btrfs: don't unconditionally call folio_start_writeback in subpage

In the normal case we check if a page is under writeback and skip it
before we attempt to begin writeback.

The exception is subpage metadata writes, where we know we don't have an
eb under writeback and we're doing it one eb at a time.  Since
b5612c3 ("mm: return void from folio_start_writeback() and related
functions") we now will BUG_ON() if we call folio_start_writeback()
on a folio that's already under writeback.  Previously
folio_start_writeback() would bail if writeback was already started.

Fix this in the subpage code by checking if we have writeback set and
skipping it if we do.  This fixes the panic we were seeing on subpage.

Reviewed-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>

fs/btrfs/subpage.c

@@ -475,7 +475,8 @@ void btrfs_subpage_set_writeback(const struct btrfs_fs_info *fs_info,
 
 	spin_lock_irqsave(&subpage->lock, flags);
 	bitmap_set(subpage->bitmaps, start_bit, len >> fs_info->sectorsize_bits);
-	folio_start_writeback(folio);
+	if (!folio_test_writeback(folio))
+		folio_start_writeback(folio);
 	spin_unlock_irqrestore(&subpage->lock, flags);
 }