Change header encoding mechanism #300
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Remove changes to do with determining whether a decoded block was pre-gingerbread based on it's gas limit.
Previously we looked for a gas limit of zero, now we use an explicit unexported field set at creation time.
This reverts commit 3b7f4b2.
…r-encoding-mechanism
alecps
approved these changes
Dec 18, 2024
karlb
approved these changes
Dec 20, 2024
karlb
pushed a commit
that referenced
this pull request
Jan 27, 2025
This changes the way that we detect pre-gingerbread headers when encoding. Closes celo-org/celo-blockchain-planning#588 Previously we would interpret a `gasLimit` of zero to indicate pre-gingerbread headers, since prior to the gingerbread hardfork celo blocks lacked the `gasLimit` field. This however caused lots of problems with the go tests (existing and ones that continue to be added) that routinely use blocks where gasLimit has not been set. So now we detect pre-gingerbread blocks when decoding and set a flag on the struct that is used to select the encoding. New blocks are assumed to be post gingerbread, which is ok because it's not possible to run op-geth without gingerbread set, because we haven't ported across the pre-gingerbread state transition logic. I was able to revert some test modifications because of this but this didn't fix the tests that broke during the celo11 rebase (skipped in [this commit](d981efb)), it seems like those problems were caused by something else.
karlb
pushed a commit
that referenced
this pull request
Jan 28, 2025
This changes the way that we detect pre-gingerbread headers when encoding. Closes celo-org/celo-blockchain-planning#588 Previously we would interpret a `gasLimit` of zero to indicate pre-gingerbread headers, since prior to the gingerbread hardfork celo blocks lacked the `gasLimit` field. This however caused lots of problems with the go tests (existing and ones that continue to be added) that routinely use blocks where gasLimit has not been set. So now we detect pre-gingerbread blocks when decoding and set a flag on the struct that is used to select the encoding. New blocks are assumed to be post gingerbread, which is ok because it's not possible to run op-geth without gingerbread set, because we haven't ported across the pre-gingerbread state transition logic. I was able to revert some test modifications because of this but this didn't fix the tests that broke during the celo11 rebase (skipped in [this commit](d981efb)), it seems like those problems were caused by something else.
karlb
pushed a commit
that referenced
this pull request
Feb 3, 2025
This changes the way that we detect pre-gingerbread headers when encoding. Closes celo-org/celo-blockchain-planning#588 Previously we would interpret a `gasLimit` of zero to indicate pre-gingerbread headers, since prior to the gingerbread hardfork celo blocks lacked the `gasLimit` field. This however caused lots of problems with the go tests (existing and ones that continue to be added) that routinely use blocks where gasLimit has not been set. So now we detect pre-gingerbread blocks when decoding and set a flag on the struct that is used to select the encoding. New blocks are assumed to be post gingerbread, which is ok because it's not possible to run op-geth without gingerbread set, because we haven't ported across the pre-gingerbread state transition logic. I was able to revert some test modifications because of this but this didn't fix the tests that broke during the celo11 rebase (skipped in [this commit](d981efb)), it seems like those problems were caused by something else.
karlb
pushed a commit
that referenced
this pull request
Feb 11, 2025
This changes the way that we detect pre-gingerbread headers when encoding. Closes celo-org/celo-blockchain-planning#588 Previously we would interpret a `gasLimit` of zero to indicate pre-gingerbread headers, since prior to the gingerbread hardfork celo blocks lacked the `gasLimit` field. This however caused lots of problems with the go tests (existing and ones that continue to be added) that routinely use blocks where gasLimit has not been set. So now we detect pre-gingerbread blocks when decoding and set a flag on the struct that is used to select the encoding. New blocks are assumed to be post gingerbread, which is ok because it's not possible to run op-geth without gingerbread set, because we haven't ported across the pre-gingerbread state transition logic. I was able to revert some test modifications because of this but this didn't fix the tests that broke during the celo11 rebase (skipped in [this commit](d981efb)), it seems like those problems were caused by something else.
piersy
added a commit
that referenced
this pull request
Feb 12, 2025
When the celo blockchain launched it did so without `gasLimit`, `difficulty` and `uncleHash` fields on the celo block. In the gingerbread hardfork these fields were added to the block, to improve celo's compatibility with ethereum, and a zero (or not) `uncleHash` was used to distinguish the old and new block formats. When transitioning to the op-stack we discovered that these fields are set to default values during genesis init if they are unset. This broke our approach to encoding, so to prevent this happening we would only set defaults in a migrated chain context (where cel2 time is set an non zero), this allowed us to skip setting defaults for migrated chains but retain the defaults for tests which rely heavily on them. In the op-stack we were using a zero gas limit as an indication of what type of block encoding to use, but this presented it's own problems for some tests that used blocks without a gas limit set, so we subsequently switched (#300) to having an explicit flag set inside headers to specify which encoding to use. When loading blocks from leveldb the RLP structure could be introspected to determine the type of block that was being loaded, our thinking was that for all migrated blocks we could rely on this approach to setting the encoding, and for any new blocks they would all be being created after the cel2 fork and so would use the most recent encoding. We removed the special casing of default setting in `toBlockWithRoot` and relied on the explicit flag. Unfortunately we forgot about what happens at genesis init, where `NewBlock` is called to create the genesis block. With the special casing removed, the defaults being set and `NewBlock` always producing blocks that use the new encoding we discovered that initing the genesis on a migrated chain resulted in a different genesis block. To fix this the special casing has been re-introduced in `toBlockWithRoot` and extended to prevent default setting of `uncleHash` inside `NewBlock` when operating in the context of a migrated chain. Also if the genesis block is a pre-gingerbread block it is marked as such so that it is encoded correctly. There's was one remaining issue which is that `TestFeeHistory`, initialises a pre-gingerbread genesis but expects it to have the new block format, so to account for that an extra condition was added to only use the old format for pre-gingerbread blocks that are being initialised as a result of executing the initGenesis command. --------- Co-authored-by: Paul Lange <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This changes the way that we detect pre-gingerbread headers when encoding.
Closes https://github.com/celo-org/celo-blockchain-planning/issues/588
Previously we would interpret a
gasLimitof zero to indicate pre-gingerbread headers, since prior to the gingerbread hardfork celo blocks lacked thegasLimitfield.This however caused lots of problems with the go tests (existing and ones that continue to be added) that routinely use blocks where gasLimit has not been set.
So now we detect pre-gingerbread blocks when decoding and set a flag on the struct that is used to select the encoding.
New blocks are assumed to be post gingerbread, which is ok because it's not possible to run op-geth without gingerbread set, because we haven't ported across the pre-gingerbread state transition logic.
I was able to revert some test modifications because of this but this didn't fix the tests that broke during the celo11 rebase (skipped in this commit), it seems like those problems were caused by something else.