Introduce cancel_proposal to rid us of those pesky proposals

[gavofyork]

Also fixes the weight of propose and introduces a hard-core blacklist option to ensure a proposal hash never returns.

polkadot companion: paritytech/polkadot#1728

TODO:

• Weights (cc @shawntabrizi )

[joepetrowski]

Breaks transaction_version in case that hasn't been updated for the next release.

[kianenigma]

self/random note that some other modules can generally use this terminology: Some data that is passed in for sanity/weight should be called witness data.

[shawntabrizi]

Other than my comments, this looks good.

[gavofyork]

@kianenigma @thiolliere @shawntabrizi ready for final review...

[gui1117]

looks good to me

[gavofyork]

@kianenigma @shawntabrizi would be good to get a final review.

[shawntabrizi]

on_initialize weights look borked when running benchmarks again:

// before
		(70826000 as Weight)
			.saturating_add((10716000 as Weight).saturating_mul(r as Weight))
			.saturating_add(T::DbWeight::get().reads(6 as Weight))
			.saturating_add(T::DbWeight::get().reads((2 as Weight).saturating_mul(r as Weight)))
			.saturating_add(T::DbWeight::get().writes(5 as Weight))
//after
		(12_357_000 as Weight)
			.saturating_add((9_372_000 as Weight).saturating_mul(r as Weight))
			.saturating_add(T::DbWeight::get().reads(5 as Weight))
			.saturating_add(T::DbWeight::get().reads((1 as Weight).saturating_mul(r as Weight)))

Missing reads and writes. I am investigating

[shawntabrizi]

I pushed the latest Weights to the PR. Note that the only relevant changes was an additional read to the propose weight, caused by checking the blacklist, and a reduction in the on_initialize_base weight due to bad benchmarking code.

To clarify the intention, on_initialize_base is supposed to be the least amount of computation needed by the on_initialize code, so in this case (5 + R) Reads.

If the on_initialize code actually triggers any of the more complex paths, like launch_next or bake_referendum, we will assume that the computation should take a full block weight. This is hardcoded into the on_initialize logic.

Tracking Issue: https://github.com/paritytech/substrate/issues/7209

[gavofyork]

bot merge

[ghost]

Trying merge.