Fixes #9570

bolt2 says the receiver of tx_init_rbf must fail (warning + disconnect, or error) if the sender previously sent splice_locked. splice_accepter() did not check this, so a peer could send splice_locked and then tx_init_rbf, and we would start a new RBF session

This adds a check on splice_state->remote_locked_txid in the WIRE_TX_INIT_RBF branch and calls peer_failed_warn() if it is set

No funds were at risk - the RBF still went through the normal balance checks and signing

Testing
test_splice_rbf passes. A CLN peer never sends this sequence, so there is no pytest regression test for it

Changelog-Fixed: channeld: reject 'tx_init_rbf' if the peer already sent 'splice_locked', as required by bolt2

Checklist

Before submitting the PR, ensure the following tasks are completed. If an item is not applicable to your PR, please mark it as checked:

• The changelog has been updated in the relevant commit(s) according to the guidelines.
• Tests have been added or modified to reflect the changes.
• Documentation has been reviewed and updated as needed.
• Related issues have been listed and linked, including any that this PR closes.
• Important All PRs must consider how to reverse any persistent changes for tools/lightning-downgrade