with conventional zones

   - fix regression, user settable data chunk limit got accidentally
     lowered and causes allocation problems on some profiles (raid0,
     raid1)"

* tag 'for-6.0-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: fix the max chunk size and stripe length calculation
  btrfs: zoned: fix mounting with conventional zones
  btrfs: zoned: set pseudo max append zone limit in zone emulation mode
  btrfs: zoned: fix API misuse of zone finish waiting

Author: torvalds

fs/btrfs/ctree.h

@@ -1088,8 +1088,6 @@ struct btrfs_fs_info {
 
 	spinlock_t zone_active_bgs_lock;
 	struct list_head zone_active_bgs;
-	/* Waiters when BTRFS_FS_NEED_ZONE_FINISH is set */
-	wait_queue_head_t zone_finish_wait;
 
 	/* Updates are not protected by any lock */
 	struct btrfs_commit_stats commit_stats;

fs/btrfs/disk-io.c

@@ -3068,7 +3068,6 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
 	init_waitqueue_head(&fs_info->transaction_blocked_wait);
 	init_waitqueue_head(&fs_info->async_submit_wait);
 	init_waitqueue_head(&fs_info->delayed_iputs_wait);
-	init_waitqueue_head(&fs_info->zone_finish_wait);
 
 	/* Usable values until the real ones are cached from the superblock */
 	fs_info->nodesize = 4096;

fs/btrfs/inode.c

@@ -1644,10 +1644,9 @@ static noinline int run_delalloc_zoned(struct btrfs_inode *inode,
 			done_offset = end;
 
 		if (done_offset == start) {
-			struct btrfs_fs_info *info = inode->root->fs_info;
-
-			wait_var_event(&info->zone_finish_wait,
-				       !test_bit(BTRFS_FS_NEED_ZONE_FINISH, &info->flags));
+			wait_on_bit_io(&inode->root->fs_info->flags,
+				       BTRFS_FS_NEED_ZONE_FINISH,
+				       TASK_UNINTERRUPTIBLE);
 			continue;
 		}
 

fs/btrfs/space-info.c

@@ -199,7 +199,7 @@ static u64 calc_chunk_size(const struct btrfs_fs_info *fs_info, u64 flags)
 	ASSERT(flags & BTRFS_BLOCK_GROUP_TYPE_MASK);
 
 	if (flags & BTRFS_BLOCK_GROUP_DATA)
-		return SZ_1G;
+		return BTRFS_MAX_DATA_CHUNK_SIZE;
 	else if (flags & BTRFS_BLOCK_GROUP_SYSTEM)
 		return SZ_32M;
 

fs/btrfs/volumes.c

@@ -5267,6 +5267,9 @@ static int decide_stripe_size_regular(struct alloc_chunk_ctl *ctl,
 				       ctl->stripe_size);
 	}
 
+	/* Stripe size should not go beyond 1G. */
+	ctl->stripe_size = min_t(u64, ctl->stripe_size, SZ_1G);
+
 	/* Align to BTRFS_STRIPE_LEN */
 	ctl->stripe_size = round_down(ctl->stripe_size, BTRFS_STRIPE_LEN);
 	ctl->chunk_size = ctl->stripe_size * data_stripes;

fs/btrfs/zoned.c

@@ -421,10 +421,19 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device, bool populate_cache)
 	 * since btrfs adds the pages one by one to a bio, and btrfs cannot
 	 * increase the metadata reservation even if it increases the number of
 	 * extents, it is safe to stick with the limit.
+	 *
+	 * With the zoned emulation, we can have non-zoned device on the zoned
+	 * mode. In this case, we don't have a valid max zone append size. So,
+	 * use max_segments * PAGE_SIZE as the pseudo max_zone_append_size.
 	 */
-	zone_info->max_zone_append_size =
-		min_t(u64, (u64)bdev_max_zone_append_sectors(bdev) << SECTOR_SHIFT,
-		      (u64)bdev_max_segments(bdev) << PAGE_SHIFT);
+	if (bdev_is_zoned(bdev)) {
+		zone_info->max_zone_append_size = min_t(u64,
+			(u64)bdev_max_zone_append_sectors(bdev) << SECTOR_SHIFT,
+			(u64)bdev_max_segments(bdev) << PAGE_SHIFT);
+	} else {
+		zone_info->max_zone_append_size =
+			(u64)bdev_max_segments(bdev) << PAGE_SHIFT;
+	}
 	if (!IS_ALIGNED(nr_sectors, zone_sectors))
 		zone_info->nr_zones++;
 
@@ -1178,7 +1187,7 @@ int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size)
  * offset.
  */
 static int calculate_alloc_pointer(struct btrfs_block_group *cache,
-				   u64 *offset_ret)
+				   u64 *offset_ret, bool new)
 {
 	struct btrfs_fs_info *fs_info = cache->fs_info;
 	struct btrfs_root *root;
@@ -1188,6 +1197,21 @@ static int calculate_alloc_pointer(struct btrfs_block_group *cache,
 	int ret;
 	u64 length;
 
+	/*
+	 * Avoid  tree lookups for a new block group, there's no use for it.
+	 * It must always be 0.
+	 *
+	 * Also, we have a lock chain of extent buffer lock -> chunk mutex.
+	 * For new a block group, this function is called from
+	 * btrfs_make_block_group() which is already taking the chunk mutex.
+	 * Thus, we cannot call calculate_alloc_pointer() which takes extent
+	 * buffer locks to avoid deadlock.
+	 */
+	if (new) {
+		*offset_ret = 0;
+		return 0;
+	}
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -1323,6 +1347,13 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
 		else
 			num_conventional++;
 
+		/*
+		 * Consider a zone as active if we can allow any number of
+		 * active zones.
+		 */
+		if (!device->zone_info->max_active_zones)
+			__set_bit(i, active);
+
 		if (!is_sequential) {
 			alloc_offsets[i] = WP_CONVENTIONAL;
 			continue;
@@ -1389,45 +1420,23 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
 			__set_bit(i, active);
 			break;
 		}
-
-		/*
-		 * Consider a zone as active if we can allow any number of
-		 * active zones.
-		 */
-		if (!device->zone_info->max_active_zones)
-			__set_bit(i, active);
 	}
 
 	if (num_sequential > 0)
 		cache->seq_zone = true;
 
 	if (num_conventional > 0) {
-		/*
-		 * Avoid calling calculate_alloc_pointer() for new BG. It
-		 * is no use for new BG. It must be always 0.
-		 *
-		 * Also, we have a lock chain of extent buffer lock ->
-		 * chunk mutex.  For new BG, this function is called from
-		 * btrfs_make_block_group() which is already taking the
-		 * chunk mutex. Thus, we cannot call
-		 * calculate_alloc_pointer() which takes extent buffer
-		 * locks to avoid deadlock.
-		 */
-
 		/* Zone capacity is always zone size in emulation */
 		cache->zone_capacity = cache->length;
-		if (new) {
-			cache->alloc_offset = 0;
-			goto out;
-		}
-		ret = calculate_alloc_pointer(cache, &last_alloc);
-		if (ret || map->num_stripes == num_conventional) {
-			if (!ret)
-				cache->alloc_offset = last_alloc;
-			else
-				btrfs_err(fs_info,
+		ret = calculate_alloc_pointer(cache, &last_alloc, new);
+		if (ret) {
+			btrfs_err(fs_info,
 			"zoned: failed to determine allocation offset of bg %llu",
-					  cache->start);
+				  cache->start);
+			goto out;
+		} else if (map->num_stripes == num_conventional) {
+			cache->alloc_offset = last_alloc;
+			cache->zone_is_active = 1;
 			goto out;
 		}
 	}
@@ -1495,13 +1504,6 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
 		goto out;
 	}
 
-	if (cache->zone_is_active) {
-		btrfs_get_block_group(cache);
-		spin_lock(&fs_info->zone_active_bgs_lock);
-		list_add_tail(&cache->active_bg_list, &fs_info->zone_active_bgs);
-		spin_unlock(&fs_info->zone_active_bgs_lock);
-	}
-
 out:
 	if (cache->alloc_offset > fs_info->zone_size) {
 		btrfs_err(fs_info,
@@ -1526,10 +1528,16 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
 		ret = -EIO;
 	}
 
-	if (!ret)
+	if (!ret) {
 		cache->meta_write_pointer = cache->alloc_offset + cache->start;
-
-	if (ret) {
+		if (cache->zone_is_active) {
+			btrfs_get_block_group(cache);
+			spin_lock(&fs_info->zone_active_bgs_lock);
+			list_add_tail(&cache->active_bg_list,
+				      &fs_info->zone_active_bgs);
+			spin_unlock(&fs_info->zone_active_bgs_lock);
+		}
+	} else {
 		kfree(cache->physical_map);
 		cache->physical_map = NULL;
 	}
@@ -2007,8 +2015,7 @@ static int do_zone_finish(struct btrfs_block_group *block_group, bool fully_writ
 	/* For active_bg_list */
 	btrfs_put_block_group(block_group);
 
-	clear_bit(BTRFS_FS_NEED_ZONE_FINISH, &fs_info->flags);
-	wake_up_all(&fs_info->zone_finish_wait);
+	clear_and_wake_up_bit(BTRFS_FS_NEED_ZONE_FINISH, &fs_info->flags);
 
 	return 0;
 }