[Bug]: Assigning public IPv4 to an existing server breaks reverse DNS entry

[cebe]

What happened?

# terraform apply
[...]
  # hcloud_rdns.cloud_ftp_ipv4[0] will be created
  + resource "hcloud_rdns" "cloud_ftp_ipv4" {
      + dns_ptr    = "ftp1.ftp.cloud2.[..]"
      + id         = (known after apply)
      + server_id  = 48738290
        # (1 unchanged attribute hidden)
    }

  # hcloud_rdns.cloud_ftp_ipv6[0] will be created
  + resource "hcloud_rdns" "cloud_ftp_ipv6" {
      + dns_ptr    = "ftp1.ftp.cloud2.[..]"
      + id         = (known after apply)
      + ip_address = "2a01:4f8:[..]"
      + server_id  = 48738290
    }

  # hcloud_server.cloud_ftp[0] will be updated in-place
  ~ resource "hcloud_server" "cloud_ftp" {
        id                         = "48738290"
        name                       = "ftp1.ftp.cloud2.[..]"
        # (22 unchanged attributes hidden)

      - public_net {
          - ipv4         = 0 -> null
          - ipv4_enabled = false -> null
          - ipv6         = 0 -> null
          - ipv6_enabled = true -> null
        }
      + public_net {
          + ipv4         = (known after apply)
          + ipv4_enabled = true
          + ipv6         = (known after apply)
          + ipv6_enabled = true
        }

        # (1 unchanged block hidden)
    }

[...]
Plan: 3 to add, 3 to change, 1 to destroy.

Do you want to perform these actions?
  Terraform will perform the actions described above.
  Only 'yes' will be accepted to approve.

  Enter a value: yes
[...]
hcloud_server.cloud_ftp[0]: Modifying... [id=48738290]
hcloud_server.cloud_ftp[0]: Still modifying... [id=48738290, 10s elapsed]
hcloud_server.cloud_ftp[0]: Modifications complete after 16s [id=48738290]
hcloud_rdns.cloud_ftp_ipv6[0]: Creating...
hcloud_rdns.cloud_ftp_ipv6[0]: Creation complete after 2s [id=s-48738290-2a01:4f8:c17:62bf::1]
╷
│ Error: Provider produced inconsistent final plan
│ 
│ When expanding the plan for hcloud_rdns.cloud_ftp_ipv4[0] to include new values learned so far during apply, provider "registry.terraform.io/hetznercloud/hcloud" produced an invalid new value for .ip_address: was cty.StringVal(""), but now cty.StringVal("157.90.248.65").
│ 
│ This is a bug in the provider, which should be reported in the provider's own issue tracker.
╵

What did you expect to happen?

I expect the new IPv4 added and the reverse DNS entry to be updated without error.

Please provide a minimal working example

terraform {
  required_providers {
    hcloud = {
      source  = "hetznercloud/hcloud"
      version = "1.47.0"
    }
  }
  required_version = ">= 1.0.0"
}

provider "hcloud" {
  token = "Neo8FwsD6w2aEM1...N9U"
}

# use SSH key from home dir
resource "hcloud_ssh_key" "instance-ssh-key" {
  name       = "instance-ssh-key"
  public_key = file("~/.ssh/id_rsa.pub")
}

resource "hcloud_server" "cloud_ftp" {
  count = 1

  name        = "ftp${count.index + 1}.ftp.example.com"
  image       = "debian-12"
  server_type = "cax21"
  location    = "hel1"
  ssh_keys    = [hcloud_ssh_key.instance-ssh-key.id]
  # in case of upgrade, keep disk size to allow downgrade later
  keep_disk   = true
  public_net {
    ipv4_enabled = false
#    ipv4_enabled = true
    ipv6_enabled = true
  }
  depends_on = [
    hcloud_ssh_key.instance-ssh-key
  ]
}

resource "hcloud_rdns" "cloud_ftp_ipv6" {
  count = 1
  #  for = toset(range(0, var.k8s_cluster_nodes - 1))
  server_id  = hcloud_server.cloud_ftp[count.index].id
  ip_address = hcloud_server.cloud_ftp[count.index].ipv6_address
  dns_ptr    = hcloud_server.cloud_ftp[count.index].name
}

#resource "hcloud_rdns" "cloud_ftp_ipv4" {
#  count = 1
#  #  for = toset(range(0, var.k8s_cluster_nodes - 1))
#  server_id  = hcloud_server.cloud_ftp[count.index].id
#  ip_address = hcloud_server.cloud_ftp[count.index].ipv4_address
#  dns_ptr    = hcloud_server.cloud_ftp[count.index].name
#}

steps to reproduce:

1. create file with the content above in an empty directory, enter your hetzner cloud api key
2. terraform init
3. terraform apply
4. set ipv4_enabled = true in servers public_net
5. uncomment resource "hcloud_rdns"
6. terraform apply

[cebe]

Workaround: Running terraform apply again, makes the correct changes.