EIP 4844 - blobhash opcode (part 1) #14757
Conversation
r0qs
added
has dependencies
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
4 times, most recently
from
01f3c80 to
367ca5f
Compare
nikola-matic
force-pushed
the
bump-evmone-and-evmc-to-11
branch
from
498252a to
9160451
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
2 times, most recently
from
35e367d to
873e62e
Compare
r0qs
changed the title
r0qs
changed the title
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
from
e65d8bc to
69122e0
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
from
69122e0 to
66e145f
Compare
cameel
force-pushed
the
bump-evmone-and-evmc-to-11
branch
3 times, most recently
from
93b8bfa to
d5119f7
Compare
cameel
force-pushed
the
eip-4844-add-blobhash
branch
from
66e145f to
9eb631b
Compare
There was a problem hiding this comment.
This is missing a changelog entry (we usually add a separate one for the opcode and for high-level support in Solidity). Also, the bits of docs from #14759 describing the opcode should really be in this PR.
Other than that I did not find any serious problems. Mostly small things, easy to tweak. I added a fixup to deal with some of it, the rest is in comments.
Since the evmone update is now merged, I rebased the PR on develop.
test/libsolidity/semanticTests/inlineAssembly/blobhash_empty.sol
Outdated
Show resolved
Hide resolved
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
3 times, most recently
from
c65588e to
08bbb84
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
from
1d6a91f to
9a7639e
Compare
There was a problem hiding this comment.
I looked through all the uses of the already existing blockhash() to make sure we're not forgetting anything we should have for blobhash() here and found only one minor test that might need it.
Other than that, we actually should change #14757 (comment) back (I'll post details in a moment). EDIT: Actually it's fine. I wanted to note that the variable assigned to the pointer will go out of scope, but it's static so it's fine.
I'm going to look through #14759 before approving this in case something from there should still end up in this PR, but if not, we'll be done here.
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
4 times, most recently
from
d762c6d to
1af000d
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
2 times, most recently
from
1f5e3b7 to
eabe457
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
2 times, most recently
from
b00b2cd to
2150f74
Compare
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
2 times, most recently
from
3cb794d to
9243997
Compare
Co-authored-by: Kamil Śliwak <kamil.sliwak@codepoets.it>
r0qs
force-pushed
the
eip-4844-add-blobhash
branch
from
9243997 to
269951e
Compare
| * Introduce global ``block.blobbasefee`` for retrieving the blob base fee of the current block. | ||
| * Yul: Introduce builtin ``blobbasefee()`` for retrieving the blob base fee of the current block. | ||
|
|
||
| * Yul: Introduce builtin ``blobhash()`` for retrieving versioned hashes of blobs associated with the transaction. |
There was a problem hiding this comment.
Oh, one more thing, let's finally stop messing with the spacing in the changelog :)
| * Yul: Introduce builtin ``blobhash()`` for retrieving versioned hashes of blobs associated with the transaction. | |
| * Yul: Introduce builtin ``blobhash()`` for retrieving versioned hashes of blobs associated with the transaction. | |
I fixed it earlier in all those PRs but it keeps reappearing again and again for some reason. Don't worry about the wrong spacing of the Compiler Features: section. We'll fix it when we sort the changelog. In PRs just add an entry for your PR and don't touch other lines.
Depends on #14743(merged)First part of #14740
This PR implements the blobhash opcode as specified in the EIP-4844 and add tests for assembler mode. The opcode is not exposed to Solidity in this part; that will be addressed in part 2. While it's intended to be non-breaking, additional tests may be required.