Actors v2 upgrade setup #854
Conversation
austinabell
force-pushed
the
austin/actv2setup
branch
from
779edb9
to
bb0c3b1
Compare
austinabell
added
Spec Change
austinabell
requested review from
dutterbutter,
ec2 and
timvermeulen
as code owners
| @@ -12,7 +12,7 @@ pub use self::message::*; | |||
| pub use self::rand_replay::*; | |||
| pub use self::stubs::*; | |||
| pub use self::tipset::*; | |||
| use actor::CHAOS_ACTOR_CODE_ID; | |||
| use actor::actorv2::CHAOS_ACTOR_CODE_ID; | |||
There was a problem hiding this comment.
Curious how you're passing v1 conformance tests with v2 chaos actor
There was a problem hiding this comment.
The code Cid is still using the /1/ path. We probably shouldn't even have chaos within actors because it isn't tied to the actor versions. But can change later
| #[derive(Debug, PartialEq, Serialize)] | ||
| #[serde(rename_all = "PascalCase")] | ||
| pub struct MinerInfo { | ||
| pub owner: Address, |
There was a problem hiding this comment.
These need to be serialized using json
There was a problem hiding this comment.
Very funky fresh setup. I focused mostly outside of the VM, so the state manager, mpool, and chain sync.
I looked at the structure of the stuff in the actors and vm code. Didnt look too hard into the correctness because if it passes conformance v1 actors and can still sync to breeze, then it should be all fine.
Tested against storage miner init functionality as well and its all good after that json fix.
Summary of changes
Changes introduced in this pull request:
This is blocked on finishing and exporting v1 actors, which is not yet but soon ready to be done, but I will just keep putting changes in this PR which are necessary until then
Reference issue to close (if applicable)
Closes #816
Closes #817
Closes #815
Closes #824
Other information and links