collation-generation + collator-protocol: collate on multiple assigned cores (#3795)

This works only for collators that implement the `collator_fn` allowing
`collation-generation` subsystem to pull collations triggered on new
heads.

Also enables
`request_v2::CollationFetchingResponse::CollationWithParentHeadData` for
test adder/undying collators.

TODO:
- [x] fix tests
- [x] new tests
- [x] PR doc

---------

Signed-off-by: Andrei Sandu <andrei-mihail@parity.io>

polkadot/node/primitives/src/lib.rs

@@ -31,8 +31,8 @@ use serde::{de, Deserialize, Deserializer, Serialize, Serializer};
 
 use polkadot_primitives::{
 	BlakeTwo256, BlockNumber, CandidateCommitments, CandidateHash, CollatorPair,
-	CommittedCandidateReceipt, CompactStatement, EncodeAs, Hash, HashT, HeadData, Id as ParaId,
-	PersistedValidationData, SessionIndex, Signed, UncheckedSigned, ValidationCode,
+	CommittedCandidateReceipt, CompactStatement, CoreIndex, EncodeAs, Hash, HashT, HeadData,
+	Id as ParaId, PersistedValidationData, SessionIndex, Signed, UncheckedSigned, ValidationCode,
 	ValidationCodeHash, ValidatorIndex, MAX_CODE_SIZE, MAX_POV_SIZE,
 };
 pub use sp_consensus_babe::{
@@ -524,6 +524,8 @@ pub struct SubmitCollationParams {
 	/// okay to just drop it. However, if it is called, it should be called with the signed
 	/// statement of a parachain validator seconding the collation.
 	pub result_sender: Option<futures::channel::oneshot::Sender<CollationSecondedSignal>>,
+	/// The core index on which the resulting candidate should be backed
+	pub core_index: CoreIndex,
 }
 
 /// This is the data we keep available for each candidate included in the relay chain.