contracts: switch from `parity-wasm`-based to `wasmi`-based module validation

[agryaznov]

We keep only the following validation checks to be enforced to a Wasm module of a contract:

• scan_imports(),
• scan_exports().

Once switched to wasmi builtin gas metering (#14084), we don't need to perform other checks anymore.

All the validation is now being done through wasmi querying. We thereby get rid of wasmparser dependency, and keep wasm-instrument/parity-wasm dependency only for benchmarks, where it is used to generate contract modules.

[agryaznov]

bot bench $ pallet dev pallet_contracts

[command-bot[bot]]

@agryaznov "$PIPELINE_SCRIPTS_DIR/commands/bench/bench.sh" pallet dev pallet_contracts (https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3113351) was cancelled in https://github.com/paritytech/substrate/pull/14449#issuecomment-1618920794

[agryaznov]

bot cancel

[command-bot[bot]]

@agryaznov Command "$PIPELINE_SCRIPTS_DIR/commands/bench/bench.sh" pallet dev pallet_contracts has finished. Result: https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3113351 has finished. If any artifacts were generated, you can download them from https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3113351/artifacts/download.

[agryaznov]

bot bench $ pallet dev pallet_contracts

[command-bot[bot]]

@agryaznov https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3114006 was started for your command "$PIPELINE_SCRIPTS_DIR/commands/bench/bench.sh" pallet dev pallet_contracts. Check out https://gitlab.parity.io/parity/mirrors/substrate/-/pipelines?page=1&scope=all&username=group_605_bot to know what else is being executed currently.

Comment bot cancel 16-4176a09d-692b-4725-b24d-d17472fed7ad to cancel this command or bot cancel to cancel all commands in this pull request.

[command-bot[bot]]

@agryaznov Command "$PIPELINE_SCRIPTS_DIR/commands/bench/bench.sh" pallet dev pallet_contracts has finished. Result: https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3114006 has finished. If any artifacts were generated, you can download them from https://gitlab.parity.io/parity/mirrors/substrate/-/jobs/3114006/artifacts/download.

[agryaznov]

Some regressions on the weights have been detected:

I'm not sure if it should be a stopper for this PR though, as we still need to re-work benchmarks for all host functions (see note in this https://github.com/paritytech/substrate/pull/14084#issuecomment-1596850814 and https://github.com/paritytech/substrate/pull/14084#issuecomment-1619112784). For now we (over)charge the engine spent part twice for every host function because of that.

[athei]

Can you please try to explain in more detail why you think the host function benchmarks overestimate? The benchmarks only measure the time it takes to call the host function which should be unaffected by out changes.

[agryaznov]

bot merge

[agryaznov]

Can you please try to explain in more detail why you think the host function benchmarks overestimate? The benchmarks only measure the time it takes to call the host function which should be unaffected by out changes.

@athei, I tried to give the explanation in this comment But this "overestimation" seems to lie far below current benchmarking error anyway, so this was just my hypothesis which failed to prove the reasons of those weights change.