Add transaction tag for fraud proof and re-define fraud proof priority

[NingLin-P]

This PR implement #2570, include changes of:

• Add transaction tag for fraud proof
  • For fraud proof that targets bad ER, for a given domain at most one fraud proof will be accepted at a time
  • For bundle equivocation fraud proof, for a given operator at most one fraud proof will be accepted at a time
• Re-define fraud proof priority
  • Fraud proof priority is defined as MAX - block_before_bad_er_confirm, thus fraud proof that targets closer to confirm bad ER (more emergency) has a higher priority to be accepted by tx pool and to be included in block.
• Reduce the priority of bundle and XDM from MAX to 1
  • So the bundle and XDM have a higher priority than other normal extrinsic (which has 0 priority), but less than the fraud proof

There is a TODO left in this PR, although an incoming fraud proof will be rejected if there is already a fraud proof with higher priority accepted by the tx pool but it is only rejected after it is validated when the cost is already made, we can avoid this cost by checking if a fraud proof with the same tag and higher priority already exists in the pool but unfortunately upstream doesn't support query tx by tag yet, I have open paritytech/polkadot-sdk#3705 for this feature.

I also tried to write a test to submit multiple fraud proofs and see the behavior, but the operator only produced fraud proof for the oldest bad ER and it is too hack to manually construct a fraud proof, but after the above TODO is resolved fraud proof will be rejected before it is validated thus we can use a dummy fraud proof, so will add the test by then.

Code contributor checklist:

• I have read, understood and followed contributing guide

[dariolina]

Please add a section like "Fraud proof priority in tx pool" with a few sentences like you have in the description to the bottom of this file https://github.com/subspace/protocol-specs/blob/main/docs/decex/fraud_proofs.md (I may find a better place for it later)

[dariolina]

The spec was updated and this looks good to me. @vedhavyas could you take a final look at the code?

[vanhauser-thc]

We also found this issue at the same time.
@dariolina mabe this PR should have the "need audit" flag?