Add basic XCM emulator tests

[acatangiu]

This PR adds XCM emulator tests for Relay chains and Asset Hub System parachains.

It adds them under new integration-tests/emulated/ dir which lives at the root of the repo and doesn't pollute the prod networks code/crates/etc in any way. As you can see from the diff, no other files outside integration-tests are touched by this PR (except workspace Cargo.toml and Cargo.lock).

Tests added by this PR are rather old/limited, but I will update them when we bump polkadot-sdk crates deps to v1.4 or v1.5 release. This will bring in more refined tests for existing scenarios and many new tests for new scenarios such as complex asset-transfers and bridging - sending XCMs over bridge, and asset transfers between AHs on different relays using bridge.

Also after bumping repo to sdk v1.4 or v1.5 release, we can also drop 90% of integration-tests/emulated/common and get it from crates.io.

Fixes #103

[acatangiu]

After bumping repo to sdk v1.4 or v1.5 release, we can drop 90% of the code in this module and get it from https://crates.io/crates/emulated-integration-tests-common/1.0.0

[NachoPal]

@acatangiu We are planing to add XCM emulator to its own repository. It is still WIP: https://github.com/paritytech/polkadot-integration-tests/tree/master

[acatangiu]

@acatangiu We are planing to add XCM emulator to its own repository. It is still WIP: https://github.com/paritytech/polkadot-integration-tests/tree/master

Why would they be on a different repo? I mean I don't see any benefits, while seeing massive drawbacks. It's already hard enough to coordinate over polkadot-sdk and polkadot-fellows, why would we make it even harder?

[NachoPal]

@acatangiu We are planing to add XCM emulator to its own repository. It is still WIP: https://github.com/paritytech/polkadot-integration-tests/tree/master

Why would they be on a different repo? I mean I don't see any benefits, while seeing massive drawbacks. It's already hard enough to coordinate over polkadot-sdk and polkadot-fellows, why would we make it even harder?

Let's continue the conversation here: #103 (comment)

[acatangiu]

@bkchr @ordian this needs another Rank 3 reviewer because it touches CI files.
(also please re-trigger the failed migration test as CI was API rate-limited when trying to set up test env)

[ordian]

I recall @rphmeier having issues with how those tests were written affecting async backing changes

[acatangiu]

I recall @rphmeier having issues with how those tests were written affecting async backing changes

I can't think of any issues with async backing - afaik the emulator doesn't build any blocks per-se - it emulates everything underneath the runtime using some dumb buffers - point being to test the runtime behavior, especially XCM message passing and execution.

These tests currently run on Rococo and Westend too, where I was under the impression async backing is already enabled. Or are you talking about some limitation on some yet unreleased async backing feature?

[joepetrowski]

The async backing issues should have all been fixed in:

• Make xcm-emulator test pass for new consensus hook paritytech/cumulus#2722
• Fix for xcm-emulator - Wrong Parachain processing message paritytech/cumulus#2800

[acatangiu]

@ordian @bkchr this is good to go

[ordian]

Didn't looked at the tests closely, but since most of it is going to be removed after bump to 1.5 is merged as you say, don't want to block this PR.

[acatangiu]

@joepetrowski bot is not counting your review after latest push, can you try to re-approve?

[acatangiu]

/merge

[acatangiu]

@bkchr @ggwpez who has merge button?

[bkchr]

@acatangiu please be patient :P https://github.com/polkadot-fellows/runtimes/actions/runs/7410194917/job/20161942771

[fellowship-merge-bot]

Enabled auto-merge in Pull Request

Available commands

• /merge: Enables auto-merge for Pull Request
• /merge cancel: Cancels auto-merge for Pull Request
• /merge help: Shows this menu

For more information see the documentation

[acatangiu]

@acatangiu please be patient :P https://github.com/polkadot-fellows/runtimes/actions/runs/7410194917/job/20161942771

ah, nice, I thought bot is ignoring me since I'm not part of fellowship, but I was just being impatient 😛

[bkchr]

Yeah, we had a lot of jobs running and github action has probably some upper number of parallel runners per project.

[ggwpez]

There are 4 failing tests here after a weight update: #129
I fixed two of them, but dont know about the other two.

Its also waiting for cumulus_pallet_dmp_queue::Event::ExecutedDownward, that event type got removed two months ago. So maybe updating to the new crates versions fixes this.

[acatangiu]

Its also waiting for cumulus_pallet_dmp_queue::Event::ExecutedDownward, that event type got removed two months ago. So maybe updating to the new crates versions fixes this.

Yes, it’s using old(er) polkadot-sdk crate versions (from v1.3.0 release). When we’ll bump deps, we’ll update tests too.

There are 4 failing tests here after a weight update: #129 I fixed two of them, but dont know about the other two.

I can look into it tomorrow.