This PR introduces powerful code size optimizations. This is achieved
with an additional IR layer in between Yul and LLVM IR. It works quite
well (see below for a comparison) because it captures optimization gains
that neither solc nor LLVM can realize.
The feature is gated behind `--newyork` or a corresponding standard JSON
input setting and **off by default**.
### How it works
We implemented an SSA IR modelled after or inspired by the [MLIR SCF
dialect](https://mlir.llvm.org/docs/Dialects/SCFDialect/). Yul is
translated into this IR and the IR implements mainly the following
optimizations:
- **Type narrowing**: narrows 256-bit values to I1..I160 where provable
(bidirectional + interprocedural), so LLVM emits native 32/64-bit ops
instead
of software 256-bit arithmetic.
- **FMP range proof**: exposes the `mload(0x40) < heap_size` bound to
LLVM
IPSCCP, eliminating trailing overflow checks on FMP-derived offsets.
Single
largest code-size contributor.
- **Heap analysis**: skips big-endian byte-swaps where native
little-endian
layout is safe.
- **mem_opt / keccak fusion+folding**, **compound outlining** (mapping
sload/sstore), **guard narrowing**, **inlining** tuned for PolkaVM, and
**exact + fuzzy function dedup**.
### Results
The optimizations implemented with the newyork IR bring substantial code
size improvements.
- Integration contracts: approximately **-20% to -50%** code size.
- OpenZeppelin Wizard corpus: around **-47%** code size.
This table shows the code size improvements on various OpenZeppelin ERC
implementations, taken from their [contracts
wizard](https://wizard.openzeppelin.com/), with almost all features
enabled.
| Contract | Default: (bytes) | With `--newyork`: (bytes) | Reduction |
|---|---:|---:|---:|
| erc1155 | 59,206 | 30,649 | -48.2% |
| erc20 | 81,697 | 42,893 | -47.5% |
| erc721 | 92,052 | 49,423 | -46.3% |
| oz_gov | 141,861 | 81,159 | -42.8% |
| oz_rwa | 77,274 | 37,975 | -50.9% |
| oz_simple_erc20 | 29,950 | 16,554 | -44.7% |
| oz_stable | 82,117 | 39,032 | -52.5% |
| **Total** | **564,157** | **297,685** | **-47.2%** |
We believe this to be a reasonably representative dataset and conclude
that the newyork IR achieves drastic code size reductions on real world
dApp implementations.
### Methodology, security and runtime performance
Since the main contributors to these improvements are the FMP range
proofs and the type narrowing, we don't expect big runtime performance
regressions (it may even benefit runtime performance in some cases).
The PR also includes docs
(`book/src/developer_guide/newyork_{optimizer,ir}.md`) and many new
integration tests, including regression tests for known FMP soundness
traps.
The PR was authored collaboratively and reviewed extensively by the
`revive` maintainers, as well as Claude Opus, Claude Fable, Qwen,
Minimax and Deepseek LLMs, over a span of many months.
The `newyork` compiler pipeline introduces no new regressions over our
integration and differential test suites (containing over 30000 test
cases). This was achieved by careful manual reviews and many LLM bughunt
loops. Additionally, a final security review by Anthropic's Fable 5 LLM
found no remaining soundness issues. Nonetheless, like any brand new
software, please consider this an experimental feature upon release.
---------
Signed-off-by: xermicus <bigcyrill@hotmail.com>
Signed-off-by: xermicus <cyrill@parity.io>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: neo <me@90f9.org>
Co-authored-by: LJ <81748770+elle-j@users.noreply.github.com>
Existing test coverage didn't catch a signed remainder overflow
semantics edge case behavior where at least one the operators is being
constant folded by LLVM instead of solc. This PR fixes the SREM
translation bug and adds missing test coverage for division ops with
constants.
Closes#524
---------
Signed-off-by: xermicus <cyrill@parity.io>
This appears to be NFC change as new pass manager handles
--passes=default<O0> correctly. Still must be useful as a safeguard and
possibly for the old pass manager that is still used by backend.
```
Contract Without (bytes) With noinline+optnone (bytes)
─────────────────────────────── ───────────────── ────────────────────────────
contract.sol:C 1,266 1,266
Fibonacci:Iterative 2,506 2,506
Storage 3,671 3,671
Fibonacci:Recursive 3,794 3,794
Computation 5,776 5,776
Fibonacci:Binet 8,786 8,786
large_div_rem 10,705 10,705
DivisionArithmetics 14,831 14,831
ERC20 32,288 32,288
ERC20Tester 34,589 34,589
```
## Description
We previously had a workaround to strip `minsize` attributes for
functions with 256-bit integer division due to a crash in the LLVM
backend. We now [patch](https://github.com/paritytech/revive/pull/500)
LLVM with the fix, allowing us to use `minsize` with correct backend
optimization, making the workaround unnecessary.
In YUL it's possible to declare closure with a same name, but in
different scope, which previously resulted to func name conflict during
LLVM IR construction.
To solve this, track YUL scope when building LLVM IR and a special
counter that increments every time function with the same name exists in
LLVM module.
There's still assert if somehow within the same scope there 2 functions
with the same name, which should be forbidden by Yul
Fixes: #474
---------
Co-authored-by: LJ <81748770+elle-j@users.noreply.github.com>
- Remove the superfluous byte swap in the `linkersymbol` builtin
translation.
- Support testing the `linkersymbol` builtin translation in the runner
specs.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
## Description
Adds default values to the compiler arguments below when parsed by
`clap`. This does not change previous user-facing behavior as these
defaults were implemented using different logic, but improves the UX and
the visibility of the exact defaults used.
**Example output from `resolc --help`:**
(Only affected arguments are shown here, and some of their descriptions
have been omitted here.)
```
Options:
-O, --optimization <OPTIMIZATION>
Set the optimization parameter -O[0 | 1 | 2 | 3 | s | z]. Use `3` for best performance and `z` for minimal size
[default: z]
--metadata-hash <METADATA_HASH>
Set the metadata hash type. Available types: `none`, `ipfs`, `keccak256`
[default: keccak256]
--heap-size <HEAP_SIZE>
The emulated EVM linear heap memory static buffer size in bytes.
...
[default: 131072]
--stack-size <STACK_SIZE>
The contracts total stack size in bytes.
...
[default: 131072]
```
## Resolved Issues
Closes#446
The OOB logic is implemented in the runtime but truncation circumvented
it.
Closes#263
---------
Signed-off-by: xermicus <cyrill@parity.io>
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
- Emit the `call_evm` and `delegate_call_evm` syscalls for contract
calls.
- The call gas is no longer ignored.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Note:
- The unstable interface in `v2509.0.0` of `polkadot-sdk` is required.
- The differential test fails against EVM in `v2509.0.0` of
`polkadot-sdk`.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
This PR synchronizes the syscall function signatures in the runtime-api
with the current revive pallet API surface.
Signed-off-by: xermicus <cyrill@parity.io>
This PR switches to the new `get_storage_or_zero` and
`set_storage_or_clear` syscalls which are more tailored towards
Solidity.
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
That's a workaround to avoid LLVM backend crash while selecting
instruction for 256-bit integer division.
The workaround needs to be removed after we switch to newest inkwell
that has a fix in RISC-V backend
---------
Signed-off-by: kvp <mammal_windier8j@icloud.com>
Function symbols can clash as we compile multiple YUL `object`
definition into the same `LLVM` module.
- Disambiguate via unique function symbols based its location (runtime
or deploy code).
- Use `LinkOnceODR` linkage for compiler builtin helpers.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Prevents unused functions in the emitted ELF object. Drive-by add a
missing test case (which misses a relocation under `-Oz` when all
internal functions are marked as private).
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Separate between compilation and linker phases to allow deploy time
linking and back-porting era compiler changes to fix#91. Unlinked
contract binaries (caused by missing libraries or missing factory
dependencies in turn) are emitted as raw ELF object.
Few drive by fixes:
- #98
- A compiler panic on missing libraries definitions.
- Fixes some incosistent type forwarding in JSON output (empty string
vs. null object).
- Remove the unused fallback for size optimization setting.
- Remove the broken `--lvm-ir` mode.
- CI workflow fixes.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Signed-off-by: xermicus <bigcyrill@hotmail.com>
Signed-off-by: xermicus <cyrill@parity.io>
The data structure can be used to build the JSON input for `resolc` too.
In that case serializing of provided custom options should not be
dismissed.
Makes the memory settings struct more modular as a drive-by.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Exposes the following PolkaVM specific options via the standard json
interface:
- Heap size
- Stack size
- Whether to emit source level debug information
Additionally it is now forbidden to specify those as CLI option in
standard JSON mode. They are bytecode altering options and having
multiple ways to specify them creates unnecessary room for confusion:
The standard JSON input description should be sufficient and succint for
reproducible builds.
Closes#290
---------
Signed-off-by: xermicus <bigcyrill@hotmail.com>
So far if no optimization level was specified, optimizations for
execution time were applied. However, we currently are a bit limited on
code size. Add to that, this setting is not available in solc and people
generally ignore the docs, generating a lot of support requests.
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
- Allow configuration of the maximum heap and stack size via CLI flags
and JSON input settings.
- Increase the default value for the stack size to 32kb.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
This PR changes the implementation of the emulated EVM heap memory:
Instead of linking in a C implementation the code is emitted directly
into the contract module. Which allows making it configurable via a
compiler parameter (a follow up PR to this).
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Closes #48
Change the code size test to no longer emit debug info as to get a more
accurate picture.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
- Lazily load function arguments so that they can be passed as pointers.
- Lazily call the immutable store function to avoid storing zero sized
immutable data.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Storage keys and values are big endian. Keeping them LE was a pre-mature
optimization because for the contract itself it this is a no-op and thus
not observable. However we should consider the storage layout as part of
the contract ABI. The endianness of transient storage values are still
kept as-is.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
- Support for passing LLVM command line options via the prcoess input or
providing one or more `--llvm-arg='..'` resolc CLI flag. This allows
more fine-grained control over the LLVM backend configuration.
- Make LLVM initialization idempotent.
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
The differential testing framework will make a second consumer. There
seems to be no re-usable Rust crate for this. But we already have
everything here, just needs a small refactor to make it fully re-usable.
- Mostly decouple the solc JSON-input-output interface types from the
`solidity` frontend crate
- Expose the JSON-input-output interface types in a dedicated crate
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
- remove the __sha3 function symbol: this is provided by the pallet
- remove the storage address spaces: they are not mapped into memory
---------
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Co-authored-by: Alexander Theißen <alex.theissen@me.com>
- Add the revive runtime function interface to minimize boiler plate
code.
- Outline heavily repeated code into dedicated functions to bring down
code size.
- The code size tests builds optimized for size.
- Function attributes are passed as slices.
This significantly brings down the code size for all OpenZeppelin wizard
contracts (using all possible features) compiled against OpenZeppelin
`v5.0.0` with size optimizations.
|contract|| `-Oz` main | `-Oz` PR || `-O3` main | `-O3` PR |
|-|-|-|-|-|-|-|
|erc1155.sol||100K|67K||114K|147K|
|erc20.sol||120K|90K||160K|191K|
|erc721.sol||128K|101K||178K|214K|
|governor.sol||226K|165K||293K|349K|
|rwa.sol||116K|85K||154K|185K|
|stable.sol||116K|86K||155K|192K|
On the flip side this introduces a heavy penalty for cycle optimized
builds. Setting the no-inline attributes for cycle optimized builds
helps a lot but heavily penalizes runtime speed (LLVM does not yet
inline everything properly - to be investigated later on).
Next steps:
- Modularize more functions
- Refactor the YUL function arguments to use pointers instead of values
- Afterwards check if LLVM still has trouble inline-ing properly on O3
or set the no-inline attribute if it does not penalize runtime
performance too bad.
