feat: add post-state-root data structs #9532
Conversation
ef7abd8 to
1d66984
Compare
| @@ -160,4 +160,4 @@ pub(super) enum PeerMessage { | |||
| VersionedStateResponse(StateResponseInfo), | |||
| } | |||
| #[cfg(target_arch = "x86_64")] // Non-x86_64 doesn't match this requirement yet but it's not bad as it's not production-ready | |||
| const _: () = assert!(std::mem::size_of::<PeerMessage>() <= 1144, "PeerMessage > 1144 bytes"); | |||
| const _: () = assert!(std::mem::size_of::<PeerMessage>() <= 1500, "PeerMessage > 1500 bytes"); | |||
There was a problem hiding this comment.
not sure what the value of having that and it breaks builds when new fields are added to block or chunk headers, so I'm increasing this value to avoid updating that after each new field is added
There was a problem hiding this comment.
Is there any comment or meaningful explanation in the PR where this was introduced? Given this is networking one guess would be something to do with MTU but I doubt that matters to us.
There was a problem hiding this comment.
nope, I couldn't find any meaningful explanation on why this is needed.
@saketh-are maybe you (as a networking expert 😄) know something about that?
| @@ -160,4 +160,4 @@ pub(super) enum PeerMessage { | |||
| VersionedStateResponse(StateResponseInfo), | |||
| } | |||
| #[cfg(target_arch = "x86_64")] // Non-x86_64 doesn't match this requirement yet but it's not bad as it's not production-ready | |||
| const _: () = assert!(std::mem::size_of::<PeerMessage>() <= 1144, "PeerMessage > 1144 bytes"); | |||
| const _: () = assert!(std::mem::size_of::<PeerMessage>() <= 1500, "PeerMessage > 1500 bytes"); | |||
There was a problem hiding this comment.
Is there any comment or meaningful explanation in the PR where this was introduced? Given this is networking one guess would be something to do with MTU but I doubt that matters to us.
|
|
||
| // V2 -> V3: Post State Root | ||
| #[derive(BorshSerialize, BorshDeserialize, Clone, PartialEq, Eq, Debug)] | ||
| pub struct ShardChunkHeaderInnerV3 { |
There was a problem hiding this comment.
I spent way too much time looking for the difference ;)
1d66984 to
8f279b3
Compare
Yep, this is just a boilerplate code. I explicitly wanted to separate this from meaningful changes to make it easier to review. |
This PR introduces the following structs:
BlockHeaderV5along withBlockHeaderInnerRestV5. For now let's keepBlockHeaderInnerLiteas it is, later we can change that if we need more fields there.ShardChunkHeaderInnerV3ShardChunkV3For now those contain the same fields as the current versions. Later we will introduce post-state-root related fields there and also potentially remove some of the pre-state-root fields.
ShardChunkV3is the only exception withprev_outgoing_receiptsbeing replaced byoutgoing_receiptsas part of this PR.This PR is a groundwork for #9450.