Adding support to send batch extrinsics to the parachain

[silva-fj]

As topic: This PR adds the structure to send batch extrinsics to the parachain asynchronously. Starting with vc_issued

[linear]

P-554 implement submitting batched callback extrinsics to the parachain

[Kailai-Wang]

Thanks! The logic looks fine, only have questions regarding the batch parameters

[BillyWooo]

In general, LGTM.
Except "too many calls", just one more point: are we considering to not use such global variables GLOBAL_PARACHAIN_EXTRINSIC_TASK in the future?
maybe @Kailai-Wang have any idea?

[Kailai-Wang]

I approve it as it's in any case an improvement to current structure.

there should be a limit of byte array size going through the ocall

Please double-check it and/or change the limit

Shall we use this thread to send other extrinsics? Like teebag callbacks? My preference is yes.

Currently it's often used as:

		self.validator_accessor
			.execute_mut_on_validator(|v| v.send_extrinsics(parentchain_extrinsics))?;

Does it have the same effect as having this global thread?

[silva-fj]

there should be a limit of byte array size going through the ocall

The limit for extrinsics_encoded_size is u32, it is big enough

[silva-fj]

Shall we use this thread to send other extrinsics? Like teebag callbacks? My preference is yes.

Currently it's often used as:

		self.validator_accessor
			.execute_mut_on_validator(|v| v.send_extrinsics(parentchain_extrinsics))?;

Does it have the same effect as having this global thread?

So, there are a couple of differences:
When using validator_access.execute_mut_on_validator(|v| v.send_extrinsics(xts))?;, after sending the extrinsics, we seal the LightValidationState, at the moment I'm not 100% sure what this implies.

The other difference is the nature of the interval batching, if we are Ok with eventual consistency there should be no problem.

@Kailai-Wang Should I make the changes to send all extrinsics to the new thread as part of this PR? or should we create a follow up issue?

[Kailai-Wang]

Please do it in a separate PR : )