Disable large contract test on firefox

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>

.cargo/config.toml

@@ -0,0 +1,17 @@
+[target.wasm32-unknown-emscripten]
+rustflags = [
+    "-Clink-arg=-sEXPORTED_FUNCTIONS=_main,_free,_malloc",
+    "-Clink-arg=-sNO_INVOKE_RUN=1",
+    "-Clink-arg=-sEXIT_RUNTIME=1",
+    "-Clink-arg=-sALLOW_MEMORY_GROWTH=1",
+    "-Clink-arg=-sEXPORTED_RUNTIME_METHODS=FS,callMain,stringToNewUTF8",
+    "-Clink-arg=-sMODULARIZE=1",
+    "-Clink-arg=-sEXPORT_NAME=createRevive",
+    "-Clink-arg=-sWASM_ASYNC_COMPILATION=0",
+    "-Clink-arg=-sDYNAMIC_EXECUTION=0",
+    "-Clink-arg=-sALLOW_TABLE_GROWTH=1",
+    "-Clink-arg=--js-library=js/embed/soljson_interface.js",
+    "-Clink-arg=--pre-js=js/embed/pre.js",
+    "-Clink-arg=-sSTACK_SIZE=128kb",
+    "-Clink-arg=-sNODEJS_CATCH_EXIT=0"
+]

.dockerignore

@@ -0,0 +1,7 @@
+emsdk
+llvm*/
+target-llvm/
+target/
+node_modules/
+utils/
+build/

.github/workflows/build-revive-debian.yml

@@ -1,38 +0,0 @@
-name: Build revive-debian
-on:
-  workflow_dispatch:
-
-env:
-  REVIVE_DEBIAN_PACKAGE: revive-debian-x86
-  DEBIAN_CONTAINER: revive-builder-debian-x86
-  DEBIAN_CONTAINER_BUILDER: build-debian-builder.sh
-  DEBIAN_CONTAINER_RUNNER: run-debian-builder.sh
-  REVIVE_DEBIAN_INSTALL: ${{ github.workspace }}/target/release
-  REVIVE_DEBIAN_BINARY: resolc
-  RUST_VERSION: "1.80"
-
-jobs:
-  build-revive-debian-x86:
-    name: debian-container-x86
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-
-      - name: build-container
-        run: |
-          (cd utils && ./${{ env.DEBIAN_CONTAINER_BUILDER}} --build-arg RUST_VERSION=${{ env.RUST_VERSION}} . )
-
-      - name: build-revive-debian
-        run: |
-          rustup show
-          cargo --version
-          rustup +nightly show
-          cargo +nightly --version
-          bash --version
-          utils/${{ env.DEBIAN_CONTAINER_RUNNER }} utils/build-revive.sh -o ${{ env.REVIVE_DEBIAN_INSTALL}}
-
-      - uses: actions/upload-artifact@v4
-        with:
-          name: ${{ env.REVIVE_DEBIAN_PACKAGE }}
-          path: ${{ env.REVIVE_DEBIAN_INSTALL }}/${{ env.REVIVE_DEBIAN_BINARY }}
-          retention-days: 1

.github/workflows/build-revive-wasm.yml

@@ -0,0 +1,143 @@
+name: Build revive-wasm
+on:
+  push:
+    branches: ["main"]
+  pull_request:
+    branches: ["main"]
+  workflow_dispatch:
+
+env:
+  CARGO_TERM_COLOR: always
+  REVIVE_WASM_INSTALL_DIR: ${{ github.workspace }}/target/wasm32-unknown-emscripten/release
+  BUN_VERSION: 1.1.43
+
+jobs:
+  build-revive-wasm:
+    runs-on: parity-large
+    defaults:
+      run:
+        shell: bash
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Cache LLVM build
+        id: cache-llvm
+        uses: actions/cache@v3
+        with:
+          path: |
+            target-llvm/emscripten/target-final
+            target-llvm/gnu/target-final
+          # Use a unique key based on LLVM version or configuration files to avoid cache invalidation
+          key: llvm-build-${{ runner.os }}-${{ hashFiles('LLVM.lock', 'Cargo.toml', 'Cargo.lock', 'crates/llvm-builder/**', '.github/workflows/build-revive-wasm.yml') }}
+
+      - name: Install system dependencies
+        run: |
+          sudo apt-get update && sudo apt-get install -y cmake ninja-build curl git libssl-dev pkg-config clang lld
+
+      - name: Install Rust stable toolchain
+        uses: actions-rs/toolchain@v1
+        with:
+          profile: minimal
+          toolchain: stable
+          components: rust-src
+          target: wasm32-unknown-emscripten
+
+      - name: Install LLVM build dependencies
+        run: |
+          make install-llvm-builder
+          revive-llvm --target-env emscripten clone
+
+      - name: Setup revive environment variables
+        run: |
+          echo "LLVM_SYS_181_PREFIX=$(pwd)/target-llvm/gnu/target-final" >> $GITHUB_ENV
+          echo "REVIVE_LLVM_TARGET_PREFIX=$(pwd)/target-llvm/emscripten/target-final" >> $GITHUB_ENV
+
+      - name: Build host LLVM
+        if: steps.cache-llvm.outputs.cache-hit != 'true'
+        run: |
+          revive-llvm build --llvm-projects lld --llvm-projects clang
+
+      - name: Build target LLVM
+        if: steps.cache-llvm.outputs.cache-hit != 'true'
+        run: |
+          source emsdk/emsdk_env.sh
+          revive-llvm --target-env emscripten build --llvm-projects lld
+
+      - run: |
+          rustup show
+          cargo --version
+          rustup +nightly show
+          cargo +nightly --version
+          cmake --version
+          bash --version
+
+      - name: Use Cached LLVM
+        if: steps.cache-llvm.outputs.cache-hit == 'true'
+        run: |
+          echo "Using cached LLVM"
+
+      - name: Build revive
+        run: |
+          source emsdk/emsdk_env.sh
+          make install-wasm
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: revive-wasm
+          path: |
+            ${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.js
+            ${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.wasm
+            ${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc_web.js
+          retention-days: 1
+
+  test-revive-wasm:
+    needs: build-revive-wasm
+    strategy:
+      matrix:
+        os: ["ubuntu-24.04", "macos-14", "windows-2022"]
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Create Target Directory
+        run: mkdir -p ${{ env.REVIVE_WASM_INSTALL_DIR }}
+
+      - name: Download Artifact
+        uses: actions/download-artifact@v4
+        with:
+          name: revive-wasm
+          path: ${{ env.REVIVE_WASM_INSTALL_DIR }}
+
+      - name: Set Up Node.js
+        uses: actions/setup-node@v3
+        with:
+          node-version: "20"
+
+      - name: Install Bun on Windows
+        if: runner.os == 'Windows'
+        run: |
+          Set-ExecutionPolicy RemoteSigned -Scope CurrentUser
+          iex (new-object net.webclient).downloadstring('https://get.scoop.sh')
+          scoop install bun@${{ env.BUN_VERSION }}
+          scoop install wget
+          Join-Path (Resolve-Path ~).Path "scoop\shims" >> $Env:GITHUB_PATH
+
+      - name: Install Bun on macOS and Linux
+        if: runner.os != 'Windows'
+        run: |
+          curl -fsSL https://bun.sh/install | bash -s bun-v${{ env.BUN_VERSION }}
+          echo "$HOME/.bun/bin" >> $GITHUB_PATH
+
+      - name: Install packages
+        run: npm install
+
+      - name: Run Playwright tests
+        run: |
+          cd js
+          npx playwright install --with-deps
+          npx playwright test
+
+      - name: Test revive
+        run: |
+          echo "Running tests for ${{ matrix.os }}"
+          npm run test:wasm

.github/workflows/rust.yml

@@ -11,7 +11,7 @@ env:
 
 jobs:
   build-ubuntu-x86:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-24.04
 
     steps:
       - uses: actions/checkout@v4
@@ -19,22 +19,26 @@ jobs:
       - name: Install solc
         run: |
           mkdir -p solc
-          curl -sSL --output solc/solc https://github.com/ethereum/solidity/releases/download/v0.8.27/solc-static-linux
+          curl -sSL --output solc/solc https://github.com/ethereum/solidity/releases/download/v0.8.28/solc-static-linux
           chmod +x solc/solc
           echo "$(pwd)/solc/" >> $GITHUB_PATH
 
       - name: Install LLVM
         run: |
-          curl -sSL --output llvm.tar.xz https://github.com/llvm/llvm-project/releases/download/llvmorg-18.1.4/clang+llvm-18.1.4-x86_64-linux-gnu-ubuntu-18.04.tar.xz
+          curl -sSL --output llvm.tar.xz https://github.com/paritytech/revive/releases/download/v0.1.0-dev.7/clang+llvm-18.1.8-x86_64-linux-gnu-ubuntu-24.04.tar.xz
-          tar Jxf llvm.tar.xz
+          mkdir llvm18
-          mv clang+llvm-18.1.4-x86_64-linux-gnu-ubuntu-18.04 llvm18/
+          tar Jxf llvm.tar.xz -C llvm18/
-          echo "$(pwd)/llvm18/bin" >> $GITHUB_PATH
+          echo "LLVM_SYS_181_PREFIX=$(pwd)/llvm18" >> $GITHUB_ENV
 
-      - name: Install apt dependencies
+      - name: Install geth
         run: |
-          sudo add-apt-repository -y ppa:ethereum/ethereum
+          git clone https://github.com/xermicus/go-ethereum --branch=cl/fix-runner-state-dump --depth=1
-          sudo apt update
+          cd go-ethereum
-          sudo apt install -y libtinfo5 ethereum
+          make all
+          echo "$(pwd)/build/bin/" >> $GITHUB_PATH
+
+      - name: Machete
+        uses: bnjbvr/cargo-machete@main
 
       - name: Format
         run: make format

.gitignore

@@ -1,4 +1,5 @@
 /target
+target-llvm
 *.dot
 .vscode/
 .DS_Store
@@ -6,11 +7,15 @@
 /*.yul
 /*.ll
 /*.s
-/llvm-project
+/llvm*
-/llvm18.0
 node_modules
 artifacts
 tmp
 package-lock.json
 /*.html
 /build
+soljson.js
+test-results
+playwright-report
+.cache
+emsdk

CHANGELOG.md

@@ -0,0 +1,120 @@
+# Changelog
+
+## Unreleased
+
+This is a development pre-release.
+
+Supported `polkadot-sdk` rev: `274a781e8ca1a9432c7ec87593bd93214abbff50`
+
+### Added
+- Support for the `coinbase` opcode.
+
+### Changed 
+- Missing the `--overwrite` flag emits an error instead of a warning.
+- The `resolc` executable prints the help by default.
+- Removed support for legacy EVM assembly (EVMLA) translation.
+- integration: identify cached code blobs on source code to fix potential confusions.
+- Setting base, include or allow paths in emscripten is now a hard error.
+- Employ a heuristic to detect `address.transfer` and `address.send` calls.
+  If detected, the re-entrant call flag is not set and 0 deposit limit is endowed.
+
+### Fixed
+- Solidity: Add the solc `--libraries` files to sources.
+- A data race in tests.
+- Fix `broken pipe` errors.
+- llvm-builder: Allow warnings.
+- solidity: Fix the custom compiler warning messages.
+
+## v0.1.0-dev.9
+
+This is a development pre-release.
+
+### Added
+
+### Changed 
+- Syscalls with more than 6 arguments now pack them into registers.
+
+### Fixed
+- Remove reloading of the resolc.js file (fix issue with relative path in web worker)
+
+## v0.1.0-dev.8
+
+This is a development pre-release.
+
+### Added
+- The `revive-llvm-builder` crate with the `revive-llvm` helper utility for streamlined management of the LLVM framework dependency.
+- Initial support for running `resolc` in the browser.
+
+### Changed
+- Suported contracts runtime is polkadot-sdk git version `d62a90c8c729acd98c7e9a5cab9803b8b211ffc5`.
+- The minimum supported Rust version is `1.81.0`.
+- Error out early instead of invoking `solc` with invalid base or include path flags.
+
+### Fixed
+- Decouple the LLVM target dependency from the LLVM host dependency.
+- Do not error out if no files and no errors were produced. This aligns resolc closer to solc.
+- Fixes input normalization in the Wasm version.
+
+## v0.1.0-dev.7
+
+This is a development pre-release.
+
+### Added
+- Implement the `GASPRICE` opcode.
+- Implement the `BASEFEE` opcode.
+- Implement the `GASLIMIT` opcode.
+
+### Changed
+- The `GAS` opcode now returns the remaining `ref_time`.
+- Contracts can now be supplied call data input of arbitrary size.
+- Some syscalls now return the value in a register, slightly improving  emitted contract code.
+- Calls forward maximum weight limits instead of 0, anticipating a change in polkadot-sdk where weight limits of 0 no longer interprets as uncapped limit.
+
+### Fixed
+- A linker bug which was preventing certain contracts from linking with the PVM linker. 
+- JS: Fix encoding conversion from JS string (UTF-16) to UTF-8.
+- The git commit hash slug is always displayed in the version string. 
+
+## v0.1.0-dev.6
+
+This is a development pre-release.
+
+# Added
+- Implement the `BLOCKHASH` opcode.
+- Implement delegate calls.
+- Implement the `GASPRICE` opcode. Currently hard-coded to return `1`.
+- The ELF shared object contract artifact is dumped into the debug output directory.
+- Initial support for emitting debug info (opt in via the `-g` flag)
+
+# Changed
+- resolc now emits 64bit PolkaVM blobs, reducing contract code size and execution time.
+- The RISC-V bit-manipulation target feature (`zbb`) is enabled.
+
+# Fixed
+- Compilation to Wasm (for usage in node and web browsers)
+
+
+## v0.1.0-dev.5
+
+This is development pre-release.
+
+# Added
+- Implement the `CODESIZE` and `EXTCODESIZE` opcodes.
+
+# Changed
+- Include the full revive version in the contract metadata.
+
+# Fixed
+
+## v0.1.0-dev-4
+
+This is development pre-release.
+
+# Added
+- Support the `ORIGIN` opcode.
+
+# Changed
+- Update polkavm to `v0.14.0`.
+- Enable the `a`, `fast-unaligned-access` and `xtheadcondmov` LLVM target features, decreasing the code size for some contracts.
+
+# Fixed

Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["crates/*"]
 
 [workspace.package]
-version = "0.1.0"
+version = "0.1.0-dev.9"
 authors = [
     "Cyrill Leutwiler <cyrill@parity.io>",
     "Parity Technologies <admin@parity.io>",
@@ -11,67 +11,73 @@ authors = [
 license = "MIT/Apache-2.0"
 edition = "2021"
 repository = "https://github.com/paritytech/revive"
-rust-version = "1.80.0"
+rust-version = "1.81.0"
 
 [workspace.dependencies]
-revive-benchmarks = { version = "0.1.0", path = "crates/benchmarks" }
+revive-benchmarks = { version = "0.1.0-dev.9", path = "crates/benchmarks" }
-revive-builtins = { version = "0.1.0", path = "crates/builtins" }
+revive-builtins = { version = "0.1.0-dev.9", path = "crates/builtins" }
-revive-common = { version = "0.1.0", path = "crates/common" }
+revive-common = { version = "0.1.0-dev.9", path = "crates/common" }
-revive-differential = { version = "0.1.0", path = "crates/differential" }
+revive-differential = { version = "0.1.0-dev.9", path = "crates/differential" }
-revive-integration = { version = "0.1.0", path = "crates/integration" }
+revive-integration = { version = "0.1.0-dev.9", path = "crates/integration" }
-revive-linker = { version = "0.1.0", path = "crates/linker" }
+revive-linker = { version = "0.1.0-dev.9", path = "crates/linker" }
-lld-sys = { version = "0.1.0", path = "crates/lld-sys" }
+lld-sys = { version = "0.1.0-dev.9", path = "crates/lld-sys" }
-revive-llvm-context = { version = "0.1.0", path = "crates/llvm-context" }
+revive-llvm-context = { version = "0.1.0-dev.9", path = "crates/llvm-context" }
-revive-runtime-api = { version = "0.1.0", path = "crates/runtime-api" }
+revive-runtime-api = { version = "0.1.0-dev.9", path = "crates/runtime-api" }
-revive-runner = { version = "0.1.0", path = "crates/runner" }
+revive-runner = { version = "0.1.0-dev.9", path = "crates/runner" }
-revive-solidity = { version = "0.1.0", path = "crates/solidity" }
+revive-solidity = { version = "0.1.0-dev.9", path = "crates/solidity" }
-revive-stdlib = { version = "0.1.0", path = "crates/stdlib" }
+revive-stdlib = { version = "0.1.0-dev.9", path = "crates/stdlib" }
+revive-build-utils = { version = "0.1.0-dev.9", path = "crates/build-utils" }
 
-hex = "0.4"
+hex = "0.4.3"
-petgraph = "0.6"
 cc = "1.0"
-libc = "0.2"
+libc = "0.2.169"
 tempfile = "3.8"
 anyhow = "1.0"
-semver = { version = "1.0", features = [ "serde" ] }
+semver = { version = "1.0", features = ["serde"] }
-itertools = "0.12"
+itertools = "0.14"
-serde = { version = "1.0", features = [ "derive" ] }
+serde = { version = "1.0", features = ["derive"] }
-serde_json = { version = "1.0", features = [ "arbitrary_precision" ] }
+serde_json = { version = "1.0", features = ["arbitrary_precision"] }
 regex = "1.10"
 once_cell = "1.19"
-num = "0.4"
+num = "0.4.3"
 sha1 = "0.10"
-sha2 = "0.10"
 sha3 = "0.10"
-md5 = "0.7"
+thiserror = "2.0"
-colored = "2.1"
+which = "7.0"
-thiserror = "1.0"
-which = "5.0"
 path-slash = "0.2"
 rayon = "1.8"
 clap = { version = "4", default-features = false, features = ["derive"] }
-rand = "0.8"
+polkavm-common = "0.19.0"
-polkavm-common = "0.10"
+polkavm-linker = "0.19.0"
-polkavm-linker = "0.10"
+polkavm-disassembler = "0.19.0"
-polkavm-disassembler = "0.10"
+polkavm = "0.19.0"
-polkavm = "0.10"
+alloy-primitives = { version = "0.8.19", features = ["serde"] }
-alloy-primitives = { version = "0.8", features = ["serde"] }
+alloy-sol-types = "0.8.19"
-alloy-sol-types = "0.8"
+alloy-genesis = "0.11.0"
-alloy-genesis = "0.3"
+alloy-serde = "0.11.0"
-alloy-serde = "0.3"
+env_logger = { version = "0.11.6", default-features = false }
-env_logger = { version = "0.10.0", default-features = false }
+serde_stacker = "0.1.11"
-serde_stacker = "0.1"
+criterion = { version = "0.5.1", features = ["html_reports"] }
-criterion = { version = "0.5", features = ["html_reports"] }
+log = { version = "0.4.25" }
-log = { version = "0.4" }
+git2 = { version = "0.20.0", default-features = false }
+downloader = "0.2.8"
+flate2 = "1.0.35"
+fs_extra = "1.3.0"
+num_cpus = "1"
+tar = "0.4.43"
+toml = "0.8.19"
+assert_cmd = "2.0.16"
+assert_fs = "1.1.2"
 
 # polkadot-sdk and friends
 codec = { version = "3.6.12", default-features = false, package = "parity-scale-codec" }
-scale-info = { version = "2.11.1", default-features = false }
+scale-info = { version = "2.11.6", default-features = false }
-polkadot-sdk = { git = "https://github.com/paritytech/polkadot-sdk", rev = "c77095f51119d2eccdc54d2f3518bed0ffbd6d53" }
+polkadot-sdk = { git = "https://github.com/paritytech/polkadot-sdk", rev = "274a781e8ca1a9432c7ec87593bd93214abbff50" }
 
 # llvm
 [workspace.dependencies.inkwell]
-version = "0.5"
+git = "https://github.com/TheDan64/inkwell.git"
+rev = "7b410298b6a93450adaa90b1841d5805a3038f12"
 default-features = false
 features = ["serde", "llvm18-0", "no-libffi-linking", "target-riscv"]
 

Dockerfile

@@ -0,0 +1,32 @@
+FROM rust:1.84.0 AS llvm-builder
+WORKDIR /opt/revive
+
+RUN apt update && \ 
+    apt upgrade -y && \
+    apt install -y cmake ninja-build curl git libssl-dev pkg-config clang lld musl
+
+COPY . .
+
+RUN make install-llvm-builder
+RUN revive-llvm --target-env musl clone
+RUN revive-llvm --target-env musl build  --llvm-projects lld --llvm-projects clang
+
+FROM messense/rust-musl-cross:x86_64-musl AS resolc-builder
+WORKDIR /opt/revive
+
+RUN apt update && \ 
+    apt upgrade -y && \
+    apt install -y pkg-config
+
+COPY . .
+COPY --from=llvm-builder /opt/revive/target-llvm /opt/revive/target-llvm
+
+ENV LLVM_SYS_181_PREFIX=/opt/revive/target-llvm/musl/target-final
+RUN make install-bin
+
+FROM alpine:latest
+ADD https://github.com/ethereum/solidity/releases/download/v0.8.28/solc-static-linux /usr/bin/solc
+COPY --from=resolc-builder /root/.cargo/bin/resolc /usr/bin/resolc
+
+RUN apk add --no-cache libc6-compat
+RUN chmod +x /usr/bin/solc 

LLVM.lock

@@ -0,0 +1,3 @@
+url = "https://github.com/llvm/llvm-project.git"
+branch = "release/18.x"
+ref = "3b5b5c1ec4a3095ab096dd780e84d7ab81f3d7ff"

Makefile

@@ -1,4 +1,24 @@
-.PHONY: install format test test-solidity test-cli test-integration test-workspace clean docs docs-build
+.PHONY: \
+	install \
+	install-bin \
+	install-npm \
+	install-wasm \
+	install-llvm-builder \
+	install-llvm \
+	format \
+	clippy \
+	machete \
+	test \
+	test-integration \
+	test-solidity \
+	test-workspace \
+	test-cli \
+	test-wasm \
+	test-llvm-builder
+	bench \
+	bench-pvm \
+	bench-evm \
+	clean
 
 install: install-bin install-npm
 
@@ -8,18 +28,28 @@ install-bin:
 install-npm:
 	npm install && npm fund
 
-# install-revive: Build and install to the directory specified in REVIVE_INSTALL_DIR
+install-wasm: install-npm
-ifeq ($(origin REVIVE_INSTALL_DIR), undefined)
+	cargo build --target wasm32-unknown-emscripten -p revive-solidity --release --no-default-features
-REVIVE_INSTALL_DIR=`pwd`/release/revive-debian
+	npm run build:package
-endif
+
-install-revive:
+install-llvm-builder:
-	cargo install --path crates/solidity --root $(REVIVE_INSTALL_DIR)
+	cargo install --path crates/llvm-builder
+
+install-llvm: install-llvm-builder
+	revive-llvm clone
+	revive-llvm build --llvm-projects lld --llvm-projects clang
 
 format:
 	cargo fmt --all --check
 
-test: format clippy test-cli test-workspace
+clippy:
-	cargo test --workspace
+	cargo clippy --all-features --workspace --tests --benches -- --deny warnings
+
+machete:
+	cargo install cargo-machete
+	cargo machete
+
+test: format clippy machete test-cli test-workspace
 
 test-integration: install-bin
 	cargo test --package revive-integration
@@ -28,11 +58,22 @@ test-solidity: install
 	cargo test --package revive-solidity
 
 test-workspace: install
-	cargo test --workspace
+	cargo test --workspace --exclude revive-llvm-builder
 
 test-cli: install
 	npm run test:cli
 
+test-wasm: install-wasm
+	npm run test:wasm
+
+test-llvm-builder:
+	@echo "warning: the llvm-builder tests will take many hours"
+	cargo test --package revive-llvm-builder -- --test-threads=1
+
+bench: install-bin
+	cargo criterion --all --all-features --message-format=json \
+	| criterion-table > crates/benchmarks/BENCHMARKS.md
+
 bench-pvm: install-bin
 	cargo criterion --bench execute --features bench-pvm-interpreter --message-format=json \
 	| criterion-table > crates/benchmarks/PVM.md
@@ -41,22 +82,13 @@ bench-evm: install-bin
 	cargo criterion --bench execute --features bench-evm --message-format=json \
 	| criterion-table > crates/benchmarks/EVM.md
 
-bench: install-bin
-	cargo criterion --all --all-features --message-format=json \
-	| criterion-table > crates/benchmarks/BENCHMARKS.md
-
-clippy:
-	cargo clippy --all-features --workspace --tests --benches
-
-docs: docs-build
-	mdbook serve --open docs/
-
-docs-build:
-	mdbook test docs/ && mdbook build docs/
-
 clean:
 	cargo clean ; \
+	revive-llvm clean ; \
 	rm -rf node_modules ; \
 	rm -rf crates/solidity/src/tests/cli-tests/artifacts ; \
 	cargo uninstall revive-solidity ; \
-	rm -f package-lock.json
+	cargo uninstall revive-llvm-builder ; \
+	rm -f package-lock.json ; \
+	rm -rf js/dist ; \
+	rm -f js/src/resolc.{wasm,js}

README.md

@@ -1,14 +1,15 @@
 ![CI](https://github.com/paritytech/revive/actions/workflows/rust.yml/badge.svg)
+[![Docs](https://img.shields.io/badge/Docs-contracts.polkadot.io-brightgreen.svg)](https://contracts.polkadot.io)
 
 # revive
 
-YUL and EVM assembly recompiler to LLVM, targetting RISC-V on [PolkaVM](https://github.com/koute/polkavm).
+YUL recompiler to LLVM, targetting RISC-V on [PolkaVM](https://github.com/koute/polkavm).
 
-[Frontend](https://github.com/matter-labs/era-compiler-solidity) and [code generator](https://github.com/matter-labs/era-compiler-llvm-context) are based of ZKSync `zksolc`.
+Visit [contracts.polkadot.io](https://contracts.polkadot.io) to learn more about contracts on Polkadot!
 
 ## Status
 
-This is experimental software in active development and not ready just yet for production usage.
+This is experimental software in active development and not ready just yet for production usage. Please do report any compiler related issues or missing features that are [not yet known to us](https://contracts.polkadot.io/known_issues/) here.
 
 Discussion around the development is hosted on the [Polkadot Forum](https://forum.polkadot.network/t/contracts-update-solidity-on-polkavm/6949#a-new-solidity-compiler-1).
 
@@ -16,18 +17,53 @@ Discussion around the development is hosted on the [Polkadot Forum](https://foru
 
 `resolc` depends on the [solc](https://github.com/ethereum/solidity) binary installed on your system.
 
+Download and install the `resolc` frontend executable for your platform from our [releases](https://github.com/paritytech/revive/releases).
+
+## Building from source
+
+Building revive requires a [stable Rust installation](https://rustup.rs/) and a C++ toolchain for building [LLVM](https://github.com/llvm/llvm-project) on your system.
+
+### LLVM
+
+`revive` depends on a custom build of LLVM `v18.1.8` with the RISC-V _embedded_ target, including the `compiler-rt` builtins. Use the provided [revive-llvm](crates/llvm-builder/README.md) utility to compile a compatible LLVM build locally and point `$LLVM_SYS_181_PREFIX` to the installation afterwards.
+
+The `Makefile` provides a shortcut target to obtain a compatible LLVM build:
+
+```bash
+make install-llvm
+export LLVM_SYS_181_PREFIX=${PWD}/target-llvm/gnu/target-final
+```
+
+### The `resolc` Solidity frontend
+
+To build the `resolc` Solidity frontend executable, make sure you have obtained a compatible LLVM build using [revive-llvm](crates/llvm-builder/README.md) and did export the `LLVM_SYS_181_PREFIX` environment variable pointing to it (see [above](#LLVM)).
+
 To install the `resolc` Solidity frontend executable:
 
 ```bash
-bash build-llvm.sh
-export PATH=${PWD}/llvm18.0/bin:$PATH
 make install-bin
 resolc --version
 ```
 
-### LLVM
+### Cross-compilation to Wasm
 
-`revive` requires a build of LLVM 18.1.4 or later including `compiler-rt`. Use the provided [build-llvm.sh](build-llvm.sh) build script to compile a compatible LLVM build locally in `$PWD/llvm18.0` (don't forget to add that to `$PATH` afterwards). 
+Cross-compile the `resolc.js` frontend executable to Wasm for running it in a Node.js or browser environment. The `REVIVE_LLVM_TARGET_PREFIX` environment variable is used to control the target environment LLVM dependency.
+
+```bash
+# Build the host LLVM dependency with PolkaVM target support
+make install-llvm
+export LLVM_SYS_181_PREFIX=${PWD}/target-llvm/gnu/target-final
+
+# Build the target LLVM dependency with PolkaVM target support
+revive-llvm --target-env emscripten clone
+source emsdk/emsdk_env.sh
+revive-llvm --target-env emscripten build --llvm-projects lld
+export REVIVE_LLVM_TARGET_PREFIX=${PWD}/target-llvm/emscripten/target-final
+
+# Build the resolc frontend executable
+make install-wasm
+make test-wasm
+```
 
 ### Development
 
@@ -35,4 +71,15 @@ Please consult the [Makefile](Makefile) targets to learn how to run tests and be
 Ensure that your branch passes `make test` locally when submitting a pull request.
 
 ## Design overview
+
 `revive` uses [solc](https://github.com/ethereum/solidity/), the Ethereum Solidity compiler, as the [Solidity frontend](crates/solidity/src/lib.rs) to process smart contracts written in Solidity. The YUL IR code (or legacy EVM assembly as a fallback for older `solc` versions) emitted by `solc` is then translated to LLVM IR, targetting [Polkadots `revive` pallet](https://docs.rs/pallet-revive/latest/pallet_revive/trait.SyscallDoc.html).
+[Frontend](https://github.com/matter-labs/era-compiler-solidity) and [code generator](https://github.com/matter-labs/era-compiler-llvm-context) are based of ZKSync `zksolc`.
+
+## Tests
+
+Before running the tests, ensure that Geth (Go Ethereum) is installed on your system. Follow the installation guide here: [Installing Geth](https://geth.ethereum.org/docs/getting-started/installing-geth).
+Once Geth is installed, you can run the tests using the following command:
+
+```bash
+make test
+```

RELEASE.md

@@ -0,0 +1,11 @@
+# Release checklist
+
+Prior to the first stable release we neither have formal release processes nor do we follow a fixed release schedule.
+
+To create a new pre-release:
+
+1. Merge a release PR which updates the `-dev.X` versions in the workspace `Cargo.toml` and updates the `CHANGELOG.md` accordingly
+2. Push a release tag to `main`
+3. Create a __pre-release__ from the tag and manually upload the `resolc` binary from docker image
+4. Manually upload `resolc.js` and `resolc.wasm` from the `build-revive-wasm` action artifacts.
+5. Update the [contract-docs](https://github.com/paritytech/contract-docs/) accordingly

SECURITY.md

@@ -0,0 +1 @@
+https://github.com/paritytech/polkadot-sdk/blob/master/docs/contributor/SECURITY.md

STATUS.md

@@ -0,0 +1,30 @@
+# Known issues
+
+The following is known and we are either working on it or it is a hard limitation. Please do not open a new issue.
+
+## Release
+
+`0.1.0-dev-2`
+
+## Missing features
+
+- [Libraries with public functions are not supported](https://github.com/paritytech/revive/issues/91)
+- [Automatic import resolution is not supported](https://github.com/paritytech/revive/issues/98)
+- The emulated EVM linear contract memory is limited to 64kb in size. Will be fixed with support for metered dynamic memory.
+- [The contract calldata is currently limited to 1kb in size](https://github.com/paritytech/revive/issues/57)
+- [EIP-4844 opcodes are not supported](https://github.com/paritytech/revive/issues/64)
+- [Delegate calls are not supported](https://github.com/paritytech/revive/issues/67)
+- [The `blockhash` opcode is not supported](https://github.com/paritytech/revive/issues/61)
+- [The `extcodesize` opcode is not supported](https://github.com/paritytech/revive/issues/58)
+- [The `origin` opcode is not supported](https://github.com/paritytech/revive/issues/59)
+- [Gas limits for contract calls are ignored](https://github.com/paritytech/revive/issues/60)
+- [Gas related opcodes are not supported](https://github.com/paritytech/revive/issues/60)
+- IPFS metadata hashes are not supported
+- [Compiled contract artifacts can exceed the pallet static memory limit and fail to deploy](https://github.com/paritytech/revive/issues/96).
+- [Transfers to inexistant accounts will fail if the transferred value lies below the ED.](https://github.com/paritytech/revive/issues/83) Will be fixed in the pallet to make the ED completely transparent for contracts.
+
+## Wontfix
+
+Please consult our documentation to learn more about Solidity and EVM features likely to remain unsupported (and why they will not be supported).
+
+TODO: Insert link to the relevant documentation section.

build-llvm.sh

@@ -1,92 +0,0 @@
-#!/usr/bin/env bash
-
-set -euo pipefail
-
-INSTALL_DIR="${PWD}/llvm18.0"
-mkdir -p ${INSTALL_DIR}
-
-
-# Clone LLVM 18 (any revision after commit bd32aaa is supposed to work)
-if [ ! -d "llvm-project" ]; then
-  git clone --depth 1 --branch release/18.x https://github.com/llvm/llvm-project.git
-fi
-
-
-# Build LLVM, clang
-LLVM_SRC_PREFIX=${PWD}/llvm-project
-LLVM_SRC_DIR=${LLVM_SRC_PREFIX}/llvm
-LLVM_BUILD_DIR=${PWD}/build/llvm
-if [ ! -d ${LLVM_BUILD_DIR} ] ; then
-	mkdir -p ${LLVM_BUILD_DIR}
-fi
-
-cmake -G Ninja \
-  -S ${LLVM_SRC_DIR} \
-  -B ${LLVM_BUILD_DIR} \
-  -DLLVM_ENABLE_ASSERTIONS=On \
-  -DLLVM_ENABLE_TERMINFO=Off \
-  -DLLVM_ENABLE_LIBXML2=Off \
-  -DLLVM_ENABLE_ZLIB=Off \
-  -DLLVM_ENABLE_PROJECTS='clang;lld' \
-  -DLLVM_TARGETS_TO_BUILD='RISCV' \
-  -DLLVM_ENABLE_ZSTD=Off \
-  -DCMAKE_BUILD_TYPE=MinSizeRel \
-  -DCMAKE_INSTALL_PREFIX=${INSTALL_DIR}
-
-cmake --build ${LLVM_BUILD_DIR}
-cmake --install ${LLVM_BUILD_DIR}
-
-# Build compiler builtins
-COMPILER_RT_SRC_DIR=${LLVM_SRC_PREFIX}/compiler-rt
-COMPILER_RT_BUILD_DIR=${PWD}/build/compiler-rt
-if [ ! -d ${COMPILER_RT_BUILD_DIR} ] ; then
-	mkdir -p ${COMPILER_RT_BUILD_DIR}
-fi
-
-build_compiler_rt() {
-	case "$1" in
-		64) TARGET_ABI=lp64e ;;
-		32) TARGET_ABI=ilp32e ;;
-		*) exit -1
-	esac
-	CFLAGS="--target=riscv${1} -march=rv${1}em -mabi=${TARGET_ABI} -mcpu=generic-rv${1} -nostdlib -nodefaultlibs"
-
-	cmake -G Ninja \
-	  -S ${COMPILER_RT_SRC_DIR} \
-	  -B ${COMPILER_RT_BUILD_DIR} \
-	  -DCMAKE_BUILD_TYPE=Release \
-	  -DCMAKE_INSTALL_PREFIX=${INSTALL_DIR} \
-	  -DCOMPILER_RT_BUILD_BUILTINS=ON \
-	  -DCOMPILER_RT_BUILD_LIBFUZZER=OFF \
-	  -DCOMPILER_RT_BUILD_MEMPROF=OFF \
-	  -DCOMPILER_RT_BUILD_PROFILE=OFF \
-	  -DCOMPILER_RT_BUILD_SANITIZERS=OFF \
-	  -DCOMPILER_RT_BUILD_XRAY=OFF \
-	  -DCMAKE_C_COMPILER=${INSTALL_DIR}/bin/clang \
-	  -DCMAKE_C_COMPILER_TARGET=riscv${1} \
-	  -DCMAKE_ASM_COMPILER_TARGET=riscv${1} \
-	  -DCMAKE_CXX_COMPILER_TARGET=riscv${1} \
-	  -DCMAKE_C_TARGET_BITS=riscv${1} \
-	  -DCMAKE_ASM_TARGET_BITS=riscv${1} \
-	  -DCMAKE_AR=${INSTALL_DIR}/bin/llvm-ar \
-	  -DCMAKE_NM=${INSTALL_DIR}/bin/llvm-nm \
-	  -DCMAKE_RANLIB=${INSTALL_DIR}/bin/llvm-ranlib \
-	  -DCOMPILER_RT_BAREMETAL_BUILD=ON \
-	  -DLLVM_CONFIG_PATH=${INSTALL_DIR}/bin/llvm-config \
-	  -DCMAKE_C_FLAGS="${CFLAGS}" \
-	  -DCMAKE_ASM_FLAGS="${CFLAGS}" \
-	  -DCOMPILER_RT_TEST_COMPILER=${INSTALL_DIR}/bin/clang \
-	  -DCMAKE_CXX_FLAGS="${CFLAGS}" \
-	  -DCMAKE_SYSTEM_NAME=unknown \
-	  -DCOMPILER_RT_DEFAULT_TARGET_ONLY=ON 
-	
-	cmake --build ${COMPILER_RT_BUILD_DIR}
-	cmake --install ${COMPILER_RT_BUILD_DIR} 
-}
-
-build_compiler_rt 32
-build_compiler_rt 64
-
-echo ""
-echo "success"
-echo "add this directory to your PATH: ${INSTALL_DIR}/bin/"

crates/benchmarks/BENCHMARKS.md

@@ -17,56 +17,56 @@
 
 |         | `EVM`                   | `PVMInterpreter`                 |
 |:--------|:------------------------|:-------------------------------- |
-| **`0`** | `5.97 us` (✅ **1.00x**) | `27.04 us` (❌ *4.53x slower*)    |
+| **`0`** | `3.36 us` (✅ **1.00x**) | `11.84 us` (❌ *3.52x slower*)    |
 
 ### OddPorduct
 
-|              | `EVM`                     | `PVMInterpreter`                 |
+|              | `EVM`                    | `PVMInterpreter`                 |
-|:-------------|:--------------------------|:-------------------------------- |
+|:-------------|:-------------------------|:-------------------------------- |
-| **`10000`**  | `4.26 ms` (✅ **1.00x**)   | `2.88 ms` (✅ **1.48x faster**)   |
+| **`10000`**  | `3.11 ms` (✅ **1.00x**)  | `1.53 ms` (🚀 **2.03x faster**)   |
-| **`100000`** | `42.37 ms` (✅ **1.00x**)  | `28.35 ms` (✅ **1.49x faster**)  |
+| **`100000`** | `30.70 ms` (✅ **1.00x**) | `15.54 ms` (🚀 **1.98x faster**)  |
-| **`300000`** | `127.88 ms` (✅ **1.00x**) | `88.43 ms` (✅ **1.45x faster**)  |
+| **`300000`** | `92.68 ms` (✅ **1.00x**) | `45.47 ms` (🚀 **2.04x faster**)  |
 
 ### TriangleNumber
 
-|              | `EVM`                     | `PVMInterpreter`                 |
+|              | `EVM`                    | `PVMInterpreter`                 |
-|:-------------|:--------------------------|:-------------------------------- |
+|:-------------|:-------------------------|:-------------------------------- |
-| **`10000`**  | `2.85 ms` (✅ **1.00x**)   | `2.37 ms` (✅ **1.20x faster**)   |
+| **`10000`**  | `2.29 ms` (✅ **1.00x**)  | `1.09 ms` (🚀 **2.11x faster**)   |
-| **`100000`** | `27.85 ms` (✅ **1.00x**)  | `23.01 ms` (✅ **1.21x faster**)  |
+| **`100000`** | `22.84 ms` (✅ **1.00x**) | `10.66 ms` (🚀 **2.14x faster**)  |
-| **`360000`** | `103.01 ms` (✅ **1.00x**) | `83.66 ms` (✅ **1.23x faster**)  |
+| **`360000`** | `82.29 ms` (✅ **1.00x**) | `37.01 ms` (🚀 **2.22x faster**)  |
 
 ### FibonacciRecursive
 
 |          | `EVM`                     | `PVMInterpreter`                  |
 |:---------|:--------------------------|:--------------------------------- |
-| **`12`** | `195.19 us` (✅ **1.00x**) | `333.53 us` (❌ *1.71x slower*)    |
+| **`12`** | `135.67 us` (✅ **1.00x**) | `125.02 us` (✅ **1.09x faster**)  |
-| **`16`** | `1.22 ms` (✅ **1.00x**)   | `1.97 ms` (❌ *1.62x slower*)      |
+| **`16`** | `903.75 us` (✅ **1.00x**) | `762.79 us` (✅ **1.18x faster**)  |
-| **`20`** | `8.14 ms` (✅ **1.00x**)   | `13.20 ms` (❌ *1.62x slower*)     |
+| **`20`** | `6.12 ms` (✅ **1.00x**)   | `4.96 ms` (✅ **1.23x faster**)    |
-| **`24`** | `55.09 ms` (✅ **1.00x**)  | `88.56 ms` (❌ *1.61x slower*)     |
+| **`24`** | `42.05 ms` (✅ **1.00x**)  | `33.86 ms` (✅ **1.24x faster**)   |
 
 ### FibonacciIterative
 
-|           | `EVM`                    | `PVMInterpreter`                  |
+|           | `EVM`                    | `PVMInterpreter`                 |
-|:----------|:-------------------------|:--------------------------------- |
+|:----------|:-------------------------|:-------------------------------- |
-| **`64`**  | `33.39 us` (✅ **1.00x**) | `86.02 us` (❌ *2.58x slower*)     |
+| **`64`**  | `15.04 us` (✅ **1.00x**) | `29.45 us` (❌ *1.96x slower*)    |
-| **`128`** | `52.91 us` (✅ **1.00x**) | `126.38 us` (❌ *2.39x slower*)    |
+| **`128`** | `26.36 us` (✅ **1.00x**) | `42.19 us` (❌ *1.60x slower*)    |
-| **`256`** | `82.33 us` (✅ **1.00x**) | `208.74 us` (❌ *2.54x slower*)    |
+| **`256`** | `48.61 us` (✅ **1.00x**) | `65.71 us` (❌ *1.35x slower*)    |
 
 ### FibonacciBinet
 
-|           | `EVM`                    | `PVMInterpreter`                  |
+|           | `EVM`                    | `PVMInterpreter`                 |
-|:----------|:-------------------------|:--------------------------------- |
+|:----------|:-------------------------|:-------------------------------- |
-| **`64`**  | `32.29 us` (✅ **1.00x**) | `161.75 us` (❌ *5.01x slower*)    |
+| **`64`**  | `15.22 us` (✅ **1.00x**) | `41.46 us` (❌ *2.72x slower*)    |
-| **`128`** | `36.02 us` (✅ **1.00x**) | `172.59 us` (❌ *4.79x slower*)    |
+| **`128`** | `17.05 us` (✅ **1.00x**) | `42.84 us` (❌ *2.51x slower*)    |
-| **`256`** | `41.21 us` (✅ **1.00x**) | `185.30 us` (❌ *4.50x slower*)    |
+| **`256`** | `19.00 us` (✅ **1.00x**) | `44.36 us` (❌ *2.34x slower*)    |
 
 ### SHA1
 
 |           | `EVM`                     | `PVMInterpreter`                  |
 |:----------|:--------------------------|:--------------------------------- |
-| **`1`**   | `160.17 us` (✅ **1.00x**) | `403.46 us` (❌ *2.52x slower*)    |
+| **`1`**   | `110.04 us` (✅ **1.00x**) | `216.11 us` (❌ *1.96x slower*)    |
-| **`64`**  | `286.69 us` (✅ **1.00x**) | `479.79 us` (❌ *1.67x slower*)    |
+| **`64`**  | `209.04 us` (✅ **1.00x**) | `309.48 us` (❌ *1.48x slower*)    |
-| **`512`** | `1.18 ms` (✅ **1.00x**)   | `1.37 ms` (❌ *1.16x slower*)      |
+| **`512`** | `903.65 us` (✅ **1.00x**) | `980.49 us` (✅ **1.09x slower**)  |
 
 ---
 Made with [criterion-table](https://github.com/nu11ptr/criterion-table)

crates/benchmarks/benches/execute.rs

@@ -46,7 +46,7 @@ fn group<'error, M>(c: &'error mut Criterion<M>, group_name: &str) -> BenchmarkG
 where
     M: Measurement,
 {
-    return c.benchmark_group(group_name);
+    c.benchmark_group(group_name)
 }
 
 fn bench_baseline(c: &mut Criterion) {

crates/build-utils/.gitignore

@@ -0,0 +1,14 @@
+# Generated by Cargo
+# will have compiled files and executables
+/target/
+
+# Remove Cargo.lock from gitignore if creating an executable, leave it for libraries
+# More information here https://doc.rust-lang.org/cargo/guide/cargo-toml-vs-cargo-lock.html
+Cargo.lock
+
+# These are backup files generated by rustfmt
+**/*.rs.bk
+
+# IDE
+/.idea/
+/.vscode/

crates/build-utils/Cargo.toml

@@ -0,0 +1,15 @@
+[package]
+name = "revive-build-utils"
+version.workspace = true
+license.workspace = true
+edition.workspace = true
+repository.workspace = true
+authors = [
+    "Cyrill Leutwiler <cyrill@parity.io>",
+]
+description = "Shared build utilities of the revive compiler"
+
+[lib]
+doctest = false
+
+[dependencies]

crates/build-utils/README.md

@@ -0,0 +1,10 @@
+# revive: Compiler Common
+
+## License
+
+This library is distributed under the terms of either
+
+- Apache License, Version 2.0, ([LICENSE-APACHE](LICENSE-APACHE) or <http://www.apache.org/licenses/LICENSE-2.0>)
+- MIT license ([LICENSE-MIT](LICENSE-MIT) or <http://opensource.org/licenses/MIT>)
+
+at your option.

crates/build-utils/src/lib.rs

@@ -0,0 +1,48 @@
+//! The compiler build utilities library.
+
+/// The revive LLVM host dependency directory prefix environment variable.
+pub const REVIVE_LLVM_HOST_PREFIX: &str = "LLVM_SYS_181_PREFIX";
+
+/// The revive LLVM target dependency directory prefix environment variable.
+pub const REVIVE_LLVM_TARGET_PREFIX: &str = "REVIVE_LLVM_TARGET_PREFIX";
+
+/// Constructs a path to the LLVM tool `name`.
+///
+/// Respects the [`REVIVE_LLVM_HOST_PREFIX`] environment variable.
+pub fn llvm_host_tool(name: &str) -> std::path::PathBuf {
+    std::env::var_os(REVIVE_LLVM_HOST_PREFIX)
+        .map(Into::<std::path::PathBuf>::into)
+        .unwrap_or_else(|| {
+            panic!(
+                "install LLVM using the revive-llvm builder and export {REVIVE_LLVM_HOST_PREFIX}",
+            )
+        })
+        .join("bin")
+        .join(name)
+}
+
+/// Returns the LLVM lib dir.
+///
+/// Respects the [`REVIVE_LLVM_HOST_PREFIX`] environment variable.
+pub fn llvm_lib_dir() -> std::path::PathBuf {
+    std::path::PathBuf::from(llvm_config("--libdir").trim())
+}
+
+/// Returns the LLVM CXX compiler flags.
+///
+/// Respects the [`REVIVE_LLVM_HOST_PREFIX`] environment variable.
+pub fn llvm_cxx_flags() -> String {
+    llvm_config("--cxxflags")
+}
+
+/// Execute the `llvm-config` utility respecting the [`REVIVE_LLVM_HOST_PREFIX`] environment variable.
+fn llvm_config(arg: &str) -> String {
+    let llvm_config = llvm_host_tool("llvm-config");
+    let output = std::process::Command::new(&llvm_config)
+        .arg(arg)
+        .output()
+        .unwrap_or_else(|error| panic!("`{} {arg}` failed: {error}", llvm_config.display()));
+
+    String::from_utf8(output.stdout)
+        .unwrap_or_else(|_| panic!("output of `{} {arg}` should be utf8", llvm_config.display()))
+}

crates/builtins/Cargo.toml

@@ -8,7 +8,5 @@ authors.workspace = true
 build = "build.rs"
 description = "compiler builtins for the revive compiler"
 
-[features]
+[build-dependencies]
-riscv-64 = []
+revive-build-utils = { workspace = true }
-
-[dependencies]

crates/builtins/build.rs

@@ -1,30 +1,34 @@
 use std::{env, fs, io::Read, path::Path, process::Command};
 
-#[cfg(not(feature = "riscv-64"))]
-pub const BUILTINS_ARCHIVE_FILE: &str = "libclang_rt.builtins-riscv32.a";
-#[cfg(feature = "riscv-64")]
 pub const BUILTINS_ARCHIVE_FILE: &str = "libclang_rt.builtins-riscv64.a";
 
 fn main() {
-    let mut llvm_lib_dir = String::new();
+    println!(
+        "cargo:rerun-if-env-changed={}",
+        revive_build_utils::REVIVE_LLVM_HOST_PREFIX
+    );
 
-    Command::new("llvm-config")
+    let llvm_config = revive_build_utils::llvm_host_tool("llvm-config");
-        .args(["--libdir"])
+    let mut llvm_lib_dir = String::new();
+    Command::new(&llvm_config)
+        .arg("--libdir")
         .output()
-        .expect("llvm-config should be able to provide LD path")
+        .unwrap_or_else(|_| {
+            panic!(
+                "{} should be able to provide LD path",
+                llvm_config.display()
+            )
+        })
         .stdout
         .as_slice()
         .read_to_string(&mut llvm_lib_dir)
         .expect("llvm-config output should be utf8");
 
-    let mut lib_path = std::path::PathBuf::from(llvm_lib_dir.trim())
+    let lib_path = revive_build_utils::llvm_lib_dir()
-        .join("linux")
+        .join("unknown")
         .join(BUILTINS_ARCHIVE_FILE);
-    if !lib_path.exists() {
+    let archive = fs::read(&lib_path).expect("clang builtins not found");
-        lib_path = std::path::PathBuf::from(std::env::var("CARGO_MANIFEST_DIR").unwrap())
+    println!("cargo:rerun-if-env-changed={}", lib_path.display());
-            .join(BUILTINS_ARCHIVE_FILE);
-    }
-    let archive = fs::read(lib_path).expect("clang builtins not found");
 
     let out_dir = env::var_os("OUT_DIR").expect("has OUT_DIR");
     let archive_path = Path::new(&out_dir).join(BUILTINS_ARCHIVE_FILE);

crates/common/Cargo.toml

@@ -13,9 +13,6 @@ description = "Shared constants of the revive compiler"
 [lib]
 doctest = false
 
-[features]
-riscv-64 = []
-
 [dependencies]
 anyhow = { workspace = true }
 serde = { workspace = true, features = ["derive"] }

crates/common/src/extension.rs

@@ -12,12 +12,6 @@ pub static EXTENSION_ABI: &str = "abi";
 /// The Yul IR file extension.
 pub static EXTENSION_YUL: &str = "yul";
 
-/// The EVM legacy assembly IR file extension.
-pub static EXTENSION_EVMLA: &str = "evmla";
-
-/// The Ethereal IR file extension.
-pub static EXTENSION_ETHIR: &str = "ethir";
-
 /// The EVM file extension.
 pub static EXTENSION_EVM: &str = "evm";
 
@@ -41,3 +35,6 @@ pub static EXTENSION_POLKAVM_ASSEMBLY: &str = "pvmasm";
 
 /// The PolkaVM bytecode file extension.
 pub static EXTENSION_POLKAVM_BINARY: &str = "pvm";
+
+/// The ELF shared object file extension.
+pub static EXTENSION_SHARED_OBJECT: &str = "so";

crates/differential/genesis.json

@@ -16,9 +16,16 @@
         "shanghaiTime": 0,
         "cancunTime": 0,
         "terminalTotalDifficulty": 0,
-        "terminalTotalDifficultyPassed": true
+        "terminalTotalDifficultyPassed": true,
+        "blobSchedule": {
+            "cancun": {
+                "target": 3,
+                "max": 6,
+                "baseFeeUpdateFraction": 3338477
+            }
+        }
     },
-    "coinbase": "0x0000000000000000000000000000000000000000",
+    "coinbase": "0xffffffffffffffffffffffffffffffffffffffff",
     "difficulty": "0x20000",
     "extraData": "",
     "gasLimit": "0xffffffff",

crates/differential/src/lib.rs

@@ -40,7 +40,7 @@ const EXECUTABLE_ARGS_BENCH: [&str; 6] = [
     "-",
 ];
 const GAS_USED_MARKER: &str = "EVM gas used:";
-const REVERT_MARKER: &str = "error: execution reverted";
+const REVERT_MARKER: &str = " error: ";
 
 /// The geth EVM state dump structure
 #[derive(Clone, Debug, Default, Serialize, Deserialize)]
@@ -413,7 +413,7 @@ impl Evm {
         let stderr = str::from_utf8(output.stderr.as_slice())
             .unwrap_or_else(|err| panic!("{EXECUTABLE_NAME} stderr failed to parse: {err}"));
 
-        let mut log: EvmLog = stdout.into();
+        let mut log: EvmLog = format!("{stdout}{stderr}").as_str().into();
         log.stderr = stderr.into();
         if self.bench {
             log.parse_gas_used_from_bench();
@@ -557,6 +557,7 @@ allocated bytes: 3711"#;
     }
 
     #[test]
+    #[ignore] // https://github.com/ethereum/go-ethereum/issues/30778
     fn bench_flipper() {
         let log_runtime = Evm::default()
             .code_blob(EVM_BIN_RUNTIME_FIXTURE.as_bytes().to_vec())

crates/integration/Cargo.toml

@@ -8,18 +8,12 @@ authors.workspace = true
 description = "revive compiler integration test cases"
 
 [dependencies]
-polkavm = { workspace = true }
 alloy-primitives = { workspace = true }
 alloy-sol-types = { workspace = true }
 hex = { workspace = true }
-env_logger = { workspace = true }
-log = { workspace = true }
 serde_json = { workspace = true }
 
 revive-solidity = { workspace = true }
-revive-differential = { workspace = true }
-revive-llvm-context = { workspace = true }
-revive-common = { workspace = true }
 revive-runner = { workspace = true }
 
 [dev-dependencies]

crates/integration/codesize.json

@@ -1,10 +1,10 @@
 {
-  "Baseline": 912,
+  "Baseline": 1110,
-  "Computation": 4413,
+  "Computation": 2389,
-  "DivisionArithmetics": 40689,
+  "DivisionArithmetics": 14822,
-  "ERC20": 54374,
+  "ERC20": 23973,
-  "Events": 1726,
+  "Events": 1605,
-  "FibonacciIterative": 3015,
+  "FibonacciIterative": 2023,
-  "Flipper": 3612,
+  "Flipper": 1989,
-  "SHA1": 32865
+  "SHA1": 17026
 }

crates/integration/contracts/Balance.sol

@@ -0,0 +1,81 @@
+// SPDX-License-Identifier: MIT
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Upload": {
+                "code": {
+                    "Solidity": {
+                        "contract": "BalanceReceiver"
+                    }
+                }
+            }
+        },
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "Balance"
+                    }
+                },
+                "value": 24
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "1c8d16b30000000000000000000000000303030303030303030303030303030303030303000000000000000000000000000000000000000000000000000000000000000a"
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "6ada15d90000000000000000000000000303030303030303030303030303030303030303000000000000000000000000000000000000000000000000000000000000000a"
+            }
+        }
+    ]
+}
+*/
+
+contract BalanceReceiver {
+    constructor() payable {}
+
+    fallback() external payable {}
+}
+
+contract Balance {
+    constructor() payable {
+        // 0 to EOA
+        transfer_to(payable(address(0xdeadbeef)), 0);
+        send_to(payable(address(0xdeadbeef)), 0);
+
+        // 1 to EOA
+        transfer_to(payable(address(0xcafebabe)), 1);
+        send_to(payable(address(0xcafebabe)), 1);
+
+        BalanceReceiver balanceReceiver = new BalanceReceiver();
+
+        // 0 to contract
+        transfer_to(payable(address(balanceReceiver)), 0);
+        send_to(payable(address(balanceReceiver)), 0);
+
+        // 1 to contract
+        transfer_to(payable(address(balanceReceiver)), 1);
+        send_to(payable(address(balanceReceiver)), 1);
+    }
+
+    function transfer_to(address payable _dest, uint _amount) public payable {
+        _dest.transfer(_amount);
+    }
+
+    function send_to(address payable _dest, uint _amount) public payable {
+        require(_dest.send(_amount));
+    }
+}

crates/integration/contracts/BaseFee.sol

@@ -0,0 +1,31 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "BaseFee"
+                    }
+                }
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        }
+    ]
+}
+*/
+
+contract BaseFee {
+    constructor() payable {
+        assert(block.basefee == 0);
+    }
+}

crates/integration/contracts/BlockHash.sol

@@ -0,0 +1,33 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "BlockHash"
+                    }
+                },
+                "data": "4545454545454545454545454545454545454545454545454545454545454545"
+            }
+        }
+    ]
+}
+*/
+
+contract BlockHash {
+    constructor(bytes32 expected) payable {
+        assert(blockhash(0) == expected);
+        assert(blockhash(1) == 0);
+        assert(
+            blockhash(
+                0xffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff
+            ) == 0
+        );
+    }
+}

crates/integration/contracts/Call.sol

@@ -21,16 +21,8 @@ pragma solidity ^0.8;
                     "Solidity": {
                         "contract": "Caller"
                     }
-                }
-            }
-        },
-        {
-            "Call": {
-                "dest": {
-                    "Instantiated": 0
                 },
-                "value": 123,
+                "value": 123
-                "data": "1eb16e5b000000000000000000000000d8b934580fce35a11b58c6d73adee468a2833fa8"
             }
         },
         {
@@ -49,11 +41,14 @@ contract Callee {
     function echo(bytes memory payload) public pure returns (bytes memory) {
         return payload;
     }
+
+    receive() external payable {}
 }
 
 contract Caller {
-    function value_transfer(address payable destination) public payable {
+    constructor() payable {
-        destination.transfer(msg.value);
+        Callee callee = new Callee();
+        payable(address(callee)).transfer(msg.value);
     }
 
     function call(bytes memory payload) public returns (bytes memory) {

crates/integration/contracts/Coinbase.sol

@@ -0,0 +1,27 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "Coinbase"
+                    }
+                }
+            }
+        }
+    ]
+}
+*/
+
+contract Coinbase {
+    constructor() payable {
+        address coinbase = address(0xFFfFfFffFFfffFFfFFfFFFFFffFFFffffFfFFFfF);
+        assert(block.coinbase == coinbase);
+    }
+}

crates/integration/contracts/Create2.sol

@@ -0,0 +1,50 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Upload": {
+                "code": {
+                    "Solidity": {
+                        "contract": "CreateA"
+                    }
+                }
+            }
+        },
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "CreateB"
+                    }
+                },
+                "value": 100000
+            }
+        }
+    ]
+}
+*/
+
+contract CreateA {
+    constructor() payable {}
+}
+
+contract CreateB {
+    constructor() payable {
+        bytes32 salt = hex"ff";
+
+        try new CreateA{salt: salt}() returns (CreateA) {} catch {
+            revert("the first instantiation should succeed");
+        }
+
+        try new CreateA{salt: salt}() returns (CreateA) {} catch {
+            return;
+        }
+
+        revert("the second instantiation should have failed");
+    }
+}

crates/integration/contracts/Delegate.sol

@@ -0,0 +1,84 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Upload": {
+                "code": {
+                    "Solidity": {
+                        "contract": "Logic"
+                    }
+                }
+            }
+        },
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "Tester"
+                    }
+                }
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "value": 123,
+                "data": "6466414b0000000000000000000000000000000000000000000000000000000000000020"
+            }
+        }
+    ]
+}
+*/
+
+contract Logic {
+    // NOTE: storage layout must be the same as contract Tester
+    uint256 public num;
+    address public sender;
+    uint256 public value;
+
+    uint public immutable multiplier = 4;
+
+    event DidSetVars();
+
+    function setVars(uint256 _num) public payable returns (uint256) {
+        num = _num * multiplier;
+        sender = msg.sender;
+        value = msg.value;
+        emit DidSetVars();
+        return _num;
+    }
+}
+
+contract Tester {
+    uint256 public num;
+    address public sender;
+    uint256 public value;
+
+    uint public immutable multiplier = 2;
+
+    function setVars(uint256 _num) public payable returns (bool, bytes memory) {
+        Logic impl = new Logic();
+
+        // Tester's storage is set, Logic is not modified.
+        (bool success, bytes memory data) = address(impl).delegatecall(
+            abi.encodeWithSignature("setVars(uint256)", _num)
+        );
+
+        assert(success);
+        assert(impl.num() == 0);
+        assert(impl.sender() == address(0));
+        assert(impl.value() == 0);
+        assert(num == _num * 4);
+        assert(sender == msg.sender);
+        assert(value == msg.value);
+
+        return (success, data);
+    }
+}

crates/integration/contracts/ERC20.sol

@@ -4,11 +4,27 @@ pragma solidity ^0.8;
 
 /* runner.json
 {
+    "differential": true,
     "actions": [
-    {
+        {
-      "Instantiate": {}
+            "Upload": {
-    }
+                "code": {
-  ]
+                    "Solidity": {
+                        "contract": "ERC20"
+                    }
+                }
+            }
+        },
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "ERC20Tester"
+                    }
+                }
+            }
+        }
+    ]
 }
 */
 
@@ -82,3 +98,26 @@ contract ERC20 is IERC20 {
         emit Transfer(msg.sender, address(0), amount);
     }
 }
+
+contract ERC20Tester {
+    constructor() {
+        address BOB = address(0xffffffffffffffffffffffffffffffffffffff);
+        ERC20 token = new ERC20();
+        assert(token.decimals() == 18);
+
+        token.mint(300);
+        assert(token.balanceOf(address(this)) == 300);
+        token.transfer(BOB, 100);
+        assert(token.balanceOf(address(this)) == 200);
+        assert(token.balanceOf(BOB) == 100);
+
+        token.approve(address(this), 100);
+
+        token.transferFrom(address(this), BOB, 100);
+        assert(token.balanceOf(BOB) == 200);
+        assert(token.balanceOf(address(this)) == 100);
+
+        token.burn(100);
+        assert(token.balanceOf(address(this)) == 0);
+    }
+}

crates/integration/contracts/ExtCode.sol

@@ -14,4 +14,16 @@ contract ExtCode {
             ret := codesize()
         }
     }
+
+    function ExtCodeHash(address who) public view returns (bytes32 ret) {
+        assembly {
+            ret := extcodehash(who)
+        }
+    }
+
+    function CodeHash() public view returns (bytes32 ret) {
+        assembly {
+            ret := extcodehash(address())
+        }
+    }
 }

crates/integration/contracts/FunctionPointer.sol

@@ -0,0 +1,42 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "FunctionPointer"
+                    }
+                }
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "26121ff0"
+            }
+        }
+    ]
+}
+*/
+
+contract FunctionPointer {
+    bool public flag = false;
+
+    function f0() public {
+        flag = true;
+    }
+
+    function f() public returns (bool) {
+        function() internal x = f0;
+        x();
+        return flag;
+    }
+}

crates/integration/contracts/GasLeft.sol

@@ -0,0 +1,32 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "GasLeft"
+                    }
+                }
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        }
+    ]
+}
+*/
+
+contract GasLeft {
+    constructor() payable {
+        assert(gasleft() > gasleft());
+        assert(gasleft() > 0 && gasleft() < 0xffffffffffffffff);
+    }
+}

crates/integration/contracts/GasLimit.sol

@@ -0,0 +1,31 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "GasLimit"
+                    }
+                }
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        }
+    ]
+}
+*/
+
+contract GasLimit {
+    constructor() payable {
+        assert(block.gaslimit == 2000000000000);
+    }
+}

crates/integration/contracts/GasPrice.sol

@@ -0,0 +1,31 @@
+// SPDX-License-Identifier: MIT
+
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "GasPrice"
+                    }
+                }
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        }
+    ]
+}
+*/
+
+contract GasPrice {
+    constructor() payable {
+        assert(tx.gasprice == 1000);
+    }
+}

crates/integration/contracts/Immutables.sol

@@ -0,0 +1,63 @@
+// SPDX-License-Identifier: MIT
+pragma solidity ^0.8;
+
+/* runner.json
+{
+  "differential": true,
+  "actions": [
+    {
+      "Upload": {
+        "code": {
+          "Solidity": {
+            "contract": "ImmutablesTester"
+          }
+        }
+      }
+    },
+    {
+      "Instantiate": {
+        "code": {
+          "Solidity": {
+            "contract": "Immutables"
+          }
+        }
+      }
+    },
+    {
+      "Call": {
+        "dest": {
+          "Instantiated": 0
+        }
+      }
+    }
+  ]
+}
+*/
+
+contract ImmutablesTester {
+    // Read should work in the runtime code
+    uint public immutable foo;
+    // Read should work in the runtime code
+    uint public immutable bar;
+    // Read should work in the runtime code
+    uint public immutable zoo;
+
+    // Assign and read should work in the constructor
+    constructor(uint _foo) payable {
+        foo = _foo;
+        bar = foo + 1;
+        zoo = bar + 2;
+
+        assert(zoo == _foo + 3);
+    }
+}
+
+contract Immutables {
+    fallback() external {
+        ImmutablesTester tester = new ImmutablesTester(127);
+
+        assert(tester.foo() == 127);
+        assert(tester.bar() == tester.foo() + 1);
+        assert(tester.zoo() == tester.bar() + 2);
+    }
+}

crates/integration/contracts/MSize.sol

@@ -4,6 +4,7 @@ pragma solidity ^0.8;
 
 /* runner.json
 {
+    "differential": true,
     "actions": [
     {
         "Instantiate": {
@@ -23,12 +24,6 @@ pragma solidity ^0.8;
             "data": "f016832c"
         }
     },
-    {
-        "VerifyCall": {
-            "success": true,
-            "output": "0000000000000000000000000000000000000000000000000000000000000060"
-        }
-    },
     {
         "Call": {
             "dest": {
@@ -36,12 +31,6 @@ pragma solidity ^0.8;
             },
             "data": "f4a63aa5"
         }
-    },
-    {
-        "VerifyCall": {
-            "success": true,
-            "output": "0000000000000000000000000000000000000000000000000000000000000084"
-        }
     }
   ]
 }

crates/integration/contracts/Send.sol

@@ -0,0 +1,75 @@
+// SPDX-License-Identifier: MIT
+pragma solidity ^0.8;
+
+/* runner.json
+{
+    "differential": false,
+    "actions": [
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "Send"
+                    }
+                },
+                "value": 211
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "1c8d16b30000000000000000000000000303030303030303030303030303030303030303000000000000000000000000000000000000000000000000000000000000000a"
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "fb9e8d050000000000000000000000000000000000000000000000000000000000000001"
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": false
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "fb9e8d050000000000000000000000000000000000000000000000000000000000000000"
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": false
+            }
+        }
+    ]
+}
+*/
+
+contract Send {
+    constructor() payable {}
+
+    function transfer_self(uint _amount) public payable {
+        transfer_to(payable(address(this)), _amount);
+    }
+
+    function transfer_to(address payable _dest, uint _amount) public payable {
+        if (_dest.send(_amount)) {}
+    }
+
+    fallback() external {}
+
+    receive() external payable {}
+}

crates/integration/contracts/Storage.sol

@@ -22,6 +22,14 @@ pragma solidity ^0.8;
                 },
                 "data": "fabc9efaffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff"
             }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "558b9f9bffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff"
+            }
         }
     ]
 }
@@ -30,10 +38,18 @@ pragma solidity ^0.8;
 contract Storage {
     function transient(uint value) public returns (uint ret) {
         assembly {
-            let slot := 123
+            let slot := 0xffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff00
             tstore(slot, value)
             let success := call(0, 0, 0, 0, 0, 0, 0)
             ret := tload(slot)
         }
     }
+
+    function persistent(uint value) public returns (uint ret) {
+        assembly {
+            let slot := 0xffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff00
+            sstore(slot, value)
+            ret := sload(slot)
+        }
+    }
 }

crates/integration/contracts/Transaction.sol

@@ -0,0 +1,54 @@
+// SPDX-License-Identifier: MIT
+pragma solidity ^0.8.0;
+
+/* runner.json
+{
+    "differential": true,
+    "actions": [
+        {
+            "Upload": {
+                "code": {
+                    "Solidity": {
+                        "contract": "TransactionOrigin"
+                    }
+                }
+            }
+        },
+        {
+            "Instantiate": {
+                "code": {
+                    "Solidity": {
+                        "contract": "TransactionTester"
+                    }
+                }
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "f8a8fd6d"
+            }
+        }
+    ]
+}
+*/
+
+contract TransactionTester {
+    constructor() payable {
+        assert(tx.origin == new TransactionOrigin().test());
+    }
+
+    function test() public payable returns (address ret) {
+        ret = tx.origin;
+    }
+}
+
+contract TransactionOrigin {
+    function test() public payable returns (address ret) {
+        assert(msg.sender != tx.origin);
+
+        ret = tx.origin;
+    }
+}

crates/integration/contracts/Transfer.sol

@@ -3,7 +3,7 @@ pragma solidity ^0.8;
 
 /* runner.json
 {
-    "differential": true,
+    "differential": false,
     "actions": [
         {
             "Instantiate": {
@@ -12,7 +12,7 @@ pragma solidity ^0.8;
                         "contract": "Transfer"
                     }
                 },
-                "value": 11
+                "value": 211
             }
         },
         {
@@ -23,12 +23,35 @@ pragma solidity ^0.8;
                 "data": "1c8d16b30000000000000000000000000303030303030303030303030303030303030303000000000000000000000000000000000000000000000000000000000000000a"
             }
         },
+        {
+            "VerifyCall": {
+                "success": true
+            }
+        },
         {
             "Call": {
                 "dest": {
                     "Instantiated": 0
                 },
-                "data": "fb9e8d0500000000000000000000000003030303030303030303030303030303030303030000000000000000000000000000000000000000000000000000000000000001"
+                "data": "fb9e8d050000000000000000000000000000000000000000000000000000000000000001"
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": false
+            }
+        },
+        {
+            "Call": {
+                "dest": {
+                    "Instantiated": 0
+                },
+                "data": "fb9e8d050000000000000000000000000000000000000000000000000000000000000000"
+            }
+        },
+        {
+            "VerifyCall": {
+                "success": false
             }
         }
     ]
@@ -36,19 +59,17 @@ pragma solidity ^0.8;
 */
 
 contract Transfer {
-    constructor() payable {
+    constructor() payable {}
-        transfer_self(msg.value);
-    }
-
-    function address_self() internal view returns (address payable) {
-        return payable(address(this));
-    }
 
     function transfer_self(uint _amount) public payable {
-        transfer_to(address_self(), _amount);
+        transfer_to(payable(address(this)), _amount);
     }
 
     function transfer_to(address payable _dest, uint _amount) public payable {
         _dest.transfer(_amount);
     }
+
+    fallback() external {}
+
+    receive() external payable {}
 }

crates/integration/contracts/Value.sol

@@ -5,6 +5,15 @@ pragma solidity ^0.8;
 {
     "differential": true,
     "actions": [
+        {
+            "Upload": {
+                "code": {
+                    "Solidity": {
+                        "contract": "ValueTester"
+                    }
+                }
+            }
+        },
         {
             "Instantiate": {
                 "value": 1024,
@@ -23,31 +32,35 @@ pragma solidity ^0.8;
                 "value": 123,
                 "data": "3fa4f245"
             }
-        },
-        {
-            "Call": {
-                "dest": {
-                    "Instantiated": 0
-                },
-                "data": "52da5fa0"
-            }
         }
     ]
 }
 */
 
-contract Value {
+contract ValueTester {
     constructor() payable {}
 
-    function value() public payable returns (uint ret) {
-        ret = msg.value;
-    }
-
     function balance_self() public view returns (uint ret) {
         ret = address(this).balance;
     }
+}
 
-    function balance_of(address _address) public view returns (uint ret) {
+contract Value {
-        ret = _address.balance;
+    constructor() payable {
+        ValueTester tester = new ValueTester{value: msg.value}();
+
+        // own account
+        assert(address(this).balance == 0);
+
+        // tester account
+        assert(address(tester).balance == msg.value);
+        assert(tester.balance_self() == msg.value);
+
+        // non-existant account
+        assert(address(0xdeadbeef).balance == 0);
+    }
+
+    function value() public payable returns (uint ret) {
+        ret = msg.value;
     }
 }

crates/integration/src/cases.rs

@@ -156,6 +156,20 @@ case!("DivisionArithmetics.sol", DivisionArithmetics, sdivCall, division_arithme
 case!("DivisionArithmetics.sol", DivisionArithmetics, modCall, division_arithmetics_mod, n: U256, d: U256);
 case!("DivisionArithmetics.sol", DivisionArithmetics, smodCall, division_arithmetics_smod, n: I256, d: I256);
 
+sol!(
+    contract Send {
+        function transfer_self(uint _amount) public payable;
+    }
+);
+case!("Send.sol", Send, transfer_selfCall, send_self, amount: U256);
+
+sol!(
+    contract Transfer {
+        function transfer_self(uint _amount) public payable;
+    }
+);
+case!("Transfer.sol", Transfer, transfer_selfCall, transfer_self, amount: U256);
+
 sol!(
     contract MStore8 {
         function mStore8(uint value) public pure returns (uint256 word);
@@ -178,10 +192,16 @@ sol!(
         function ExtCodeSize(address who) public view returns (uint ret);
 
         function CodeSize() public pure returns (uint ret);
+
+        function ExtCodeHash(address who) public view returns (bytes32 ret);
+
+        function CodeHash() public view returns (bytes32 ret);
     }
 );
 case!("ExtCode.sol", ExtCode, ExtCodeSizeCall, ext_code_size, address: Address);
 case!("ExtCode.sol", ExtCode, CodeSizeCall, code_size,);
+case!("ExtCode.sol", ExtCode, ExtCodeHashCall, ext_code_hash, address: Address);
+case!("ExtCode.sol", ExtCode, CodeHashCall, code_hash,);
 
 sol!(
     contract MCopy {

crates/integration/src/tests.rs

@@ -37,11 +37,24 @@ test_spec!(events, "Events", "Events.sol");
 test_spec!(storage, "Storage", "Storage.sol");
 test_spec!(mstore8, "MStore8", "MStore8.sol");
 test_spec!(address, "Context", "Context.sol");
-test_spec!(balance, "Value", "Value.sol");
+test_spec!(value, "Value", "Value.sol");
 test_spec!(create, "CreateB", "Create.sol");
 test_spec!(call, "Caller", "Call.sol");
-test_spec!(transfer, "Transfer", "Transfer.sol");
+test_spec!(balance, "Balance", "Balance.sol");
 test_spec!(return_data_oob, "ReturnDataOob", "ReturnDataOob.sol");
+test_spec!(immutables, "Immutables", "Immutables.sol");
+test_spec!(transaction, "Transaction", "Transaction.sol");
+test_spec!(block_hash, "BlockHash", "BlockHash.sol");
+test_spec!(delegate, "Delegate", "Delegate.sol");
+test_spec!(gas_price, "GasPrice", "GasPrice.sol");
+test_spec!(gas_left, "GasLeft", "GasLeft.sol");
+test_spec!(gas_limit, "GasLimit", "GasLimit.sol");
+test_spec!(base_fee, "BaseFee", "BaseFee.sol");
+test_spec!(coinbase, "Coinbase", "Coinbase.sol");
+test_spec!(create2, "CreateB", "Create2.sol");
+test_spec!(transfer, "Transfer", "Transfer.sol");
+test_spec!(send, "Send", "Send.sol");
+test_spec!(function_pointer, "FunctionPointer", "FunctionPointer.sol");
 
 fn instantiate(path: &str, contract: &str) -> Vec<SpecsAction> {
     vec![Instantiate {
@@ -53,7 +66,6 @@ fn instantiate(path: &str, contract: &str) -> Vec<SpecsAction> {
             path: Some(path.into()),
             contract: contract.to_string(),
             solc_optimizer: None,
-            pipeline: None,
         },
         data: vec![],
         salt: OptionalHex::default(),
@@ -229,84 +241,243 @@ fn signed_remainder() {
             data: Contract::division_arithmetics_smod(n, d).calldata,
         })
     }
+
+    run_differential(actions);
 }
 
-/*
-// These test were implement for the mock-runtime and need to be ported yet.
-
 #[test]
-fn create2_failure() {
+fn ext_code_hash() {
-    let mut state = State::default();
+    let mut actions = instantiate("contracts/ExtCode.sol", "ExtCode");
-    let contract_a = Contract::create_a();
-    state.upload_code(&contract_a.pvm_runtime);
 
-    let contract = Contract::create_b();
+    // First do contract instantiation to figure out address and code hash
-    let (state, output) = state
+    let results = Specs {
-        .transaction()
+        actions: actions.clone(),
-        .with_default_account(&contract.pvm_runtime)
+        ..Default::default()
-        .calldata(contract.calldata.clone())
+    }
-        .call();
+    .run();
+    let (addr, code_hash) = match results.first().cloned() {
+        Some(CallResult::Instantiate {
+            result, code_hash, ..
+        }) => (result.result.unwrap().addr, code_hash),
+        _ => panic!("instantiate contract failed"),
+    };
 
-    assert_eq!(output.flags, ReturnFlags::Success);
+    // code hash of itself
+    actions.push(Call {
+        origin: TestAddress::Alice,
+        dest: TestAddress::Instantiated(0),
+        value: 0,
+        gas_limit: None,
+        storage_deposit_limit: None,
+        data: Contract::code_hash().calldata,
+    });
+    actions.push(VerifyCall(VerifyCallExpectation {
+        success: true,
+        output: OptionalHex::from(code_hash.as_bytes().to_vec()),
+        gas_consumed: None,
+    }));
 
-    // The address already exists, which should cause the contract to revert
+    // code hash for a given contract address
+    actions.push(Call {
+        origin: TestAddress::Alice,
+        dest: TestAddress::Instantiated(0),
+        value: 0,
+        gas_limit: None,
+        storage_deposit_limit: None,
+        data: Contract::ext_code_hash(Address::from(addr.to_fixed_bytes())).calldata,
+    });
+    actions.push(VerifyCall(VerifyCallExpectation {
+        success: true,
+        output: OptionalHex::from(code_hash.as_bytes().to_vec()),
+        gas_consumed: None,
+    }));
 
-    let (_, output) = state
+    // EOA returns fixed hash
-        .transaction()
+    actions.push(Call {
-        .with_default_account(&contract.pvm_runtime)
+        origin: TestAddress::Alice,
-        .calldata(contract.calldata)
+        dest: TestAddress::Instantiated(0),
-        .call();
+        value: 0,
+        gas_limit: None,
+        storage_deposit_limit: None,
+        data: Contract::ext_code_hash(Address::from(CHARLIE.to_fixed_bytes())).calldata,
+    });
+    actions.push(VerifyCall(VerifyCallExpectation {
+        success: true,
+        output: OptionalHex::from(
+            hex!("c5d2460186f7233c927e7db2dcc703c0e500b653ca82273b7bfad8045d85a470").to_vec(),
+        ),
+        gas_consumed: None,
+    }));
 
-    assert_eq!(output.flags, ReturnFlags::Revert);
+    // non-existing account
-}
+    actions.push(Call {
+        origin: TestAddress::Alice,
+        dest: TestAddress::Instantiated(0),
+        value: 0,
+        gas_limit: None,
+        storage_deposit_limit: None,
+        data: Contract::ext_code_hash(Address::from([8u8; 20])).calldata,
+    });
+    actions.push(VerifyCall(VerifyCallExpectation {
+        success: true,
+        output: OptionalHex::from([0u8; 32].to_vec()),
+        gas_consumed: None,
+    }));
 
-
+    Specs {
-#[test]
+        actions,
-fn balance() {
+        ..Default::default()
-    let (_, output) = assert_success(&Contract::value_balance_of(Default::default()), false);
+    }
-
+    .run();
-    let expected = U256::ZERO;
-    let received = U256::from_be_slice(&output.data);
-    assert_eq!(expected, received);
-
-    let expected = U256::from(54589);
-    let (mut state, address) = State::new_deployed(Contract::value_balance_of(Default::default()));
-    state.accounts_mut().get_mut(&address).unwrap().value = expected;
-
-    let contract = Contract::value_balance_of(address);
-    let (_, output) = state
-        .transaction()
-        .with_default_account(&contract.pvm_runtime)
-        .calldata(contract.calldata)
-        .call();
-
-    assert_eq!(ReturnFlags::Success, output.flags);
-
-    let received = U256::from_be_slice(&output.data);
-    assert_eq!(expected, received)
 }
 
 #[test]
 fn ext_code_size() {
-    let contract = Contract::ext_code_size(Transaction::default_address());
+    let alice = Address::from(ALICE.0);
-    let (_, output) = assert_success(&contract, false);
+    let own_address = alice.create(0);
-    let received = U256::from_be_slice(&output.data);
+    let baseline_address = alice.create2([0u8; 32], keccak256(Contract::baseline().pvm_runtime));
-    let expected = U256::from(contract.pvm_runtime.len());
-    assert_eq!(received, expected);
 
-    let contract = Contract::ext_code_size(Default::default());
+    let own_code_size = U256::from(
-    let (_, output) = assert_success(&contract, false);
+        Contract::ext_code_size(Default::default())
-    let received = U256::from_be_slice(&output.data);
+            .pvm_runtime
-    let expected = U256::ZERO;
+            .len(),
-    assert_eq!(received, expected);
+    );
+    let baseline_code_size = U256::from(Contract::baseline().pvm_runtime.len());
+
+    Specs {
+        actions: vec![
+            // Instantiate the test contract
+            instantiate("contracts/ExtCode.sol", "ExtCode").remove(0),
+            // Instantiate the baseline contract
+            Instantiate {
+                origin: TestAddress::Alice,
+                value: 0,
+                gas_limit: Some(GAS_LIMIT),
+                storage_deposit_limit: None,
+                code: Code::Solidity {
+                    path: Some("contracts/Baseline.sol".into()),
+                    contract: "Baseline".to_string(),
+                    solc_optimizer: None,
+                },
+                data: vec![],
+                salt: OptionalHex::from([0; 32]),
+            },
+            // Alice is not a contract and returns a code size of 0
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value: 0,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::ext_code_size(alice).calldata,
+            },
+            VerifyCall(VerifyCallExpectation {
+                success: true,
+                output: OptionalHex::from([0u8; 32].to_vec()),
+                gas_consumed: None,
+            }),
+            // Unknown address returns a code size of 0
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value: 0,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::ext_code_size(Address::from([0xff; 20])).calldata,
+            },
+            VerifyCall(VerifyCallExpectation {
+                success: true,
+                output: OptionalHex::from([0u8; 32].to_vec()),
+                gas_consumed: None,
+            }),
+            // Own address via extcodesize returns own code size
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value: 0,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::ext_code_size(own_address).calldata,
+            },
+            VerifyCall(VerifyCallExpectation {
+                success: true,
+                output: OptionalHex::from(own_code_size.to_be_bytes::<32>().to_vec()),
+                gas_consumed: None,
+            }),
+            // Own address via codesize returns own code size
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value: 0,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::code_size().calldata,
+            },
+            VerifyCall(VerifyCallExpectation {
+                success: true,
+                output: OptionalHex::from(own_code_size.to_be_bytes::<32>().to_vec()),
+                gas_consumed: None,
+            }),
+            // Baseline address returns the baseline code size
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value: 0,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::ext_code_size(baseline_address).calldata,
+            },
+            VerifyCall(VerifyCallExpectation {
+                success: true,
+                output: OptionalHex::from(baseline_code_size.to_be_bytes::<32>().to_vec()),
+                gas_consumed: None,
+            }),
+        ],
+        ..Default::default()
+    }
+    .run();
 }
 
 #[test]
-fn code_size() {
+#[should_panic(expected = "ReentranceDenied")]
-    let contract = Contract::code_size();
+fn send_denies_reentrancy() {
-    let (_, output) = assert_success(&contract, false);
+    let value = 1000;
-    let expected = U256::from(contract.pvm_runtime.len());
+    Specs {
-    let received = U256::from_be_slice(&output.data);
+        actions: vec![
-    assert_eq!(expected, received);
+            instantiate("contracts/Send.sol", "Send").remove(0),
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::send_self(U256::from(value)).calldata,
+            },
+        ],
+        differential: false,
+        ..Default::default()
+    }
+    .run();
+}
+
+#[test]
+#[should_panic(expected = "ReentranceDenied")]
+fn transfer_denies_reentrancy() {
+    let value = 1000;
+    Specs {
+        actions: vec![
+            instantiate("contracts/Transfer.sol", "Transfer").remove(0),
+            Call {
+                origin: TestAddress::Alice,
+                dest: TestAddress::Instantiated(0),
+                value,
+                gas_limit: None,
+                storage_deposit_limit: None,
+                data: Contract::transfer_self(U256::from(value)).calldata,
+            },
+        ],
+        differential: false,
+        ..Default::default()
+    }
+    .run();
 }
-*/

crates/linker/Cargo.toml

@@ -7,14 +7,9 @@ repository.workspace = true
 authors.workspace = true
 description = "revive compiler linker utils"
 
-[features]
-riscv-64 = []
-
 [dependencies]
-inkwell = { workspace = true }
 tempfile = { workspace = true }
 polkavm-linker = { workspace = true }
-polkavm-common = { workspace = true }
 libc = { workspace = true }
 anyhow = { workspace = true }
 

crates/linker/src/lib.rs

@@ -8,14 +8,7 @@ SECTIONS {
     .text : { KEEP(*(.text.polkavm_export)) *(.text .text.*) }
 }"#;
 
-#[cfg(not(feature = "riscv-64"))]
-const BUILTINS_ARCHIVE_FILE: &str = "libclang_rt.builtins-riscv32.a";
-#[cfg(feature = "riscv-64")]
 const BUILTINS_ARCHIVE_FILE: &str = "libclang_rt.builtins-riscv64.a";
-
-#[cfg(not(feature = "riscv-64"))]
-const BUILTINS_LIB_NAME: &str = "clang_rt.builtins-riscv32";
-#[cfg(feature = "riscv-64")]
 const BUILTINS_LIB_NAME: &str = "clang_rt.builtins-riscv64";
 
 fn invoke_lld(cmd_args: &[&str]) -> bool {
@@ -29,9 +22,9 @@ fn invoke_lld(cmd_args: &[&str]) -> bool {
     unsafe { LLDELFLink(args.as_ptr(), args.len()) == 0 }
 }
 
-fn polkavm_linker<T: AsRef<[u8]>>(code: T) -> anyhow::Result<Vec<u8>> {
+pub fn polkavm_linker<T: AsRef<[u8]>>(code: T, strip_binary: bool) -> anyhow::Result<Vec<u8>> {
     let mut config = polkavm_linker::Config::default();
-    config.set_strip(true);
+    config.set_strip(strip_binary);
     config.set_optimize(true);
 
     polkavm_linker::program_from_elf(config, code.as_ref())
@@ -64,6 +57,7 @@ pub fn link<T: AsRef<[u8]>>(input: T) -> anyhow::Result<Vec<u8>> {
         "--relocatable",
         "--emit-relocs",
         "--no-relax",
+        "--unique",
         "--gc-sections",
         "--library-path",
         dir.path().to_str().expect("should be utf8"),
@@ -79,10 +73,5 @@ pub fn link<T: AsRef<[u8]>>(input: T) -> anyhow::Result<Vec<u8>> {
         return Err(anyhow::anyhow!("ld.lld failed"));
     }
 
-    if env::var("PVM_LINKER_DUMP_SO").is_ok() {
+    Ok(fs::read(&output_path)?)
-        fs::copy(&output_path, "/tmp/out.so")?;
-    };
-
-    let blob = fs::read(&output_path)?;
-    polkavm_linker(blob)
 }

crates/lld-sys/Cargo.toml

@@ -13,3 +13,4 @@ libc = { workspace = true }
 
 [build-dependencies]
 cc = { workspace = true }
+revive-build-utils = { workspace = true }

crates/lld-sys/build.rs

@@ -1,17 +1,16 @@
-fn llvm_config(arg: &str) -> String {
-    let output = std::process::Command::new("llvm-config")
-        .args([arg])
-        .output()
-        .unwrap_or_else(|_| panic!("`llvm-config {arg}` failed"));
-
-    String::from_utf8(output.stdout)
-        .unwrap_or_else(|_| panic!("output of `llvm-config {arg}` should be utf8"))
-}
-
 fn set_rustc_link_flags() {
-    println!("cargo:rustc-link-search=native={}", llvm_config("--libdir"));
+    let llvm_lib_path = match std::env::var(revive_build_utils::REVIVE_LLVM_TARGET_PREFIX) {
+        Ok(path) => std::path::PathBuf::from(path).join("lib"),
+        _ => revive_build_utils::llvm_lib_dir(),
+    };
+
+    println!(
+        "cargo:rustc-link-search=native={}",
+        llvm_lib_path.to_string_lossy()
+    );
 
     for lib in [
+        // These are required by ld.lld
         "lldELF",
         "lldCommon",
         "lldMachO",
@@ -22,19 +21,79 @@ fn set_rustc_link_flags() {
         "LLVMTargetParser",
         "LLVMBinaryFormat",
         "LLVMDemangle",
+        // The `llvm-sys` crate relies on `llvm-config` to obtain a list of required LLVM libraries
+        // during the build process. This works well in typical native environments, where `llvm-config`
+        // can accurately list the necessary libraries.
+        // However, when cross-compiling to WebAssembly using Emscripten, `llvm-config` fails to recognize
+        // JavaScript-based libraries, making it necessary to manually inject the required dependencies.
+        "LLVMRISCVDisassembler",
+        "LLVMRISCVAsmParser",
+        "LLVMRISCVCodeGen",
+        "LLVMRISCVDesc",
+        "LLVMRISCVInfo",
+        "LLVMExecutionEngine",
+        "LLVMOption",
+        "LLVMMCDisassembler",
+        "LLVMPasses",
+        "LLVMHipStdPar",
+        "LLVMCFGuard",
+        "LLVMCoroutines",
+        "LLVMipo",
+        "LLVMVectorize",
+        "LLVMInstrumentation",
+        "LLVMFrontendOpenMP",
+        "LLVMFrontendOffloading",
+        "LLVMGlobalISel",
+        "LLVMAsmPrinter",
+        "LLVMSelectionDAG",
+        "LLVMCodeGen",
+        "LLVMTarget",
+        "LLVMObjCARCOpts",
+        "LLVMCodeGenTypes",
+        "LLVMIRPrinter",
+        "LLVMScalarOpts",
+        "LLVMInstCombine",
+        "LLVMAggressiveInstCombine",
+        "LLVMTransformUtils",
+        "LLVMBitWriter",
+        "LLVMAnalysis",
+        "LLVMProfileData",
+        "LLVMDebugInfoDWARF",
+        "LLVMObject",
+        "LLVMMCParser",
+        "LLVMIRReader",
+        "LLVMAsmParser",
+        "LLVMMC",
+        "LLVMDebugInfoCodeView",
+        "LLVMBitReader",
+        "LLVMRemarks",
+        "LLVMBitstreamReader",
     ] {
         println!("cargo:rustc-link-lib=static={lib}");
     }
 
-    #[cfg(target_os = "linux")]
+    let target_os = std::env::var("CARGO_CFG_TARGET_OS").unwrap_or_default();
-    {
+    let target_env = std::env::var("CARGO_CFG_TARGET_ENV").unwrap_or_default();
-        println!("cargo:rustc-link-lib=dylib=stdc++");
+    if target_os == "linux" {
-        println!("cargo:rustc-link-lib=tinfo");
+        if target_env == "musl" {
+            println!("cargo:rustc-link-lib=static=c++");
+        } else {
+            println!("cargo:rustc-link-lib=dylib=stdc++");
+        }
     }
 }
 
 fn main() {
-    llvm_config("--cxxflags")
+    println!(
+        "cargo:rerun-if-env-changed={}",
+        revive_build_utils::REVIVE_LLVM_HOST_PREFIX
+    );
+    println!(
+        "cargo:rerun-if-env-changed={}",
+        revive_build_utils::REVIVE_LLVM_TARGET_PREFIX
+    );
+
+    revive_build_utils::llvm_cxx_flags()
         .split_whitespace()
         .fold(&mut cc::Build::new(), |builder, flag| builder.flag(flag))
         .flag("-Wno-unused-parameter")

crates/llvm-builder/Cargo.toml

@@ -0,0 +1,38 @@
+[package]
+name = "revive-llvm-builder"
+description = "revive LLVM compiler framework builder"
+authors = [
+    "Oleksandr Zarudnyi <a.zarudnyy@matterlabs.dev>",
+    "Anton Baliasnikov <aba@matterlabs.dev>",
+    "Cyrill Leutwiler <cyrill@parity.io>",
+]
+version.workspace = true
+license.workspace = true
+edition.workspace = true
+repository.workspace = true
+
+[[bin]]
+name = "revive-llvm"
+path = "src/revive_llvm/main.rs"
+
+[lib]
+doctest = false
+
+[dependencies]
+clap = { workspace = true, features = ["help", "std", "derive"] }
+anyhow = { workspace = true }
+serde = { workspace = true, features = [ "derive" ] }
+toml = { workspace = true }
+num_cpus = { workspace = true }
+fs_extra = { workspace = true }
+path-slash = { workspace = true }
+regex = { workspace = true }
+downloader = { workspace = true }
+tar = { workspace = true }
+flate2 = { workspace = true }
+env_logger = { workspace = true }
+log = { workspace = true }
+
+[dev-dependencies]
+assert_cmd = { workspace = true }
+assert_fs = { workspace = true }

crates/llvm-builder/README.md

@@ -0,0 +1,132 @@
+# revive LLVM builder
+
+Parity fork of the [Matter Labs zksync LLVM builder](https://github.com/matter-labs/era-compiler-llvm-builder) helper utility for compiling [revive](https://github.com/paritytech/revive) compatible LLVM builds.
+
+## Installation and usage
+
+The LLVM compiler framework for revive must be built with our tool called `revive-llvm`.
+This is because the revive compiler has requirements not fullfilled in upstream builds:
+- Special builds for compiling the frontend into statically linked ELF binaries and also Wasm executables
+- The RISC-V target (the PolkaVM target)
+- The compiler-rt builtins for the PolkaVM target
+- We want to leave the assertions always on
+- Various other specific configurations and optimization may be applied
+
+Obtain a compatible build for your host platform from the release section of this repository (TODO). Alternatively follow below steps to get a custom build:
+
+<details>
+<summary>1. Install the system prerequisites.</summary>
+
+   * Linux (Debian):
+
+      Install the following packages:
+      ```shell
+      apt install cmake ninja-build curl git libssl-dev pkg-config clang lld
+      ```
+   * Linux (Arch):
+
+      Install the following packages:
+      ```shell
+      pacman -Syu which cmake ninja curl git pkg-config clang lld
+      ```
+
+   * MacOS:
+
+      * Install the [HomeBrew](https://brew.sh) package manager.
+      * Install the following packages:
+
+         ```shell
+         brew install cmake ninja coreutils
+         ```
+
+      * Install your choice of a recent LLVM/[Clang](https://clang.llvm.org) compiler, e.g. via [Xcode](https://developer.apple.com/xcode/), [Apple’s Command Line Tools](https://developer.apple.com/library/archive/technotes/tn2339/_index.html), or your preferred package manager.
+</details>
+
+<details>
+<summary>2. Install Rust.</summary>
+
+   * Follow the latest [official instructions](https://www.rust-lang.org/tools/install:
+      ```shell
+      curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
+      . ${HOME}/.cargo/env
+      ```
+
+      > Currently we are not pinned to any specific version of Rust, so just install the latest stable build for your   platform.
+</details>
+
+<details>
+<summary>3. Install the revive LLVM framework builder.</summary>
+
+   * Install the builder using `cargo`:
+      ```shell
+      cargo install --git https://github.com/paritytech/revive-llvm-builder --force --locked
+      ```
+
+      > The builder is not the LLVM framework itself, but a tool that clones its repository and runs a sequence of build commands. By default it is installed in `~/.cargo/bin/`, which is recommended to be added to your `$PATH`.
+
+</details>
+
+<details>
+<summary>4. (Optional) Create the `LLVM.lock` file.</summary>
+
+   * The `LLVM.lock` dictates the LLVM source tree being used.
+     A default `./LLVM.lock` pointing to the release used for development is already provided.
+
+</details>
+
+<details>
+<summary>5. Build LLVM.</summary>
+
+   * Clone and build the LLVM framework using the `revive-llvm` tool.
+
+     The clang and lld projects are required for the `resolc` Solidity frontend executable; they are enabled by default. LLVM assertions are also enabled by default.
+
+      ```shell
+      revive-llvm clone
+      revive-llvm build --llvm-projects lld --llvm-projects clang
+      ```
+
+      Build artifacts end up in the `./target-llvm/gnu/target-final/` directory by default.
+      The `gnu` directory depends on the supported archticture and will either be `gnu`, `musl` or `emscripten`.
+      You now need to export the final target directory `$LLVM_SYS_181_PREFIX`: `export LLVM_SYS_181_PREFIX=${PWD}/target-llvm/gnu/target-final`
+      If built with the `--enable-tests` option, test tools will be in the `./target-llvm/gnu/build-final/` directory, along with copies of the build artifacts. For all supported build options, run `revive-llvm build --help`.
+
+</details>
+
+## Supported target architectures
+
+The following target platforms are supported:
+- Linux GNU (x86)
+- Linux MUSL (x86)
+- MacOS (aarch64)
+- Windows GNU (x86)
+- Emscripten (wasm32)
+
+<details>
+<summary>Building for MUSL</summary>
+
+   * Via a musl build we can build revive into fully static ELF binaries.
+     Which is desirable for reproducible Solidity contracts builds.
+     The resulting binary is also very portable, akin to the`solc` frontend binary distribution.
+
+     Clone and build the LLVM framework using the `revive-llvm` tool:
+      ```shell
+      revive-llvm --target-env musl clone
+      revive-llvm --target-env musl build --enable-assertions --llvm-projects clang --llvm-projects lld 
+      ```
+
+</details>
+
+<details>
+<summary>Building for Emscripten</summary>
+
+   * Via an emsdk build we can run revive in the browser and on node.js.
+
+     Clone and build the LLVM framework using the `revive-llvm` tool:
+      ```shell
+      revive-llvm --target-env emscripten clone
+      revive-llvm --target-env emscripten build --enable-assertions --llvm-projects clang --llvm-projects lld 
+      ```
+
+</details>
+

crates/llvm-builder/src/build_type.rs

@@ -0,0 +1,39 @@
+//! The revive LLVM build type.
+
+/// The revive LLVM build type.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum BuildType {
+    /// The debug build.
+    Debug,
+    /// The release build.
+    Release,
+    /// The release with debug info build.
+    RelWithDebInfo,
+    /// The minimal size release build.
+    MinSizeRel,
+}
+
+impl std::str::FromStr for BuildType {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value {
+            "Debug" => Ok(Self::Debug),
+            "Release" => Ok(Self::Release),
+            "RelWithDebInfo" => Ok(Self::RelWithDebInfo),
+            "MinSizeRel" => Ok(Self::MinSizeRel),
+            value => Err(format!("Unsupported build type: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for BuildType {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Self::Debug => write!(f, "Debug"),
+            Self::Release => write!(f, "Release"),
+            Self::RelWithDebInfo => write!(f, "RelWithDebInfo"),
+            Self::MinSizeRel => write!(f, "MinSizeRel"),
+        }
+    }
+}

crates/llvm-builder/src/builtins.rs

@@ -0,0 +1,137 @@
+//! Utilities for compiling the LLVM compiler-rt builtins.
+
+/// Static CFLAGS variable passed to the compiler building the compiler-rt builtins.
+const C_FLAGS: [&str; 6] = [
+    "--target=riscv64",
+    "-march=rv64emac",
+    "-mabi=lp64e",
+    "-mcpu=generic-rv64",
+    "-nostdlib",
+    "-nodefaultlibs",
+];
+
+/// Static CMAKE arguments for building the compiler-rt builtins.
+const CMAKE_STATIC_ARGS: [&str; 14] = [
+    "-DCOMPILER_RT_BUILD_BUILTINS='On'",
+    "-DCOMPILER_RT_BUILD_LIBFUZZER='Off'",
+    "-DCOMPILER_RT_BUILD_MEMPROF='Off'",
+    "-DCOMPILER_RT_BUILD_PROFILE='Off'",
+    "-DCOMPILER_RT_BUILD_SANITIZERS='Off'",
+    "-DCOMPILER_RT_BUILD_XRAY='Off'",
+    "-DCOMPILER_RT_DEFAULT_TARGET_ONLY='On'",
+    "-DCOMPILER_RT_BAREMETAL_BUILD='On'",
+    "-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
+    "-DCMAKE_EXPORT_COMPILE_COMMANDS='On'",
+    "-DCMAKE_SYSTEM_NAME='unknown'",
+    "-DCMAKE_C_COMPILER_TARGET='riscv64'",
+    "-DCMAKE_ASM_COMPILER_TARGET='riscv64'",
+    "-DCMAKE_CXX_COMPILER_TARGET='riscv64'",
+];
+
+/// Dynamic cmake arguments for building the compiler-rt builtins.
+fn cmake_dynamic_args(
+    build_type: crate::BuildType,
+    target_env: crate::target_env::TargetEnv,
+) -> anyhow::Result<[String; 13]> {
+    let llvm_compiler_rt_target = crate::LLVMPath::llvm_target_compiler_rt()?;
+
+    // The Emscripten target needs to use the host LLVM tools.
+    let llvm_target_host = if target_env == crate::target_env::TargetEnv::Emscripten {
+        crate::LLVMPath::llvm_build_host()?
+    } else {
+        crate::LLVMPath::llvm_target_final()?
+    };
+
+    let mut clang_path = llvm_target_host.to_path_buf();
+    clang_path.push("bin/clang");
+
+    let mut clangxx_path = llvm_target_host.to_path_buf();
+    clangxx_path.push("bin/clang++");
+
+    let mut llvm_config_path = llvm_target_host.to_path_buf();
+    llvm_config_path.push("bin/llvm-config");
+
+    let mut ar_path = llvm_target_host.to_path_buf();
+    ar_path.push("bin/llvm-ar");
+
+    let mut nm_path = llvm_target_host.to_path_buf();
+    nm_path.push("bin/llvm-nm");
+
+    let mut ranlib_path = llvm_target_host.to_path_buf();
+    ranlib_path.push("bin/llvm-ranlib");
+
+    let mut linker_path = llvm_target_host.to_path_buf();
+    linker_path.push("bin/ld.lld");
+
+    Ok([
+        format!(
+            "-DCMAKE_INSTALL_PREFIX='{}'",
+            llvm_compiler_rt_target.to_string_lossy().as_ref(),
+        ),
+        format!("-DCMAKE_BUILD_TYPE='{build_type}'"),
+        format!(
+            "-DCOMPILER_RT_TEST_COMPILER='{}'",
+            clang_path.to_string_lossy()
+        ),
+        format!("-DCMAKE_C_FLAGS='{}'", C_FLAGS.join(" ")),
+        format!("-DCMAKE_ASM_FLAGS='{}'", C_FLAGS.join(" ")),
+        format!("-DCMAKE_CXX_FLAGS='{}'", C_FLAGS.join(" ")),
+        format!("-DCMAKE_C_COMPILER='{}'", clang_path.to_string_lossy()),
+        format!("-DCMAKE_ASM_COMPILER='{}'", clang_path.to_string_lossy()),
+        format!("-DCMAKE_CXX_COMPILER='{}'", clangxx_path.to_string_lossy()),
+        format!("-DCMAKE_AR='{}'", ar_path.to_string_lossy()),
+        format!("-DCMAKE_NM='{}'", nm_path.to_string_lossy()),
+        format!("-DCMAKE_RANLIB='{}'", ranlib_path.to_string_lossy()),
+        format!(
+            "-DLLVM_CONFIG_PATH='{}'",
+            llvm_config_path.to_string_lossy()
+        ),
+    ])
+}
+
+/// Build the compiler-rt builtins library.
+pub fn build(
+    build_type: crate::BuildType,
+    target_env: crate::target_env::TargetEnv,
+    default_target: Option<crate::TargetTriple>,
+    extra_args: &[String],
+    ccache_variant: Option<crate::ccache_variant::CcacheVariant>,
+    sanitizer: Option<crate::sanitizer::Sanitizer>,
+) -> anyhow::Result<()> {
+    log::info!("building compiler-rt for rv64emac");
+
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_compiler_rt = crate::LLVMPath::llvm_module_compiler_rt()?;
+    let llvm_compiler_rt_build = crate::LLVMPath::llvm_build_compiler_rt()?;
+
+    crate::utils::command(
+        std::process::Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_compiler_rt.to_string_lossy().as_ref(),
+                "-B",
+                llvm_compiler_rt_build.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+            ])
+            .args(CMAKE_STATIC_ARGS)
+            .args(cmake_dynamic_args(build_type, target_env)?)
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            )),
+        "LLVM compiler-rt building cmake",
+    )?;
+
+    crate::utils::ninja(&llvm_compiler_rt_build)?;
+
+    Ok(())
+}

crates/llvm-builder/src/ccache_variant.rs

@@ -0,0 +1,31 @@
+//! Compiler cache variants.
+
+/// The list compiler cache variants to be used as constants.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum CcacheVariant {
+    /// Standard ccache.
+    Ccache,
+    /// Mozilla's sccache.
+    Sccache,
+}
+
+impl std::str::FromStr for CcacheVariant {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value {
+            "ccache" => Ok(Self::Ccache),
+            "sccache" => Ok(Self::Sccache),
+            value => Err(format!("Unsupported ccache variant: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for CcacheVariant {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        match self {
+            Self::Ccache => write!(f, "ccache"),
+            Self::Sccache => write!(f, "sccache"),
+        }
+    }
+}

crates/llvm-builder/src/lib.rs

@@ -0,0 +1,340 @@
+//! The revive LLVM builder library.
+
+pub mod build_type;
+pub mod builtins;
+pub mod ccache_variant;
+pub mod llvm_path;
+pub mod llvm_project;
+pub mod lock;
+pub mod platforms;
+pub mod sanitizer;
+pub mod target_env;
+pub mod target_triple;
+pub mod utils;
+
+pub use self::build_type::BuildType;
+pub use self::llvm_path::LLVMPath;
+pub use self::lock::Lock;
+pub use self::platforms::Platform;
+
+use std::collections::HashSet;
+use std::path::{Path, PathBuf};
+use std::process::Command;
+pub use target_env::TargetEnv;
+pub use target_triple::TargetTriple;
+
+/// Executes the LLVM repository cloning.
+pub fn clone(lock: Lock, deep: bool, target_env: TargetEnv) -> anyhow::Result<()> {
+    utils::check_presence("git")?;
+
+    if target_env == TargetEnv::Emscripten {
+        utils::install_emsdk()?;
+    }
+
+    let destination_path = PathBuf::from(LLVMPath::DIRECTORY_LLVM_SOURCE);
+    if destination_path.exists() {
+        log::warn!(
+            "LLVM repository directory {} already exists, falling back to checkout",
+            destination_path.display()
+        );
+        return checkout(lock, false);
+    }
+
+    let mut clone_args = vec!["clone", "--branch", lock.branch.as_str()];
+    if !deep {
+        clone_args.push("--depth");
+        clone_args.push("1");
+    }
+
+    utils::command(
+        Command::new("git")
+            .args(clone_args)
+            .arg(lock.url.as_str())
+            .arg(destination_path.to_string_lossy().as_ref()),
+        "LLVM repository cloning",
+    )?;
+
+    if let Some(r#ref) = lock.r#ref {
+        utils::command(
+            Command::new("git")
+                .args(["checkout", r#ref.as_str()])
+                .current_dir(destination_path.to_string_lossy().as_ref()),
+            "LLVM repository commit checking out",
+        )?;
+    }
+
+    Ok(())
+}
+
+/// Executes the checkout of the specified branch.
+pub fn checkout(lock: Lock, force: bool) -> anyhow::Result<()> {
+    let destination_path = PathBuf::from(LLVMPath::DIRECTORY_LLVM_SOURCE);
+
+    utils::command(
+        Command::new("git")
+            .current_dir(destination_path.as_path())
+            .args(["fetch", "--all", "--tags"]),
+        "LLVM repository data fetching",
+    )?;
+
+    if force {
+        utils::command(
+            Command::new("git")
+                .current_dir(destination_path.as_path())
+                .args(["clean", "-d", "-x", "--force"]),
+            "LLVM repository cleaning",
+        )?;
+    }
+
+    utils::command(
+        Command::new("git")
+            .current_dir(destination_path.as_path())
+            .args(["checkout", "--force", lock.branch.as_str()]),
+        "LLVM repository data pulling",
+    )?;
+
+    if let Some(r#ref) = lock.r#ref {
+        let mut checkout_command = Command::new("git");
+        checkout_command.current_dir(destination_path.as_path());
+        checkout_command.arg("checkout");
+        if force {
+            checkout_command.arg("--force");
+        }
+        checkout_command.arg(r#ref);
+        utils::command(&mut checkout_command, "LLVM repository checking out")?;
+    }
+
+    Ok(())
+}
+
+/// Executes the building of the LLVM framework for the platform determined by the cfg macro.
+/// Since cfg is evaluated at compile time, overriding the platform with a command-line
+/// argument is not possible. So for cross-platform testing, comment out all but the
+/// line to be tested, and perhaps also checks in the platform-specific build method.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    target_env: TargetEnv,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<llvm_project::LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<ccache_variant::CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<sanitizer::Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    log::trace!("build type: {:?}", build_type);
+    log::trace!("target env: {:?}", target_env);
+    log::trace!("targets: {:?}", targets);
+    log::trace!("llvm projects: {:?}", llvm_projects);
+    log::trace!("enable rtti: {:?}", enable_rtti);
+    log::trace!("default target: {:?}", default_target);
+    log::trace!("eneable tests: {:?}", enable_tests);
+    log::trace!("enable_coverage: {:?}", enable_coverage);
+    log::trace!("extra args: {:?}", extra_args);
+    log::trace!("sanitzer: {:?}", sanitizer);
+    log::trace!("enable valgrind: {:?}", enable_valgrind);
+
+    if !PathBuf::from(LLVMPath::DIRECTORY_LLVM_SOURCE).exists() {
+        log::error!(
+            "LLVM project source directory {} does not exist (run `revive-llvm --target-env {} clone`)",
+            LLVMPath::DIRECTORY_LLVM_SOURCE,
+            target_env
+        )
+    }
+
+    std::fs::create_dir_all(llvm_path::DIRECTORY_LLVM_TARGET.get().unwrap())?;
+
+    if cfg!(target_arch = "x86_64") {
+        if cfg!(target_os = "linux") {
+            if target_env == TargetEnv::MUSL {
+                platforms::x86_64_linux_musl::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else if target_env == TargetEnv::GNU {
+                platforms::x86_64_linux_gnu::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else if target_env == TargetEnv::Emscripten {
+                platforms::wasm32_emscripten::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else {
+                anyhow::bail!("Unsupported target environment for x86_64 and Linux");
+            }
+        } else if cfg!(target_os = "macos") {
+            platforms::x86_64_macos::build(
+                build_type,
+                targets,
+                llvm_projects,
+                enable_rtti,
+                default_target,
+                enable_tests,
+                enable_coverage,
+                extra_args,
+                ccache_variant,
+                enable_assertions,
+                sanitizer,
+            )?;
+        } else if cfg!(target_os = "windows") {
+            platforms::x86_64_windows_gnu::build(
+                build_type,
+                targets,
+                llvm_projects,
+                enable_rtti,
+                default_target,
+                enable_tests,
+                enable_coverage,
+                extra_args,
+                ccache_variant,
+                enable_assertions,
+                sanitizer,
+            )?;
+        } else {
+            anyhow::bail!("Unsupported target OS for x86_64");
+        }
+    } else if cfg!(target_arch = "aarch64") {
+        if cfg!(target_os = "linux") {
+            if target_env == TargetEnv::MUSL {
+                platforms::aarch64_linux_musl::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else if target_env == TargetEnv::GNU {
+                platforms::aarch64_linux_gnu::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else {
+                anyhow::bail!("Unsupported target environment for aarch64 and Linux");
+            }
+        } else if cfg!(target_os = "macos") {
+            if target_env == TargetEnv::Emscripten {
+                platforms::wasm32_emscripten::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                    enable_valgrind,
+                )?;
+            } else {
+                platforms::aarch64_macos::build(
+                    build_type,
+                    targets,
+                    llvm_projects,
+                    enable_rtti,
+                    default_target,
+                    enable_tests,
+                    enable_coverage,
+                    extra_args,
+                    ccache_variant,
+                    enable_assertions,
+                    sanitizer,
+                )?;
+            }
+        } else {
+            anyhow::bail!("Unsupported target OS for aarch64");
+        }
+    } else {
+        anyhow::bail!("Unsupported target architecture");
+    }
+
+    crate::builtins::build(
+        build_type,
+        target_env,
+        default_target,
+        extra_args,
+        ccache_variant,
+        sanitizer,
+    )?;
+
+    Ok(())
+}
+
+/// Executes the build artifacts cleaning.
+pub fn clean() -> anyhow::Result<()> {
+    let remove_if_exists = |path: &Path| {
+        if !path.exists() {
+            return Ok(());
+        }
+        log::info!("deleting {}", path.display());
+        std::fs::remove_dir_all(path)
+    };
+
+    remove_if_exists(
+        llvm_path::DIRECTORY_LLVM_TARGET
+            .get()
+            .expect("target_env is always set because of the default value")
+            .parent()
+            .expect("target_env parent directory is target-llvm"),
+    )?;
+    remove_if_exists(&PathBuf::from(LLVMPath::DIRECTORY_EMSDK_SOURCE))?;
+    remove_if_exists(&PathBuf::from(LLVMPath::DIRECTORY_LLVM_SOURCE))?;
+    remove_if_exists(&PathBuf::from(LLVMPath::DIRECTORY_LLVM_HOST_SOURCE))?;
+
+    Ok(())
+}

crates/llvm-builder/src/llvm_path.rs

@@ -0,0 +1,151 @@
+//! The revive LLVM builder constants.
+
+use std::path::PathBuf;
+use std::sync::OnceLock;
+
+pub static DIRECTORY_LLVM_TARGET: OnceLock<PathBuf> = OnceLock::new();
+
+/// The LLVM path resolver.
+pub struct LLVMPath {}
+
+impl LLVMPath {
+    /// The LLVM source directory.
+    pub const DIRECTORY_LLVM_SOURCE: &'static str = "./llvm/";
+
+    /// The LLVM host source directory for stage 1 of multistage MUSL and Emscripten builds.
+    ///
+    /// We use upstream LLVM anyways; re-use the same tree for host and target builds.
+    pub const DIRECTORY_LLVM_HOST_SOURCE: &'static str = Self::DIRECTORY_LLVM_SOURCE;
+
+    /// The Emscripten SDK source directory.
+    pub const DIRECTORY_EMSDK_SOURCE: &'static str = "./emsdk/";
+
+    /// Returns the path to the `llvm` stage 1 host LLVM source module directory.
+    pub fn llvm_host_module_llvm() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(Self::DIRECTORY_LLVM_HOST_SOURCE);
+        path.push("llvm");
+        crate::utils::absolute_path(path).inspect(|absolute_path| {
+            log::debug!(
+                "llvm stage 1 host llvm source module: {}",
+                absolute_path.display()
+            )
+        })
+    }
+
+    /// Returns the path to the `llvm` LLVM source module directory.
+    pub fn llvm_module_llvm() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(Self::DIRECTORY_LLVM_SOURCE);
+        path.push("llvm");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm source module: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the MUSL source.
+    pub fn musl_source(name: &str) -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push(name);
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("musl source: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the MUSL build directory.
+    pub fn musl_build(source_directory: &str) -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push(source_directory);
+        path.push("build");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("musl build: '{}'", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM CRT build directory.
+    pub fn llvm_build_crt() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("build-crt");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm build crt: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM host build directory.
+    pub fn llvm_build_host() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("build-host");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm build host: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM final build directory.
+    pub fn llvm_build_final() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("build-final");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm build final: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the MUSL target directory.
+    pub fn musl_target() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("target-musl");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("musl target: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM CRT target directory.
+    pub fn llvm_target_crt() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("target-crt");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm crt target: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM host target directory.
+    pub fn llvm_target_host() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("target-host");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm host target: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM final target directory.
+    pub fn llvm_target_final() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("target-final");
+        crate::utils::absolute_path(path)
+            .inspect(|absolute_path| log::debug!("llvm final target: {}", absolute_path.display()))
+    }
+
+    /// Returns the path to the LLVM compiler builtin target directory.
+    pub fn llvm_module_compiler_rt() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(Self::DIRECTORY_LLVM_SOURCE);
+        path.push("compiler-rt");
+        crate::utils::absolute_path(path).inspect(|absolute_path| {
+            log::debug!("compiler-rt source dir: {}", absolute_path.display())
+        })
+    }
+
+    /// Returns the path to the LLVM compiler-rt target directory.
+    pub fn llvm_target_compiler_rt() -> anyhow::Result<PathBuf> {
+        Self::llvm_target_final()
+    }
+
+    /// Returns the path to the LLVM compiler-rt build directory.
+    pub fn llvm_build_compiler_rt() -> anyhow::Result<PathBuf> {
+        let mut path = PathBuf::from(DIRECTORY_LLVM_TARGET.get().unwrap());
+        path.push("build-compiler-rt");
+        crate::utils::absolute_path(path).inspect(|absolute_path| {
+            log::debug!("llvm compiler-rt build: {}", absolute_path.display())
+        })
+    }
+
+    /// Returns the path to the LLVM target final bin path.
+    ///
+    pub fn llvm_target_final_bin(
+        target_env: crate::target_env::TargetEnv,
+    ) -> anyhow::Result<PathBuf> {
+        let mut path = Self::llvm_target_final()?;
+        path.push("bin");
+        path.push(format!("{target_env}"));
+        crate::utils::absolute_path(path).inspect(|absolute_path| {
+            log::debug!("llvm target final bin: {}", absolute_path.display())
+        })
+    }
+}

crates/llvm-builder/src/llvm_project.rs

@@ -0,0 +1,39 @@
+//! The LLVM projects to enable during the build.
+
+/// The list of LLVM projects used as constants.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum LLVMProject {
+    /// The Clang compiler.
+    CLANG,
+    /// LLD, the LLVM linker.
+    LLD,
+    /// The LLVM debugger.
+    LLDB,
+    /// The MLIR compiler.
+    MLIR,
+}
+
+impl std::str::FromStr for LLVMProject {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value.to_lowercase().as_str() {
+            "clang" => Ok(Self::CLANG),
+            "lld" => Ok(Self::LLD),
+            "lldb" => Ok(Self::LLDB),
+            "mlir" => Ok(Self::MLIR),
+            value => Err(format!("Unsupported LLVM project to enable: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for LLVMProject {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        match self {
+            Self::CLANG => write!(f, "clang"),
+            Self::LLD => write!(f, "lld"),
+            Self::LLDB => write!(f, "lldb"),
+            Self::MLIR => write!(f, "mlir"),
+        }
+    }
+}

crates/llvm-builder/src/lock.rs

@@ -0,0 +1,37 @@
+//! The revive LLVM builder lock file.
+
+use anyhow::Context;
+use std::fs::File;
+use std::io::Read;
+use std::path::PathBuf;
+
+use serde::Deserialize;
+use serde::Serialize;
+
+/// The default lock file location.
+pub const LLVM_LOCK_DEFAULT_PATH: &str = "LLVM.lock";
+
+/// The lock file data.
+///
+/// This file describes the exact reference of the LLVM framework.
+#[derive(Debug, Deserialize, Serialize)]
+pub struct Lock {
+    /// The LLVM repository URL.
+    pub url: String,
+    /// The LLVM repository branch.
+    pub branch: String,
+    /// The LLVM repository commit reference.
+    pub r#ref: Option<String>,
+}
+
+impl TryFrom<&PathBuf> for Lock {
+    type Error = anyhow::Error;
+
+    fn try_from(path: &PathBuf) -> Result<Self, Self::Error> {
+        let mut config_str = String::new();
+        let mut config_file =
+            File::open(path).with_context(|| format!("Error opening {:?} file", path))?;
+        config_file.read_to_string(&mut config_str)?;
+        Ok(toml::from_str(&config_str)?)
+    }
+}

crates/llvm-builder/src/platforms/aarch64_linux_gnu.rs

@@ -0,0 +1,111 @@
+//! The revive LLVM arm64 `linux-gnu` builder.
+
+use std::collections::HashSet;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    crate::utils::check_presence("lld")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_llvm.to_string_lossy().as_ref(),
+                "-B",
+                llvm_build_final.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    llvm_target_final.to_string_lossy().as_ref(),
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                "-DLLVM_USE_LINKER='lld'",
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::GNU,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_valgrind(
+                enable_valgrind,
+            )),
+        "LLVM building cmake",
+    )?;
+    crate::utils::ninja(llvm_build_final.as_ref())?;
+    Ok(())
+}

crates/llvm-builder/src/platforms/aarch64_linux_musl.rs

@@ -0,0 +1,394 @@
+//! The revive LLVM arm64 `linux-musl` builder.
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+use std::collections::HashSet;
+use std::path::Path;
+use std::process::Command;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    crate::utils::check_presence("lld")?;
+    crate::utils::check_presence("ninja")?;
+
+    let musl_name = "musl-1.2.3";
+    let musl_build = LLVMPath::musl_build(musl_name)?;
+    let musl_target = LLVMPath::musl_target()?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_host_module_llvm = LLVMPath::llvm_host_module_llvm()?;
+
+    let llvm_build_crt = LLVMPath::llvm_build_crt()?;
+    let llvm_target_crt = LLVMPath::llvm_target_crt()?;
+
+    let llvm_build_host = LLVMPath::llvm_build_host()?;
+    let llvm_target_host = LLVMPath::llvm_target_host()?;
+
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    if !LLVMPath::musl_source(musl_name)?.exists() {
+        crate::utils::download_musl(musl_name)?;
+    }
+    crate::platforms::shared::build_musl(musl_build.as_path(), musl_target.as_path())?;
+    build_crt(
+        targets.clone(),
+        llvm_host_module_llvm.as_path(),
+        llvm_build_crt.as_path(),
+        llvm_target_crt.as_path(),
+        ccache_variant,
+    )?;
+    build_host(
+        llvm_host_module_llvm.as_path(),
+        llvm_build_host.as_path(),
+        llvm_target_host.as_path(),
+        musl_target.as_path(),
+        llvm_target_crt.as_path(),
+        ccache_variant,
+    )?;
+    build_target(
+        build_type,
+        targets,
+        llvm_projects,
+        enable_rtti,
+        default_target,
+        llvm_module_llvm.as_path(),
+        llvm_build_final.as_path(),
+        llvm_target_final.as_path(),
+        musl_target.as_path(),
+        llvm_target_host.as_path(),
+        enable_tests,
+        enable_coverage,
+        extra_args,
+        ccache_variant,
+        enable_assertions,
+        sanitizer,
+        enable_valgrind,
+    )?;
+
+    Ok(())
+}
+
+///
+/// The `crt` building sequence.
+///
+fn build_crt(
+    mut targets: HashSet<Platform>,
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    ccache_variant: Option<CcacheVariant>,
+) -> anyhow::Result<()> {
+    targets.insert(Platform::AArch64);
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_BUILD_TYPE='Release'",
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                "-DLLVM_ENABLE_PROJECTS='compiler-rt'",
+                format!("-DLLVM_TARGETS_TO_BUILD='{}'", Platform::AArch64).as_str(),
+                "-DLLVM_DEFAULT_TARGET_TRIPLE='aarch64-unknown-linux-musl'",
+                "-DLLVM_BUILD_TESTS='Off'",
+                "-DLLVM_BUILD_RUNTIMES='Off'",
+                "-DLLVM_BUILD_UTILS='Off'",
+                "-DLLVM_INCLUDE_TESTS='Off'",
+                "-DLLVM_INCLUDE_RUNTIMES='Off'",
+                "-DLLVM_INCLUDE_UTILS='Off'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ARCH='aarch64'",
+                "-DCOMPILER_RT_BUILD_CRT='On'",
+                "-DCOMPILER_RT_BUILD_BUILTINS='On'",
+                "-DCOMPILER_RT_BUILD_SANITIZERS='Off'",
+                "-DCOMPILER_RT_BUILD_XRAY='Off'",
+                "-DCOMPILER_RT_BUILD_LIBFUZZER='Off'",
+                "-DCOMPILER_RT_BUILD_PROFILE='Off'",
+                "-DCOMPILER_RT_BUILD_MEMPROF='Off'",
+                "-DCOMPILER_RT_BUILD_ORC='Off'",
+            ])
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            )),
+        "CRT building cmake",
+    )?;
+
+    crate::utils::command(
+        Command::new("ninja")
+            .arg("-C")
+            .arg(build_directory)
+            .arg("install-crt"),
+        "CRT building ninja",
+    )?;
+
+    Ok(())
+}
+
+///
+/// The host toolchain building sequence.
+///
+fn build_host(
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    musl_target_directory: &Path,
+    crt_target_directory: &Path,
+    ccache_variant: Option<CcacheVariant>,
+) -> anyhow::Result<()> {
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DDEFAULT_SYSROOT='{}'",
+                    musl_target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_BUILD_TYPE='Release'",
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                "-DCLANG_DEFAULT_CXX_STDLIB='libc++'",
+                "-DCLANG_DEFAULT_RTLIB='compiler-rt'",
+                "-DLLVM_DEFAULT_TARGET_TRIPLE='aarch64-unknown-linux-musl'",
+                "-DLLVM_TARGETS_TO_BUILD='AArch64'",
+                "-DLLVM_BUILD_TESTS='Off'",
+                "-DLLVM_BUILD_UTILS='Off'",
+                "-DLLVM_INCLUDE_TESTS='Off'",
+                "-DLLVM_INCLUDE_UTILS='Off'",
+                "-DLLVM_ENABLE_PROJECTS='clang;lld'",
+                "-DLLVM_ENABLE_RUNTIMES='compiler-rt;libcxx;libcxxabi;libunwind'",
+                "-DLIBCXX_CXX_ABI='libcxxabi'",
+                "-DLIBCXX_HAS_MUSL_LIBC='On'",
+                "-DLIBCXX_ENABLE_SHARED='Off'",
+                "-DLIBCXX_ENABLE_STATIC='On'",
+                "-DLIBCXX_ENABLE_STATIC_ABI_LIBRARY='On'",
+                "-DLIBCXXABI_ENABLE_SHARED='Off'",
+                "-DLIBCXXABI_ENABLE_STATIC='On'",
+                "-DLIBCXXABI_ENABLE_STATIC_UNWINDER='On'",
+                "-DLIBCXXABI_USE_LLVM_UNWINDER='On'",
+                "-DLIBCXXABI_USE_COMPILER_RT='On'",
+                "-DLIBUNWIND_ENABLE_STATIC='On'",
+                "-DLIBUNWIND_ENABLE_SHARED='Off'",
+                "-DCOMPILER_RT_BUILD_CRT='On'",
+                "-DCOMPILER_RT_BUILD_SANITIZERS='Off'",
+                "-DCOMPILER_RT_BUILD_XRAY='Off'",
+                "-DCOMPILER_RT_BUILD_LIBFUZZER='Off'",
+                "-DCOMPILER_RT_BUILD_PROFILE='On'",
+                "-DCOMPILER_RT_BUILD_MEMPROF='Off'",
+                "-DCOMPILER_RT_BUILD_ORC='Off'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ARCH='aarch64'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ONLY='On'",
+            ])
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            )),
+        "LLVM host building cmake",
+    )?;
+
+    let mut crt_lib_directory = crt_target_directory.to_path_buf();
+    crt_lib_directory.push("lib/");
+
+    let mut build_lib_directory = build_directory.to_path_buf();
+    build_lib_directory.push("lib/");
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        content_only: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy(crt_lib_directory, build_lib_directory, &copy_options)?;
+
+    crate::utils::command(
+        Command::new("ninja")
+            .arg("-C")
+            .arg(build_directory)
+            .arg("install"),
+        "LLVM host building ninja",
+    )?;
+
+    Ok(())
+}
+
+///
+/// The target toolchain building sequence.
+///
+#[allow(clippy::too_many_arguments)]
+fn build_target(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    musl_target_directory: &Path,
+    host_target_directory: &Path,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    let mut clang_path = host_target_directory.to_path_buf();
+    clang_path.push("bin/clang");
+
+    let mut clang_cxx_path = host_target_directory.to_path_buf();
+    clang_cxx_path.push("bin/clang++");
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                "-DBUILD_SHARED_LIBS='Off'",
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                format!("-DCMAKE_C_COMPILER='{}'", clang_path.to_string_lossy()).as_str(),
+                format!(
+                    "-DCMAKE_CXX_COMPILER='{}'",
+                    clang_cxx_path.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_FIND_LIBRARY_SUFFIXES='.a'",
+                "-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
+                "-DCMAKE_EXE_LINKER_FLAGS='-fuse-ld=lld -static'",
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::MUSL,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_valgrind(
+                enable_valgrind,
+            )),
+        "LLVM target building cmake",
+    )?;
+
+    crate::utils::ninja(build_directory)?;
+
+    let mut musl_lib_directory = musl_target_directory.to_path_buf();
+    musl_lib_directory.push("lib/");
+
+    let mut host_lib_directory = host_target_directory.to_path_buf();
+    host_lib_directory.push("lib/aarch64-unknown-linux-musl/");
+
+    let mut target_lib_directory = target_directory.to_path_buf();
+    target_lib_directory.push("lib/");
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        content_only: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy(
+        musl_lib_directory,
+        target_lib_directory.as_path(),
+        &copy_options,
+    )?;
+    fs_extra::dir::copy(
+        host_lib_directory,
+        target_lib_directory.as_path(),
+        &copy_options,
+    )?;
+
+    Ok(())
+}

crates/llvm-builder/src/platforms/aarch64_macos.rs

@@ -0,0 +1,105 @@
+//! The revive LLVM arm64 `macos-aarch64` builder.
+
+use std::collections::HashSet;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_llvm.to_string_lossy().as_ref(),
+                "-B",
+                llvm_build_final.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    llvm_target_final.to_string_lossy().as_ref(),
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                "-DCMAKE_OSX_DEPLOYMENT_TARGET='11.0'",
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::GNU,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::macos_build_opts_ignore_dupicate_libs_warnings())
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            )),
+        "LLVM building cmake",
+    )?;
+
+    crate::utils::ninja(llvm_build_final.as_ref())?;
+
+    Ok(())
+}

crates/llvm-builder/src/platforms/mod.rs

@@ -0,0 +1,45 @@
+//! The revive LLVM builder platforms.
+
+pub mod aarch64_linux_gnu;
+pub mod aarch64_linux_musl;
+pub mod aarch64_macos;
+pub mod shared;
+pub mod wasm32_emscripten;
+pub mod x86_64_linux_gnu;
+pub mod x86_64_linux_musl;
+pub mod x86_64_macos;
+pub mod x86_64_windows_gnu;
+
+use std::str::FromStr;
+
+/// The list of platforms used as constants.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum Platform {
+    /// The native X86 platform.
+    X86,
+    /// The native AArch64 platform.
+    AArch64,
+    /// The PolkaVM RISC-V platform.
+    PolkaVM,
+}
+
+impl FromStr for Platform {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value {
+            "PolkaVM" => Ok(Self::PolkaVM),
+            value => Err(format!("Unsupported platform: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for Platform {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        match self {
+            Self::X86 => write!(f, "X86"),
+            Self::AArch64 => write!(f, "AArch64"),
+            Self::PolkaVM => write!(f, "RISCV"),
+        }
+    }
+}

crates/llvm-builder/src/platforms/shared.rs

@@ -0,0 +1,224 @@
+//! The shared options for building various platforms.
+
+use crate::ccache_variant::CcacheVariant;
+use crate::sanitizer::Sanitizer;
+use crate::target_env::TargetEnv;
+use crate::target_triple::TargetTriple;
+use std::path::Path;
+use std::process::Command;
+
+/// The build options shared by all platforms.
+pub const SHARED_BUILD_OPTS: [&str; 19] = [
+    "-DPACKAGE_VENDOR='Parity Technologies'",
+    "-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
+    "-DLLVM_BUILD_DOCS='Off'",
+    "-DLLVM_INCLUDE_DOCS='Off'",
+    "-DLLVM_INCLUDE_BENCHMARKS='Off'",
+    "-DLLVM_INCLUDE_EXAMPLES='Off'",
+    "-DLLVM_ENABLE_DOXYGEN='Off'",
+    "-DLLVM_ENABLE_SPHINX='Off'",
+    "-DLLVM_ENABLE_OCAMLDOC='Off'",
+    "-DLLVM_ENABLE_ZLIB='Off'",
+    "-DLLVM_ENABLE_ZSTD='Off'",
+    "-DLLVM_ENABLE_LIBXML2='Off'",
+    "-DLLVM_ENABLE_BINDINGS='Off'",
+    "-DLLVM_ENABLE_TERMINFO='Off'",
+    "-DLLVM_ENABLE_LIBEDIT='Off'",
+    "-DLLVM_ENABLE_LIBPFM='Off'",
+    "-DCMAKE_EXPORT_COMPILE_COMMANDS='On'",
+    "-DPython3_FIND_REGISTRY='LAST'", // Use Python version from $PATH, not from registry
+    "-DBUG_REPORT_URL='https://github.com/paritytech/contract-issues/issues/'",
+];
+
+/// The build options shared by all platforms except MUSL.
+pub const SHARED_BUILD_OPTS_NOT_MUSL: [&str; 4] = [
+    "-DLLVM_OPTIMIZED_TABLEGEN='On'",
+    "-DLLVM_BUILD_RUNTIME='Off'",
+    "-DLLVM_BUILD_RUNTIMES='Off'",
+    "-DLLVM_INCLUDE_RUNTIMES='Off'",
+];
+
+/// The shared build options to treat warnings as errors.
+///
+/// Disabled because it makes the build very brittle.
+pub fn shared_build_opts_werror(_target_env: TargetEnv) -> Vec<String> {
+    vec!["-DLLVM_ENABLE_WERROR='Off'".to_string()]
+}
+
+/// The build options to set the default target.
+pub fn shared_build_opts_default_target(target: Option<TargetTriple>) -> Vec<String> {
+    match target {
+        Some(target) => vec![format!(
+            "-DLLVM_DEFAULT_TARGET_TRIPLE='{}'",
+            target.to_string()
+        )],
+        None => vec![format!(
+            "-DLLVM_DEFAULT_TARGET_TRIPLE='{}'",
+            TargetTriple::PolkaVM
+        )],
+    }
+}
+
+/// The `musl` building sequence.
+pub fn build_musl(build_directory: &Path, target_directory: &Path) -> anyhow::Result<()> {
+    std::fs::create_dir_all(build_directory)?;
+    std::fs::create_dir_all(target_directory)?;
+
+    crate::utils::command(
+        Command::new("../configure")
+            .current_dir(build_directory)
+            .arg(format!("--prefix={}", target_directory.to_string_lossy()))
+            .arg(format!(
+                "--syslibdir={}/lib/",
+                target_directory.to_string_lossy()
+            ))
+            .arg("--enable-wrapper='clang'"),
+        "MUSL configuring",
+    )?;
+    crate::utils::command(
+        Command::new("make")
+            .current_dir(build_directory)
+            .arg("-j")
+            .arg(num_cpus::get().to_string()),
+        "MUSL building",
+    )?;
+    crate::utils::command(
+        Command::new("make")
+            .current_dir(build_directory)
+            .arg("install"),
+        "MUSL installing",
+    )?;
+
+    let mut include_directory = target_directory.to_path_buf();
+    include_directory.push("include/");
+
+    let mut asm_include_directory = include_directory.clone();
+    asm_include_directory.push("asm/");
+    std::fs::create_dir_all(asm_include_directory.as_path())?;
+
+    let mut types_header_path = asm_include_directory.clone();
+    types_header_path.push("types.h");
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy("/usr/include/linux", include_directory, &copy_options)?;
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        content_only: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy(
+        "/usr/include/asm-generic",
+        asm_include_directory,
+        &copy_options,
+    )?;
+
+    crate::utils::command(
+        Command::new("sed")
+            .arg("-i")
+            .arg("s/asm-generic/asm/")
+            .arg(types_header_path),
+        "types_header asm signature replacement",
+    )?;
+
+    Ok(())
+}
+
+/// The build options to enable assertions.
+pub fn shared_build_opts_assertions(enabled: bool) -> Vec<String> {
+    vec![format!(
+        "-DLLVM_ENABLE_ASSERTIONS='{}'",
+        if enabled { "On" } else { "Off" },
+    )]
+}
+
+/// The build options to build with RTTI support.
+pub fn shared_build_opts_rtti(enabled: bool) -> Vec<String> {
+    vec![format!(
+        "-DLLVM_ENABLE_RTTI='{}'",
+        if enabled { "On" } else { "Off" },
+    )]
+}
+
+/// The build options to enable sanitizers.
+pub fn shared_build_opts_sanitizers(sanitizer: Option<Sanitizer>) -> Vec<String> {
+    match sanitizer {
+        Some(sanitizer) => vec![format!("-DLLVM_USE_SANITIZER='{}'", sanitizer)],
+        None => vec![],
+    }
+}
+
+/// The build options to enable Valgrind for LLVM regression tests.
+pub fn shared_build_opts_valgrind(enabled: bool) -> Vec<String> {
+    if enabled {
+        vec!["-DLLVM_LIT_ARGS='-sv --vg --vg-leak'".to_owned()]
+    } else {
+        vec![]
+    }
+}
+
+/// The LLVM tests build options shared by all platforms.
+pub fn shared_build_opts_tests(enabled: bool) -> Vec<String> {
+    vec![
+        format!(
+            "-DLLVM_BUILD_UTILS='{}'",
+            if enabled { "On" } else { "Off" },
+        ),
+        format!(
+            "-DLLVM_BUILD_TESTS='{}'",
+            if enabled { "On" } else { "Off" },
+        ),
+        format!(
+            "-DLLVM_INCLUDE_UTILS='{}'",
+            if enabled { "On" } else { "Off" },
+        ),
+        format!(
+            "-DLLVM_INCLUDE_TESTS='{}'",
+            if enabled { "On" } else { "Off" },
+        ),
+    ]
+}
+
+/// The code coverage build options shared by all platforms.
+pub fn shared_build_opts_coverage(enabled: bool) -> Vec<String> {
+    vec![format!(
+        "-DLLVM_BUILD_INSTRUMENTED_COVERAGE='{}'",
+        if enabled { "On" } else { "Off" },
+    )]
+}
+
+/// Use of compiler cache (ccache) to speed up the build process.
+pub fn shared_build_opts_ccache(ccache_variant: Option<CcacheVariant>) -> Vec<String> {
+    match ccache_variant {
+        Some(ccache_variant) => vec![
+            format!(
+                "-DCMAKE_C_COMPILER_LAUNCHER='{}'",
+                ccache_variant.to_string()
+            ),
+            format!(
+                "-DCMAKE_CXX_COMPILER_LAUNCHER='{}'",
+                ccache_variant.to_string()
+            ),
+        ],
+        None => vec![],
+    }
+}
+
+/// Ignore duplicate libraries warnings for MacOS with XCode>=15.
+pub fn macos_build_opts_ignore_dupicate_libs_warnings() -> Vec<String> {
+    let xcode_version =
+        crate::utils::get_xcode_version().unwrap_or(crate::utils::XCODE_MIN_VERSION);
+    if xcode_version >= crate::utils::XCODE_VERSION_15 {
+        vec![
+            "-DCMAKE_EXE_LINKER_FLAGS='-Wl,-no_warn_duplicate_libraries'".to_owned(),
+            "-DCMAKE_SHARED_LINKER_FLAGS='-Wl,-no_warn_duplicate_libraries'".to_owned(),
+        ]
+    } else {
+        vec![]
+    }
+}

crates/llvm-builder/src/platforms/wasm32_emscripten.rs

@@ -0,0 +1,224 @@
+//! The revive LLVM `wasm32_unknown_emscripten` builder.
+//!
+//! Cross-compiling LLVM for Emscripten requires llvm-tblgen, clang-tblgen and llvm-config.
+
+use std::{collections::HashSet, path::Path, process::Command};
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: crate::BuildType,
+    targets: HashSet<crate::Platform>,
+    llvm_projects: HashSet<crate::llvm_project::LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<crate::TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<crate::ccache_variant::CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<crate::sanitizer::Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("ninja")?;
+    crate::utils::check_presence("emsdk")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    if cfg!(target_os = "linux") {
+        crate::utils::check_presence("lld")?;
+    }
+
+    let llvm_module_llvm = crate::LLVMPath::llvm_module_llvm()?;
+    let llvm_host_module_llvm = crate::LLVMPath::llvm_host_module_llvm()?;
+
+    let llvm_build_host = crate::LLVMPath::llvm_build_host()?;
+    let llvm_target_host = crate::LLVMPath::llvm_target_host()?;
+
+    let llvm_build_final = crate::LLVMPath::llvm_build_final()?;
+    let llvm_target_final = crate::LLVMPath::llvm_target_final()?;
+
+    build_host(
+        llvm_host_module_llvm.as_path(),
+        llvm_build_host.as_path(),
+        llvm_target_host.as_path(),
+        ccache_variant,
+    )?;
+
+    build_target(
+        build_type,
+        targets,
+        llvm_projects,
+        enable_rtti,
+        default_target,
+        llvm_module_llvm.as_path(),
+        llvm_build_final.as_path(),
+        llvm_target_final.as_path(),
+        llvm_target_host.as_path(),
+        enable_tests,
+        enable_coverage,
+        extra_args,
+        ccache_variant,
+        enable_assertions,
+        sanitizer,
+        enable_valgrind,
+    )?;
+
+    Ok(())
+}
+
+/// The host toolchain building sequence.
+fn build_host(
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    ccache_variant: Option<crate::ccache_variant::CcacheVariant>,
+) -> anyhow::Result<()> {
+    log::info!("building the LLVM Emscripten host utilities");
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                &format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                ),
+                "-DLLVM_BUILD_SHARED_LIBS='Off'",
+                "-DCMAKE_BUILD_TYPE='Release'",
+                &format!(
+                    "-DLLVM_TARGETS_TO_BUILD='WebAssembly;{}'",
+                    crate::Platform::PolkaVM
+                ),
+                "-DLLVM_ENABLE_PROJECTS='clang;lld'",
+            ])
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            )),
+        "LLVM host building cmake config",
+    )?;
+
+    crate::utils::ninja(build_directory)?;
+
+    Ok(())
+}
+
+/// The target toolchain building sequence.
+#[allow(clippy::too_many_arguments)]
+fn build_target(
+    build_type: crate::BuildType,
+    targets: HashSet<crate::Platform>,
+    llvm_projects: HashSet<crate::llvm_project::LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<crate::TargetTriple>,
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    host_target_directory: &Path,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<crate::ccache_variant::CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<crate::sanitizer::Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    let mut llvm_tblgen_path = host_target_directory.to_path_buf();
+    llvm_tblgen_path.push("bin/llvm-tblgen");
+
+    let mut clang_tblgen_path = host_target_directory.to_path_buf();
+    clang_tblgen_path.push("bin/clang-tblgen");
+
+    crate::utils::command(
+        Command::new("emcmake")
+            .env("EMCC_DEBUG", "2")
+            .env("CXXFLAGS", "-Dwait4=__syscall_wait4")
+            .env("LDFLAGS", "-lnodefs.js -s NO_INVOKE_RUN=1 -s EXIT_RUNTIME=1 -s ALLOW_MEMORY_GROWTH=1 -s EXPORTED_RUNTIME_METHODS=FS,callMain,NODEFS -s MODULARIZE=1 -s WASM_BIGINT=1 -s ALLOW_TABLE_GROWTH=1 -s NODEJS_CATCH_EXIT=0 -sDYNAMIC_EXECUTION=0")
+            .arg("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                "-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
+                // Enable thin LTO but emscripten has various issues with it.
+                // FIXME: https://github.com/paritytech/revive/issues/148
+                //"-DLLVM_ENABLE_LTO='Thin'",
+                //"-DCMAKE_EXE_LINKER_FLAGS='-Wl,-u,htons -Wl,-u,htonl -Wl,-u,fileno -Wl,-u,ntohs'",
+                &format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                ),
+                &format!("-DCMAKE_BUILD_TYPE='{build_type}'"),
+                &format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                ),
+                &format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                ),
+                "-DLLVM_BUILD_SHARED_LIBS='Off'",
+                "-DLLVM_ENABLE_DUMP='Off'",
+                "-DLLVM_ENABLE_EXPENSIVE_CHECKS='Off'",
+                "-DLLVM_ENABLE_BACKTRACES='Off'",
+                "-DLLVM_ENABLE_BACKTRACES='Off'",
+                "-DLLVM_ENABLE_THREADS='Off'",
+                "-DLLVM_BUILD_TOOLS='Off'",
+                &format!("-DLLVM_TABLEGEN='{}'", llvm_tblgen_path.to_string_lossy()),
+                &format!("-DCLANG_TABLEGEN='{}'", clang_tblgen_path.to_string_lossy()),
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(crate::target_env::TargetEnv::Emscripten))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_valgrind(
+                enable_valgrind,
+            )),
+        "LLVM target building cmake",
+    )?;
+
+    crate::utils::ninja(build_directory)?;
+
+    Ok(())
+}

crates/llvm-builder/src/platforms/x86_64_linux_gnu.rs

@@ -0,0 +1,111 @@
+//! The revive LLVM amd64 `linux-gnu` builder.
+
+use std::collections::HashSet;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    crate::utils::check_presence("lld")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_llvm.to_string_lossy().as_ref(),
+                "-B",
+                llvm_build_final.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    llvm_target_final.to_string_lossy().as_ref(),
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                "-DLLVM_USE_LINKER='lld'",
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::GNU,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_valgrind(
+                enable_valgrind,
+            )),
+        "LLVM building cmake",
+    )?;
+    crate::utils::ninja(llvm_build_final.as_ref())?;
+    Ok(())
+}

crates/llvm-builder/src/platforms/x86_64_linux_musl.rs

@@ -0,0 +1,394 @@
+//! The revive LLVM amd64 `linux-musl` builder.
+
+use std::collections::HashSet;
+use std::path::Path;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    log::info!("building for target x86_64_linux_musl");
+
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    crate::utils::check_presence("lld")?;
+    crate::utils::check_presence("ninja")?;
+
+    let musl_name = "musl-1.2.3";
+    let musl_build = LLVMPath::musl_build(musl_name)?;
+    let musl_target = LLVMPath::musl_target()?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_host_module_llvm = LLVMPath::llvm_host_module_llvm()?;
+
+    let llvm_build_crt = LLVMPath::llvm_build_crt()?;
+    let llvm_target_crt = LLVMPath::llvm_target_crt()?;
+
+    let llvm_build_host = LLVMPath::llvm_build_host()?;
+    let llvm_target_host = LLVMPath::llvm_target_host()?;
+
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    if !LLVMPath::musl_source(musl_name)?.exists() {
+        crate::utils::download_musl(musl_name)?;
+    }
+    crate::platforms::shared::build_musl(musl_build.as_path(), musl_target.as_path())?;
+    build_crt(
+        targets.clone(),
+        llvm_host_module_llvm.as_path(),
+        llvm_build_crt.as_path(),
+        llvm_target_crt.as_path(),
+        ccache_variant,
+    )?;
+    build_host(
+        llvm_host_module_llvm.as_path(),
+        llvm_build_host.as_path(),
+        llvm_target_host.as_path(),
+        musl_target.as_path(),
+        llvm_target_crt.as_path(),
+        ccache_variant,
+    )?;
+    build_target(
+        build_type,
+        targets,
+        llvm_projects,
+        enable_rtti,
+        default_target,
+        llvm_module_llvm.as_path(),
+        llvm_build_final.as_path(),
+        llvm_target_final.as_path(),
+        musl_target.as_path(),
+        llvm_target_host.as_path(),
+        enable_tests,
+        enable_coverage,
+        extra_args,
+        ccache_variant,
+        enable_assertions,
+        sanitizer,
+        enable_valgrind,
+    )?;
+
+    Ok(())
+}
+
+/// The `crt` building sequence.
+fn build_crt(
+    mut targets: HashSet<Platform>,
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    ccache_variant: Option<CcacheVariant>,
+) -> anyhow::Result<()> {
+    targets.insert(Platform::X86);
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_BUILD_TYPE='Release'",
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                "-DLLVM_ENABLE_PROJECTS='compiler-rt'",
+                format!("-DLLVM_TARGETS_TO_BUILD='{}'", Platform::X86).as_str(),
+                "-DLLVM_DEFAULT_TARGET_TRIPLE='x86_64-pc-linux-musl'",
+                "-DLLVM_BUILD_TESTS='Off'",
+                "-DLLVM_BUILD_RUNTIMES='Off'",
+                "-DLLVM_BUILD_UTILS='Off'",
+                "-DLLVM_INCLUDE_TESTS='Off'",
+                "-DLLVM_INCLUDE_RUNTIMES='Off'",
+                "-DLLVM_INCLUDE_UTILS='Off'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ARCH='x86_64'",
+                "-DCOMPILER_RT_BUILD_CRT='On'",
+                "-DCOMPILER_RT_BUILD_SANITIZERS='Off'",
+                "-DCOMPILER_RT_BUILD_XRAY='Off'",
+                "-DCOMPILER_RT_BUILD_LIBFUZZER='Off'",
+                "-DCOMPILER_RT_BUILD_PROFILE='Off'",
+                "-DCOMPILER_RT_BUILD_MEMPROF='Off'",
+                "-DCOMPILER_RT_BUILD_ORC='Off'",
+            ])
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            )),
+        "CRT building cmake",
+    )?;
+
+    crate::utils::command(
+        Command::new("ninja")
+            .arg("-C")
+            .arg(build_directory)
+            .arg("install-crt"),
+        "CRT building ninja",
+    )?;
+
+    Ok(())
+}
+
+/// The host toolchain building sequence.
+fn build_host(
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    musl_target_directory: &Path,
+    crt_target_directory: &Path,
+    ccache_variant: Option<CcacheVariant>,
+) -> anyhow::Result<()> {
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DDEFAULT_SYSROOT='{}'",
+                    musl_target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_BUILD_TYPE='Release'",
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                "-DCLANG_DEFAULT_CXX_STDLIB='libc++'",
+                "-DCLANG_DEFAULT_RTLIB='compiler-rt'",
+                "-DLLVM_DEFAULT_TARGET_TRIPLE='x86_64-pc-linux-musl'",
+                "-DLLVM_TARGETS_TO_BUILD='X86'",
+                "-DLLVM_BUILD_TESTS='Off'",
+                "-DLLVM_BUILD_UTILS='Off'",
+                "-DLLVM_INCLUDE_TESTS='Off'",
+                "-DLLVM_INCLUDE_UTILS='Off'",
+                "-DLLVM_ENABLE_PROJECTS='clang;lld'",
+                "-DLLVM_ENABLE_RUNTIMES='compiler-rt;libcxx;libcxxabi;libunwind'",
+                "-DLIBCXX_CXX_ABI='libcxxabi'",
+                "-DLIBCXX_HAS_MUSL_LIBC='On'",
+                "-DLIBCXX_ENABLE_SHARED='Off'",
+                "-DLIBCXX_ENABLE_STATIC='On'",
+                "-DLIBCXX_ENABLE_STATIC_ABI_LIBRARY='On'",
+                "-DLIBCXXABI_ENABLE_SHARED='Off'",
+                "-DLIBCXXABI_ENABLE_STATIC='On'",
+                "-DLIBCXXABI_ENABLE_STATIC_UNWINDER='On'",
+                "-DLIBCXXABI_USE_LLVM_UNWINDER='On'",
+                "-DLIBCXXABI_USE_COMPILER_RT='On'",
+                "-DLIBUNWIND_ENABLE_STATIC='On'",
+                "-DLIBUNWIND_ENABLE_SHARED='Off'",
+                "-DCOMPILER_RT_BUILD_CRT='On'",
+                "-DCOMPILER_RT_BUILD_SANITIZERS='Off'",
+                "-DCOMPILER_RT_BUILD_XRAY='Off'",
+                "-DCOMPILER_RT_BUILD_LIBFUZZER='Off'",
+                "-DCOMPILER_RT_BUILD_PROFILE='On'",
+                "-DCOMPILER_RT_BUILD_MEMPROF='Off'",
+                "-DCOMPILER_RT_BUILD_ORC='Off'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ARCH='x86_64'",
+                "-DCOMPILER_RT_DEFAULT_TARGET_ONLY='On'",
+                "-DLIBCLANG_BUILD_STATIC='On'",
+                "-DBUILD_SHARED_LIBS='Off'",
+            ])
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            )),
+        "LLVM host building cmake",
+    )?;
+
+    let mut crt_lib_directory = crt_target_directory.to_path_buf();
+    crt_lib_directory.push("lib/");
+
+    let mut build_lib_directory = build_directory.to_path_buf();
+    build_lib_directory.push("lib/");
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        content_only: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy(crt_lib_directory, build_lib_directory, &copy_options)?;
+
+    crate::utils::command(
+        Command::new("ninja")
+            .arg("-C")
+            .arg(build_directory)
+            .arg("install"),
+        "LLVM host building ninja",
+    )?;
+
+    Ok(())
+}
+
+/// The target toolchain building sequence.
+#[allow(clippy::too_many_arguments)]
+fn build_target(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    source_directory: &Path,
+    build_directory: &Path,
+    target_directory: &Path,
+    musl_target_directory: &Path,
+    host_target_directory: &Path,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+    enable_valgrind: bool,
+) -> anyhow::Result<()> {
+    let mut clang_path = host_target_directory.to_path_buf();
+    clang_path.push("bin/clang");
+
+    let mut clang_cxx_path = host_target_directory.to_path_buf();
+    clang_cxx_path.push("bin/clang++");
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                source_directory.to_string_lossy().as_ref(),
+                "-B",
+                build_directory.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                "-DBUILD_SHARED_LIBS='Off'",
+                "-DLIBCLANG_BUILD_STATIC='On'",
+                "-DLLVM_BUILD_STATIC='On'",
+                "-DLINKER_SUPPORTS_COLOR_DIAGNOSTICS=0",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    target_directory.to_string_lossy()
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                format!("-DCMAKE_C_COMPILER='{}'", clang_path.to_string_lossy()).as_str(),
+                format!(
+                    "-DCMAKE_CXX_COMPILER='{}'",
+                    clang_cxx_path.to_string_lossy()
+                )
+                .as_str(),
+                "-DCMAKE_FIND_LIBRARY_SUFFIXES='.a'",
+                "-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
+                "-DCMAKE_EXE_LINKER_FLAGS='-fuse-ld=lld -static'",
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::MUSL,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_valgrind(
+                enable_valgrind,
+            )),
+        "LLVM target building cmake",
+    )?;
+
+    crate::utils::ninja(build_directory)?;
+
+    let mut musl_lib_directory = musl_target_directory.to_path_buf();
+    musl_lib_directory.push("lib/");
+
+    let mut host_lib_directory = host_target_directory.to_path_buf();
+    host_lib_directory.push("lib/x86_64-pc-linux-musl/");
+
+    let mut target_lib_directory = target_directory.to_path_buf();
+    target_lib_directory.push("lib/");
+
+    let copy_options = fs_extra::dir::CopyOptions {
+        overwrite: true,
+        copy_inside: true,
+        content_only: true,
+        ..Default::default()
+    };
+    fs_extra::dir::copy(
+        musl_lib_directory,
+        target_lib_directory.as_path(),
+        &copy_options,
+    )?;
+    fs_extra::dir::copy(
+        host_lib_directory,
+        target_lib_directory.as_path(),
+        &copy_options,
+    )?;
+
+    Ok(())
+}

crates/llvm-builder/src/platforms/x86_64_macos.rs

@@ -0,0 +1,105 @@
+//! The revive LLVM amd64 `macos` builder.
+
+use std::collections::HashSet;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_llvm = LLVMPath::llvm_module_llvm()?;
+    let llvm_build_final = LLVMPath::llvm_build_final()?;
+    let llvm_target_final = LLVMPath::llvm_target_final()?;
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_llvm.to_string_lossy().as_ref(),
+                "-B",
+                llvm_build_final.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    llvm_target_final.to_string_lossy().as_ref(),
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                "-DCMAKE_OSX_DEPLOYMENT_TARGET='11.0'",
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::GNU,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::macos_build_opts_ignore_dupicate_libs_warnings())
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            )),
+        "LLVM building cmake",
+    )?;
+
+    crate::utils::ninja(llvm_build_final.as_ref())?;
+
+    Ok(())
+}

crates/llvm-builder/src/platforms/x86_64_windows_gnu.rs

@@ -0,0 +1,127 @@
+//! The revive LLVM amd64 `windows-gnu` builder.
+
+use std::collections::HashSet;
+use std::path::PathBuf;
+use std::process::Command;
+
+use crate::build_type::BuildType;
+use crate::ccache_variant::CcacheVariant;
+use crate::llvm_path::LLVMPath;
+use crate::llvm_project::LLVMProject;
+use crate::platforms::Platform;
+use crate::sanitizer::Sanitizer;
+use crate::target_triple::TargetTriple;
+
+/// The building sequence.
+#[allow(clippy::too_many_arguments)]
+pub fn build(
+    build_type: BuildType,
+    targets: HashSet<Platform>,
+    llvm_projects: HashSet<LLVMProject>,
+    enable_rtti: bool,
+    default_target: Option<TargetTriple>,
+    enable_tests: bool,
+    enable_coverage: bool,
+    extra_args: &[String],
+    ccache_variant: Option<CcacheVariant>,
+    enable_assertions: bool,
+    sanitizer: Option<Sanitizer>,
+) -> anyhow::Result<()> {
+    crate::utils::check_presence("cmake")?;
+    crate::utils::check_presence("clang")?;
+    crate::utils::check_presence("clang++")?;
+    crate::utils::check_presence("lld")?;
+    crate::utils::check_presence("ninja")?;
+
+    let llvm_module_llvm =
+        LLVMPath::llvm_module_llvm().and_then(crate::utils::path_windows_to_unix)?;
+    let llvm_build_final =
+        LLVMPath::llvm_build_final().and_then(crate::utils::path_windows_to_unix)?;
+    let llvm_target_final =
+        LLVMPath::llvm_target_final().and_then(crate::utils::path_windows_to_unix)?;
+
+    crate::utils::command(
+        Command::new("cmake")
+            .args([
+                "-S",
+                llvm_module_llvm.to_string_lossy().as_ref(),
+                "-B",
+                llvm_build_final.to_string_lossy().as_ref(),
+                "-G",
+                "Ninja",
+                format!(
+                    "-DCMAKE_INSTALL_PREFIX='{}'",
+                    llvm_target_final.to_string_lossy().as_ref(),
+                )
+                .as_str(),
+                format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
+                "-DCMAKE_C_COMPILER='clang'",
+                "-DCMAKE_CXX_COMPILER='clang++'",
+                format!(
+                    "-DLLVM_TARGETS_TO_BUILD='{}'",
+                    targets
+                        .into_iter()
+                        .map(|platform| platform.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                format!(
+                    "-DLLVM_ENABLE_PROJECTS='{}'",
+                    llvm_projects
+                        .into_iter()
+                        .map(|project| project.to_string())
+                        .collect::<Vec<String>>()
+                        .join(";")
+                )
+                .as_str(),
+                "-DLLVM_USE_LINKER='lld'",
+            ])
+            .args(crate::platforms::shared::shared_build_opts_default_target(
+                default_target,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_tests(
+                enable_tests,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_coverage(
+                enable_coverage,
+            ))
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS)
+            .args(crate::platforms::shared::SHARED_BUILD_OPTS_NOT_MUSL)
+            .args(crate::platforms::shared::shared_build_opts_werror(
+                crate::target_env::TargetEnv::GNU,
+            ))
+            .args(extra_args)
+            .args(crate::platforms::shared::shared_build_opts_ccache(
+                ccache_variant,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_assertions(
+                enable_assertions,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_rtti(
+                enable_rtti,
+            ))
+            .args(crate::platforms::shared::shared_build_opts_sanitizers(
+                sanitizer,
+            )),
+        "LLVM building cmake",
+    )?;
+
+    crate::utils::ninja(llvm_build_final.as_ref())?;
+
+    let libstdcpp_source_path = match std::env::var("LIBSTDCPP_SOURCE_PATH") {
+        Ok(libstdcpp_source_path) => PathBuf::from(libstdcpp_source_path),
+        Err(error) => anyhow::bail!(
+            "The `LIBSTDCPP_SOURCE_PATH` must be set to the path to the libstdc++.a static library: {}", error
+        ),
+    };
+    let mut libstdcpp_destination_path = llvm_target_final;
+    libstdcpp_destination_path.push("./lib/libstdc++.a");
+    fs_extra::file::copy(
+        crate::utils::path_windows_to_unix(libstdcpp_source_path)?,
+        crate::utils::path_windows_to_unix(libstdcpp_destination_path)?,
+        &fs_extra::file::CopyOptions::default(),
+    )?;
+
+    Ok(())
+}

crates/llvm-builder/src/revive_llvm/arguments.rs

@@ -0,0 +1,113 @@
+//! The revive LLVM builder arguments.
+
+use clap::Parser;
+use revive_llvm_builder::ccache_variant::CcacheVariant;
+
+/// The revive LLVM builder arguments.
+#[derive(Debug, Parser)]
+#[command(version, about)]
+pub struct Arguments {
+    /// Target environment to build LLVM (gnu, musl, emscripten).
+    #[arg(long, default_value_t = revive_llvm_builder::target_env::TargetEnv::GNU)]
+    pub target_env: revive_llvm_builder::target_env::TargetEnv,
+
+    #[command(subcommand)]
+    pub subcommand: Subcommand,
+}
+
+/// The revive LLVM builder arguments.
+#[derive(Debug, clap::Subcommand)]
+pub enum Subcommand {
+    /// Clone the branch specified in `LLVM.lock`.
+    Clone {
+        /// Clone with full commits history.
+        #[arg(long)]
+        deep: bool,
+    },
+
+    /// Build the LLVM framework.
+    Build {
+        /// LLVM build type (`Debug`, `Release`, `RelWithDebInfo`, or `MinSizeRel`).
+        #[arg(long, default_value_t = revive_llvm_builder::BuildType::Release)]
+        build_type: revive_llvm_builder::BuildType,
+
+        /// Additional targets to build LLVM with.
+        #[arg(long)]
+        targets: Vec<String>,
+
+        /// LLVM projects to build LLVM with.
+        #[arg(long)]
+        llvm_projects: Vec<revive_llvm_builder::llvm_project::LLVMProject>,
+
+        /// Whether to build LLVM with run-time type information (RTTI) enabled.
+        #[arg(long)]
+        enable_rtti: bool,
+
+        /// The default target to build LLVM with.
+        #[arg(long)]
+        default_target: Option<revive_llvm_builder::target_triple::TargetTriple>,
+
+        /// Whether to build the LLVM tests.
+        #[arg(long)]
+        enable_tests: bool,
+
+        /// Whether to build LLVM for source-based code coverage.
+        #[arg(long)]
+        enable_coverage: bool,
+
+        /// Extra arguments to pass to CMake.  
+        /// A leading backslash will be unescaped.
+        #[arg(long)]
+        extra_args: Vec<String>,
+
+        /// Whether to use compiler cache (ccache) to speed-up builds.
+        #[arg(long)]
+        ccache_variant: Option<CcacheVariant>,
+
+        /// Whether to build with assertions enabled or not.
+        #[arg(long, default_value_t = true)]
+        enable_assertions: bool,
+
+        /// Build LLVM with sanitizer enabled (`Address`, `Memory`, `MemoryWithOrigins`, `Undefined`, `Thread`, `DataFlow`, or `Address;Undefined`).
+        #[arg(long)]
+        sanitizer: Option<revive_llvm_builder::sanitizer::Sanitizer>,
+
+        /// Whether to run LLVM unit tests under valgrind or not.
+        #[arg(long)]
+        enable_valgrind: bool,
+    },
+
+    /// Checkout the branch specified in `LLVM.lock`.
+    Checkout {
+        /// Remove all artifacts preventing the checkout (removes all local changes!).
+        #[arg(long)]
+        force: bool,
+    },
+
+    /// Clean the build artifacts.
+    Clean,
+
+    /// Build the LLVM compiler-rt builtins for the PolkaVM target.
+    Builtins {
+        /// LLVM build type (`Debug`, `Release`, `RelWithDebInfo`, or `MinSizeRel`).
+        #[arg(long, default_value_t = revive_llvm_builder::BuildType::Release)]
+        build_type: revive_llvm_builder::BuildType,
+
+        /// The default target to build LLVM with.
+        #[arg(long)]
+        default_target: Option<revive_llvm_builder::target_triple::TargetTriple>,
+
+        /// Extra arguments to pass to CMake.  
+        /// A leading backslash will be unescaped.
+        #[arg(long)]
+        extra_args: Vec<String>,
+
+        /// Whether to use compiler cache (ccache) to speed-up builds.
+        #[arg(long)]
+        ccache_variant: Option<CcacheVariant>,
+
+        /// Build LLVM with sanitizer enabled (`Address`, `Memory`, `MemoryWithOrigins`, `Undefined`, `Thread`, `DataFlow`, or `Address;Undefined`).
+        #[arg(long)]
+        sanitizer: Option<revive_llvm_builder::sanitizer::Sanitizer>,
+    },
+}

crates/llvm-builder/src/revive_llvm/main.rs

@@ -0,0 +1,141 @@
+//! The revive LLVM builder.
+
+pub(crate) mod arguments;
+
+use std::collections::HashSet;
+use std::path::PathBuf;
+use std::str::FromStr;
+
+use anyhow::Context;
+use clap::Parser;
+
+use self::arguments::{Arguments, Subcommand};
+
+fn main() {
+    env_logger::init();
+
+    match main_inner() {
+        Ok(()) => std::process::exit(0),
+        Err(error) => {
+            log::error!("{error:?}");
+            std::process::exit(1)
+        }
+    }
+}
+
+fn main_inner() -> anyhow::Result<()> {
+    let arguments = Arguments::parse();
+
+    revive_llvm_builder::utils::directory_target_llvm(arguments.target_env);
+
+    match arguments.subcommand {
+        Subcommand::Clone { deep } => {
+            let lock = revive_llvm_builder::Lock::try_from(&PathBuf::from(
+                revive_llvm_builder::lock::LLVM_LOCK_DEFAULT_PATH,
+            ))?;
+            revive_llvm_builder::clone(lock, deep, arguments.target_env)?;
+        }
+
+        Subcommand::Build {
+            build_type,
+            targets,
+            llvm_projects,
+            enable_rtti,
+            default_target,
+            enable_tests,
+            enable_coverage,
+            extra_args,
+            ccache_variant,
+            enable_assertions,
+            sanitizer,
+            enable_valgrind,
+        } => {
+            let mut targets = targets
+                .into_iter()
+                .map(|target| revive_llvm_builder::Platform::from_str(target.as_str()))
+                .collect::<Result<HashSet<revive_llvm_builder::Platform>, String>>()
+                .map_err(|platform| anyhow::anyhow!("Unknown platform `{}`", platform))?;
+            targets.insert(revive_llvm_builder::Platform::PolkaVM);
+
+            log::info!("build targets: {:?}", &targets);
+
+            let extra_args_unescaped: Vec<String> = extra_args
+                .iter()
+                .map(|argument| {
+                    argument
+                        .strip_prefix('\\')
+                        .unwrap_or(argument.as_str())
+                        .to_owned()
+                })
+                .collect();
+
+            log::debug!("extra_args: {:#?}", extra_args);
+            log::debug!("extra_args_unescaped: {:#?}", extra_args_unescaped);
+
+            if let Some(ccache_variant) = ccache_variant {
+                revive_llvm_builder::utils::check_presence(ccache_variant.to_string().as_str())?;
+            }
+
+            let mut projects = llvm_projects
+                .into_iter()
+                .map(|project| {
+                    revive_llvm_builder::llvm_project::LLVMProject::from_str(
+                        project.to_string().as_str(),
+                    )
+                })
+                .collect::<Result<HashSet<revive_llvm_builder::llvm_project::LLVMProject>, String>>(
+                )
+                .map_err(|project| anyhow::anyhow!("Unknown LLVM project `{}`", project))?;
+            projects.insert(revive_llvm_builder::llvm_project::LLVMProject::LLD);
+
+            log::info!("build projects: {:?}", &projects);
+
+            revive_llvm_builder::build(
+                build_type,
+                arguments.target_env,
+                targets,
+                projects,
+                enable_rtti,
+                default_target,
+                enable_tests,
+                enable_coverage,
+                &extra_args_unescaped,
+                ccache_variant,
+                enable_assertions,
+                sanitizer,
+                enable_valgrind,
+            )?;
+        }
+
+        Subcommand::Checkout { force } => {
+            let lock = revive_llvm_builder::Lock::try_from(&PathBuf::from(
+                revive_llvm_builder::lock::LLVM_LOCK_DEFAULT_PATH,
+            ))?;
+            revive_llvm_builder::checkout(lock, force)?;
+        }
+
+        Subcommand::Clean => {
+            revive_llvm_builder::clean()
+                .with_context(|| "Unable to remove target LLVM directory")?;
+        }
+
+        Subcommand::Builtins {
+            build_type,
+            default_target,
+            extra_args,
+            ccache_variant,
+            sanitizer,
+        } => {
+            revive_llvm_builder::builtins::build(
+                build_type,
+                arguments.target_env,
+                default_target,
+                &extra_args,
+                ccache_variant,
+                sanitizer,
+            )?;
+        }
+    }
+
+    Ok(())
+}

crates/llvm-builder/src/sanitizer.rs

@@ -0,0 +1,50 @@
+//! LLVM sanitizers.
+
+/// LLVM sanitizers.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum Sanitizer {
+    /// The address sanitizer.
+    Address,
+    /// The memory sanitizer.
+    Memory,
+    /// The memory with origins sanitizer.
+    MemoryWithOrigins,
+    /// The undefined behavior sanitizer
+    Undefined,
+    /// The thread sanitizer.
+    Thread,
+    /// The data flow sanitizer.
+    DataFlow,
+    /// Combine address and undefined behavior sanitizer.
+    AddressUndefined,
+}
+
+impl std::str::FromStr for Sanitizer {
+    type Err = String;
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value.to_lowercase().as_str() {
+            "address" => Ok(Self::Address),
+            "memory" => Ok(Self::Memory),
+            "memorywithorigins" => Ok(Self::MemoryWithOrigins),
+            "undefined" => Ok(Self::Undefined),
+            "thread" => Ok(Self::Thread),
+            "dataflow" => Ok(Self::DataFlow),
+            "address;undefined" => Ok(Self::AddressUndefined),
+            value => Err(format!("Unsupported sanitizer: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for Sanitizer {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        match self {
+            Self::Address => write!(f, "Address"),
+            Self::Memory => write!(f, "Memory"),
+            Self::MemoryWithOrigins => write!(f, "MemoryWithOrigins"),
+            Self::Undefined => write!(f, "Undefined"),
+            Self::Thread => write!(f, "Thread"),
+            Self::DataFlow => write!(f, "DataFlow"),
+            Self::AddressUndefined => write!(f, "Address;Undefined"),
+        }
+    }
+}

crates/llvm-builder/src/target_env.rs

@@ -0,0 +1,36 @@
+//! The target environments to build LLVM.
+
+/// The list of target environments used as constants.
+#[derive(Default, Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum TargetEnv {
+    /// The GNU target environment.
+    #[default]
+    GNU,
+    /// The MUSL target environment.
+    MUSL,
+    /// The wasm32 Emscripten environment.
+    Emscripten,
+}
+
+impl std::str::FromStr for TargetEnv {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value {
+            "gnu" => Ok(Self::GNU),
+            "musl" => Ok(Self::MUSL),
+            "emscripten" => Ok(Self::Emscripten),
+            value => Err(format!("Unsupported target environment: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for TargetEnv {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Self::GNU => write!(f, "gnu"),
+            Self::MUSL => write!(f, "musl"),
+            Self::Emscripten => write!(f, "emscripten"),
+        }
+    }
+}

crates/llvm-builder/src/target_triple.rs

@@ -0,0 +1,29 @@
+//! The PolkaVM LLVM target triples.
+
+/// The list of target triples used as constants.
+///
+/// It must be in the lowercase.
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub enum TargetTriple {
+    /// The PolkaVM RISC-V target triple.
+    PolkaVM,
+}
+
+impl std::str::FromStr for TargetTriple {
+    type Err = String;
+
+    fn from_str(value: &str) -> Result<Self, Self::Err> {
+        match value {
+            "polkavm" => Ok(Self::PolkaVM),
+            value => Err(format!("Unsupported target triple: `{}`", value)),
+        }
+    }
+}
+
+impl std::fmt::Display for TargetTriple {
+    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        match self {
+            Self::PolkaVM => write!(f, "riscv64-unknown-elf"),
+        }
+    }
+}

crates/llvm-builder/src/utils.rs

@@ -0,0 +1,230 @@
+//! The LLVM builder utilities.
+
+use std::fs::File;
+use std::path::Path;
+use std::path::PathBuf;
+use std::process::Command;
+use std::process::Stdio;
+use std::time::Duration;
+
+use path_slash::PathBufExt;
+
+/// The LLVM host repository URL.
+pub const LLVM_HOST_SOURCE_URL: &str = "https://github.com/llvm/llvm-project";
+
+/// The LLVM host repository tag.
+pub const LLVM_HOST_SOURCE_TAG: &str = "llvmorg-18.1.8";
+
+/// The minimum required XCode version.
+pub const XCODE_MIN_VERSION: u32 = 11;
+
+/// The XCode version 15.
+pub const XCODE_VERSION_15: u32 = 15;
+
+/// The number of download retries if failed.
+pub const DOWNLOAD_RETRIES: u16 = 16;
+
+/// The number of parallel download requests.
+pub const DOWNLOAD_PARALLEL_REQUESTS: u16 = 1;
+
+/// The download timeout in seconds.
+pub const DOWNLOAD_TIMEOUT_SECONDS: u64 = 300;
+
+/// The musl snapshots URL.
+pub const MUSL_SNAPSHOTS_URL: &str = "https://git.musl-libc.org/cgit/musl/snapshot";
+
+/// The emscripten SDK git URL.
+pub const EMSDK_SOURCE_URL: &str = "https://github.com/emscripten-core/emsdk.git";
+
+/// The emscripten SDK version.
+pub const EMSDK_VERSION: &str = "3.1.64";
+
+/// The subprocess runner.
+///
+/// Checks the status and prints `stderr`.
+pub fn command(command: &mut Command, description: &str) -> anyhow::Result<()> {
+    log::debug!("executing '{command:?}' ({description})");
+
+    if std::env::var("DRY_RUN").is_ok() {
+        log::warn!("Only a dry run; not executing the command.");
+        return Ok(());
+    }
+
+    let status = command
+        .status()
+        .map_err(|error| anyhow::anyhow!("{} process: {}", description, error))?;
+
+    if !status.success() {
+        log::error!("the command '{command:?}' failed!");
+        anyhow::bail!("{} failed", description);
+    }
+
+    Ok(())
+}
+
+/// Download a file from the URL to the path.
+pub fn download(url: &str, path: &str) -> anyhow::Result<()> {
+    log::trace!("downloading '{url}' into '{path}'");
+
+    let mut downloader = downloader::Downloader::builder()
+        .download_folder(Path::new(path))
+        .parallel_requests(DOWNLOAD_PARALLEL_REQUESTS)
+        .retries(DOWNLOAD_RETRIES)
+        .timeout(Duration::from_secs(DOWNLOAD_TIMEOUT_SECONDS))
+        .build()?;
+    while let Err(error) = downloader.download(&[downloader::Download::new(url)]) {
+        log::error!("MUSL download from `{url}` failed: {error}");
+    }
+    Ok(())
+}
+
+/// Unpack a tarball.
+pub fn unpack_tar(filename: PathBuf, path: &str) -> anyhow::Result<()> {
+    let tar_gz = File::open(filename)?;
+    let tar = flate2::read::GzDecoder::new(tar_gz);
+    let mut archive = tar::Archive::new(tar);
+    archive.unpack(path)?;
+    Ok(())
+}
+
+/// The `musl` downloading sequence.
+pub fn download_musl(name: &str) -> anyhow::Result<()> {
+    log::info!("downloading musl {name}");
+    let tar_file_name = format!("{name}.tar.gz");
+    let url = format!("{}/{tar_file_name}", MUSL_SNAPSHOTS_URL);
+    let target_path = crate::llvm_path::DIRECTORY_LLVM_TARGET
+        .get()
+        .unwrap()
+        .to_string_lossy();
+    download(url.as_str(), &target_path)?;
+    let musl_tarball = crate::LLVMPath::musl_source(tar_file_name.as_str())?;
+    unpack_tar(musl_tarball, &target_path)?;
+    Ok(())
+}
+
+/// Call ninja to build the LLVM.
+pub fn ninja(build_dir: &Path) -> anyhow::Result<()> {
+    let mut ninja = Command::new("ninja");
+    ninja.args(["-C", build_dir.to_string_lossy().as_ref()]);
+    if std::env::var("DRY_RUN").is_ok() {
+        ninja.arg("-n");
+    }
+    command(ninja.arg("install"), "Running ninja install")?;
+    Ok(())
+}
+
+/// Create an absolute path, appending it to the current working directory.
+pub fn absolute_path<P: AsRef<Path>>(path: P) -> anyhow::Result<PathBuf> {
+    let mut full_path = std::env::current_dir()?;
+    full_path.push(path);
+    Ok(full_path)
+}
+
+///
+/// Converts a Windows path into a Unix path.
+///
+pub fn path_windows_to_unix<P: AsRef<Path> + PathBufExt>(path: P) -> anyhow::Result<PathBuf> {
+    path.to_slash()
+        .map(|pathbuf| PathBuf::from(pathbuf.to_string()))
+        .ok_or_else(|| anyhow::anyhow!("Windows-to-Unix path conversion error"))
+}
+
+/// Checks if the tool exists in the system.
+pub fn check_presence(name: &str) -> anyhow::Result<()> {
+    let description = &format!("checking the `{name}` executable");
+    log::info!("{description}");
+
+    command(Command::new("which").arg(name), description)
+        .map_err(|_| anyhow::anyhow!("Tool `{}` is missing. Please install", name))
+}
+
+/// Identify XCode version using `pkgutil`.
+pub fn get_xcode_version() -> anyhow::Result<u32> {
+    let pkgutil = Command::new("pkgutil")
+        .args(["--pkg-info", "com.apple.pkg.CLTools_Executables"])
+        .stdout(Stdio::piped())
+        .spawn()
+        .map_err(|error| anyhow::anyhow!("`pkgutil` process: {}", error))?;
+    let grep_version = Command::new("grep")
+        .arg("version")
+        .stdin(Stdio::from(pkgutil.stdout.expect(
+            "Failed to identify XCode version - XCode or CLI tools are not installed",
+        )))
+        .output()
+        .map_err(|error| anyhow::anyhow!("`grep` process: {}", error))?;
+    let version_string = String::from_utf8(grep_version.stdout)?;
+    let version_regex = regex::Regex::new(r"version: (\d+)\..*")?;
+    let captures = version_regex
+        .captures(version_string.as_str())
+        .ok_or(anyhow::anyhow!(
+            "Failed to parse XCode version: {version_string}"
+        ))?;
+    let xcode_version: u32 = captures
+        .get(1)
+        .expect("Always has a major version")
+        .as_str()
+        .parse()
+        .map_err(|error| anyhow::anyhow!("Failed to parse XCode version: {error}"))?;
+    Ok(xcode_version)
+}
+
+/// Install the Emscripten SDK.
+pub fn install_emsdk() -> anyhow::Result<()> {
+    log::info!("installing emsdk v{EMSDK_VERSION}");
+
+    let emsdk_source_path = PathBuf::from(crate::LLVMPath::DIRECTORY_EMSDK_SOURCE);
+
+    if emsdk_source_path.exists() {
+        log::warn!(
+            "emsdk source path {emsdk_source_path:?} already exists.
+            Skipping the emsdk installation, delete the source path for re-installation"
+        );
+        return Ok(());
+    }
+
+    crate::utils::command(
+        Command::new("git")
+            .arg("clone")
+            .arg(crate::utils::EMSDK_SOURCE_URL)
+            .arg(emsdk_source_path.to_string_lossy().as_ref()),
+        "Emscripten SDK repository cloning",
+    )?;
+
+    crate::utils::command(
+        Command::new("git")
+            .arg("checkout")
+            .arg(format!("tags/{}", crate::utils::EMSDK_VERSION))
+            .current_dir(&emsdk_source_path),
+        "Emscripten SDK repository version checkout",
+    )?;
+
+    crate::utils::command(
+        Command::new("./emsdk")
+            .arg("install")
+            .arg(EMSDK_VERSION)
+            .current_dir(&emsdk_source_path),
+        "Emscripten SDK installation",
+    )?;
+
+    crate::utils::command(
+        Command::new("./emsdk")
+            .arg("activate")
+            .arg(EMSDK_VERSION)
+            .current_dir(&emsdk_source_path),
+        "Emscripten SDK activation",
+    )?;
+
+    log::warn!(
+        "run 'source {}emsdk_env.sh' to finish the emsdk installation",
+        emsdk_source_path.display()
+    );
+
+    Ok(())
+}
+
+/// The LLVM target directory default path.
+pub fn directory_target_llvm(target_env: crate::target_env::TargetEnv) -> PathBuf {
+    crate::llvm_path::DIRECTORY_LLVM_TARGET
+        .get_or_init(|| PathBuf::from(format!("./target-llvm/{}/", target_env)))
+        .clone()
+}

crates/llvm-builder/tests/build.rs

@@ -0,0 +1,158 @@
+pub mod common;
+
+use std::process::Command;
+
+use assert_cmd::prelude::*;
+
+/// This test verifies that the LLVM repository can be successfully cloned, built, and cleaned.
+#[test]
+fn clone_build_and_clean() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("build")
+        .arg("--llvm-projects")
+        .arg("clang")
+        .arg("--llvm-projects")
+        .arg("lld")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("builtins")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clean")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// This test verifies that the LLVM repository can be successfully cloned, built, and cleaned
+/// with 2-staged build using MUSL as sysroot.
+#[test]
+#[cfg(target_os = "linux")]
+fn clone_build_and_clean_musl() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .arg("clone")
+        .current_dir(test_dir.path())
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .arg("--target-env")
+        .arg("musl")
+        .arg("build")
+        .arg("--llvm-projects")
+        .arg("clang")
+        .arg("--llvm-projects")
+        .arg("lld")
+        .current_dir(test_dir.path())
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("builtins")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clean")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// This test verifies that the LLVM repository can be successfully cloned and built in debug mode
+/// with tests and coverage enabled.
+#[test]
+fn debug_build_with_tests_coverage() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("build")
+        .arg("--enable-coverage")
+        .arg("--enable-tests")
+        .arg("--build-type")
+        .arg("Debug")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// This test verifies that the LLVM repository can be successfully built with address sanitizer.
+#[test]
+fn build_with_sanitizers() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("build")
+        .arg("--sanitizer")
+        .arg("Address")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// Tests the clone, build, and clean process of the LLVM repository for the emscripten target.
+#[test]
+#[cfg(any(target_os = "linux", target_os = "macos"))]
+fn clone_build_and_clean_emscripten() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+    let command = Command::cargo_bin(common::REVIVE_LLVM)?;
+    let program = command.get_program().to_string_lossy();
+    let emsdk_wrapped_build_command = format!(
+        "{program} --target-env emscripten clone && \
+        source {}emsdk_env.sh && \
+        {program} --target-env emscripten build --llvm-projects clang --llvm-projects lld",
+        revive_llvm_builder::LLVMPath::DIRECTORY_EMSDK_SOURCE,
+    );
+
+    Command::new("sh")
+        .arg("-c")
+        .arg(emsdk_wrapped_build_command)
+        .current_dir(test_dir.path())
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .arg("clean")
+        .current_dir(test_dir.path())
+        .assert()
+        .success();
+
+    Ok(())
+}

crates/llvm-builder/tests/checkout.rs

@@ -0,0 +1,48 @@
+pub mod common;
+
+use std::process::Command;
+
+use assert_cmd::prelude::*;
+
+/// This test verifies that after cloning the LLVM repository, checking out a specific branch
+/// or reference works as expected.
+#[test]
+fn checkout_after_clone() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("checkout")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// This test verifies that after cloning the LLVM repository, checking out a specific branch
+/// or reference with the `--force` option works as expected.
+#[test]
+fn force_checkout() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("checkout")
+        .arg("--force")
+        .assert()
+        .success();
+
+    Ok(())
+}

crates/llvm-builder/tests/clone.rs

@@ -0,0 +1,36 @@
+pub mod common;
+
+use std::process::Command;
+
+use assert_cmd::prelude::*;
+
+/// This test verifies that the LLVM repository can be successfully cloned using a specific branch
+/// and reference.
+#[test]
+fn clone() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .assert()
+        .success();
+
+    Ok(())
+}
+
+/// This test verifies that the LLVM repository can be successfully cloned using a specific branch
+/// and reference with --deep option.
+#[test]
+fn clone_deep() -> anyhow::Result<()> {
+    let test_dir = common::TestDir::with_lockfile(None)?;
+
+    Command::cargo_bin(common::REVIVE_LLVM)?
+        .current_dir(test_dir.path())
+        .arg("clone")
+        .arg("--deep")
+        .assert()
+        .success();
+
+    Ok(())
+}

crates/llvm-builder/tests/common.rs

@@ -0,0 +1,36 @@
+use assert_fs::fixture::FileWriteStr;
+
+pub const REVIVE_LLVM: &str = "revive-llvm";
+pub const REVIVE_LLVM_REPO_URL: &str = "https://github.com/llvm/llvm-project";
+pub const REVIVE_LLVM_REPO_TEST_BRANCH: &str = "release/18.x";
+
+pub struct TestDir {
+    _lockfile: assert_fs::NamedTempFile,
+    path: std::path::PathBuf,
+}
+
+/// Creates a temporary lock file for testing.
+impl TestDir {
+    pub fn with_lockfile(reference: Option<String>) -> anyhow::Result<Self> {
+        let file =
+            assert_fs::NamedTempFile::new(revive_llvm_builder::lock::LLVM_LOCK_DEFAULT_PATH)?;
+        let lock = revive_llvm_builder::Lock {
+            url: REVIVE_LLVM_REPO_URL.to_string(),
+            branch: REVIVE_LLVM_REPO_TEST_BRANCH.to_string(),
+            r#ref: reference,
+        };
+        file.write_str(toml::to_string(&lock)?.as_str())?;
+
+        Ok(Self {
+            path: file
+                .parent()
+                .expect("lockfile parent dir always exists")
+                .into(),
+            _lockfile: file,
+        })
+    }
+
+    pub fn path(&self) -> &std::path::Path {
+        &self.path
+    }
+}

crates/llvm-context/Cargo.toml

@@ -13,22 +13,14 @@ description = "Shared front end code of the revive PolkaVM compilers"
 [lib]
 doctest = false
 
-[features]
-riscv-zbb = []
-riscv-64 = []
-
 [dependencies]
 anyhow = { workspace = true }
 semver = { workspace = true }
 itertools = { workspace = true }
 serde = { workspace = true, features = ["derive"] }
-regex = { workspace = true }
-once_cell = { workspace = true }
 num = { workspace = true }
 hex = { workspace = true }
-sha2 = { workspace = true }
 sha3 = { workspace = true }
-md5 = { workspace = true }
 inkwell = { workspace = true }
 polkavm-disassembler = { workspace = true }
 polkavm-common = { workspace = true }
@@ -36,5 +28,4 @@ polkavm-common = { workspace = true }
 revive-common = { workspace = true }
 revive-runtime-api = { workspace = true }
 revive-linker = { workspace = true }
-revive-builtins = { workspace = true }
 revive-stdlib = { workspace = true }

crates/llvm-context/src/debug_config/ir_type.rs

@@ -1,20 +1,17 @@
 //! The debug IR type.
 
 /// The debug IR type.
-#[allow(non_camel_case_types)]
 #[allow(clippy::upper_case_acronyms)]
 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
 pub enum IRType {
     /// Whether to dump the Yul code.
     Yul,
-    /// Whether to dump the EVM legacy assembly code.
-    EVMLA,
-    /// Whether to dump the Ethereal IR code.
-    EthIR,
     /// Whether to dump the LLVM IR code.
     LLVM,
     /// Whether to dump the assembly code.
     Assembly,
+    /// Whether to dump the ELF shared object
+    SO,
     /// Whether to jump JSON
     #[cfg(debug_assertions)]
     JSON,
@@ -25,12 +22,11 @@ impl IRType {
     pub fn file_extension(&self) -> &'static str {
         match self {
             Self::Yul => revive_common::EXTENSION_YUL,
-            Self::EthIR => revive_common::EXTENSION_ETHIR,
-            Self::EVMLA => revive_common::EXTENSION_EVMLA,
             Self::LLVM => revive_common::EXTENSION_LLVM_SOURCE,
             Self::Assembly => revive_common::EXTENSION_POLKAVM_ASSEMBLY,
             #[cfg(debug_assertions)]
             Self::JSON => revive_common::EXTENSION_JSON,
+            Self::SO => revive_common::EXTENSION_SHARED_OBJECT,
         }
     }
 }

crates/llvm-context/src/debug_config/mod.rs

@@ -13,41 +13,28 @@ use self::ir_type::IRType;
 #[derive(Debug, Default, Serialize, Deserialize, Clone)]
 pub struct DebugConfig {
     /// The directory to dump the IRs to.
-    pub output_directory: PathBuf,
+    pub output_directory: Option<PathBuf>,
+    /// Whether debug info should be emitted.
+    pub emit_debug_info: bool,
 }
 
 impl DebugConfig {
     /// A shortcut constructor.
-    pub fn new(output_directory: PathBuf) -> Self {
+    pub const fn new(output_directory: Option<PathBuf>, emit_debug_info: bool) -> Self {
-        Self { output_directory }
+        Self {
+            output_directory,
+            emit_debug_info,
+        }
     }
 
     /// Dumps the Yul IR.
     pub fn dump_yul(&self, contract_path: &str, code: &str) -> anyhow::Result<()> {
-        let mut file_path = self.output_directory.to_owned();
+        if let Some(output_directory) = self.output_directory.as_ref() {
-        let full_file_name = Self::full_file_name(contract_path, None, IRType::Yul);
+            let mut file_path = output_directory.to_owned();
-        file_path.push(full_file_name);
+            let full_file_name = Self::full_file_name(contract_path, None, IRType::Yul);
-        std::fs::write(file_path, code)?;
+            file_path.push(full_file_name);
-
+            std::fs::write(file_path, code)?;
-        Ok(())
+        }
-    }
-
-    /// Dumps the EVM legacy assembly IR.
-    pub fn dump_evmla(&self, contract_path: &str, code: &str) -> anyhow::Result<()> {
-        let mut file_path = self.output_directory.to_owned();
-        let full_file_name = Self::full_file_name(contract_path, None, IRType::EVMLA);
-        file_path.push(full_file_name);
-        std::fs::write(file_path, code)?;
-
-        Ok(())
-    }
-
-    /// Dumps the Ethereal IR.
-    pub fn dump_ethir(&self, contract_path: &str, code: &str) -> anyhow::Result<()> {
-        let mut file_path = self.output_directory.to_owned();
-        let full_file_name = Self::full_file_name(contract_path, None, IRType::EthIR);
-        file_path.push(full_file_name);
-        std::fs::write(file_path, code)?;
 
         Ok(())
     }
@@ -58,12 +45,15 @@ impl DebugConfig {
         contract_path: &str,
         module: &inkwell::module::Module,
     ) -> anyhow::Result<()> {
-        let llvm_code = module.print_to_string().to_string();
+        if let Some(output_directory) = self.output_directory.as_ref() {
+            let llvm_code = module.print_to_string().to_string();
 
-        let mut file_path = self.output_directory.to_owned();
+            let mut file_path = output_directory.to_owned();
-        let full_file_name = Self::full_file_name(contract_path, Some("unoptimized"), IRType::LLVM);
+            let full_file_name =
-        file_path.push(full_file_name);
+                Self::full_file_name(contract_path, Some("unoptimized"), IRType::LLVM);
-        std::fs::write(file_path, llvm_code)?;
+            file_path.push(full_file_name);
+            std::fs::write(file_path, llvm_code)?;
+        }
 
         Ok(())
     }
@@ -74,22 +64,39 @@ impl DebugConfig {
         contract_path: &str,
         module: &inkwell::module::Module,
     ) -> anyhow::Result<()> {
-        let llvm_code = module.print_to_string().to_string();
+        if let Some(output_directory) = self.output_directory.as_ref() {
+            let llvm_code = module.print_to_string().to_string();
 
-        let mut file_path = self.output_directory.to_owned();
+            let mut file_path = output_directory.to_owned();
-        let full_file_name = Self::full_file_name(contract_path, Some("optimized"), IRType::LLVM);
+            let full_file_name =
-        file_path.push(full_file_name);
+                Self::full_file_name(contract_path, Some("optimized"), IRType::LLVM);
-        std::fs::write(file_path, llvm_code)?;
+            file_path.push(full_file_name);
+            std::fs::write(file_path, llvm_code)?;
+        }
 
         Ok(())
     }
 
     /// Dumps the assembly.
     pub fn dump_assembly(&self, contract_path: &str, code: &str) -> anyhow::Result<()> {
-        let mut file_path = self.output_directory.to_owned();
+        if let Some(output_directory) = self.output_directory.as_ref() {
-        let full_file_name = Self::full_file_name(contract_path, None, IRType::Assembly);
+            let mut file_path = output_directory.to_owned();
-        file_path.push(full_file_name);
+            let full_file_name = Self::full_file_name(contract_path, None, IRType::Assembly);
-        std::fs::write(file_path, code)?;
+            file_path.push(full_file_name);
+            std::fs::write(file_path, code)?;
+        }
+
+        Ok(())
+    }
+
+    /// Dumps the code object.
+    pub fn dump_object(&self, contract_path: &str, code: &[u8]) -> anyhow::Result<()> {
+        if let Some(output_directory) = self.output_directory.as_ref() {
+            let mut file_path = output_directory.to_owned();
+            let full_file_name = Self::full_file_name(contract_path, None, IRType::SO);
+            file_path.push(full_file_name);
+            std::fs::write(file_path, code)?;
+        }
 
         Ok(())
     }
@@ -102,10 +109,12 @@ impl DebugConfig {
         contract_suffix: Option<&str>,
         stage_json: &Vec<u8>,
     ) -> anyhow::Result<()> {
-        let mut file_path = self.output_directory.to_owned();
+        if let Some(output_directory) = self.output_directory.as_ref() {
-        let full_file_name = Self::full_file_name(contract_path, contract_suffix, IRType::JSON);
+            let mut file_path = output_directory.to_owned();
-        file_path.push(full_file_name);
+            let full_file_name = Self::full_file_name(contract_path, contract_suffix, IRType::JSON);
-        std::fs::write(file_path, stage_json)?;
+            file_path.push(full_file_name);
+            std::fs::write(file_path, stage_json)?;
+        }
 
         Ok(())
     }

crates/llvm-context/src/lib.rs

@@ -16,20 +16,18 @@ pub use self::polkavm::context::argument::Argument as PolkaVMArgument;
 pub use self::polkavm::context::attribute::Attribute as PolkaVMAttribute;
 pub use self::polkavm::context::build::Build as PolkaVMBuild;
 pub use self::polkavm::context::code_type::CodeType as PolkaVMCodeType;
-pub use self::polkavm::context::evmla_data::EVMLAData as PolkaVMContextEVMLAData;
+pub use self::polkavm::context::debug_info::DebugInfo;
-pub use self::polkavm::context::function::block::evmla_data::key::Key as PolkaVMFunctionBlockKey;
-pub use self::polkavm::context::function::block::evmla_data::EVMLAData as PolkaVMFunctionBlockEVMLAData;
-pub use self::polkavm::context::function::block::Block as PolkaVMFunctionBlock;
 pub use self::polkavm::context::function::declaration::Declaration as PolkaVMFunctionDeclaration;
-pub use self::polkavm::context::function::evmla_data::EVMLAData as PolkaVMFunctionEVMLAData;
 pub use self::polkavm::context::function::intrinsics::Intrinsics as PolkaVMIntrinsicFunction;
 pub use self::polkavm::context::function::llvm_runtime::LLVMRuntime as PolkaVMLLVMRuntime;
 pub use self::polkavm::context::function::r#return::Return as PolkaVMFunctionReturn;
 pub use self::polkavm::context::function::runtime::deploy_code::DeployCode as PolkaVMDeployCodeFunction;
 pub use self::polkavm::context::function::runtime::entry::Entry as PolkaVMEntryFunction;
+pub use self::polkavm::context::function::runtime::immutable_data_load::ImmutableDataLoad as PolkaVMImmutableDataLoadFunction;
 pub use self::polkavm::context::function::runtime::runtime_code::RuntimeCode as PolkaVMRuntimeCodeFunction;
 pub use self::polkavm::context::function::runtime::FUNCTION_DEPLOY_CODE as PolkaVMFunctionDeployCode;
 pub use self::polkavm::context::function::runtime::FUNCTION_ENTRY as PolkaVMFunctionEntry;
+pub use self::polkavm::context::function::runtime::FUNCTION_LOAD_IMMUTABLE_DATA as PolkaVMFunctionImmutableDataLoad;
 pub use self::polkavm::context::function::runtime::FUNCTION_RUNTIME_CODE as PolkaVMFunctionRuntimeCode;
 pub use self::polkavm::context::function::yul_data::YulData as PolkaVMFunctionYulData;
 pub use self::polkavm::context::function::Function as PolkaVMFunction;
@@ -58,7 +56,6 @@ pub use self::polkavm::evm::return_data as polkavm_evm_return_data;
 pub use self::polkavm::evm::storage as polkavm_evm_storage;
 pub use self::polkavm::metadata_hash::MetadataHash as PolkaVMMetadataHash;
 pub use self::polkavm::r#const as polkavm_const;
-pub use self::polkavm::utils as polkavm_utils;
 pub use self::polkavm::Dependency as PolkaVMDependency;
 pub use self::polkavm::DummyDependency as PolkaVMDummyDependency;
 pub use self::polkavm::DummyLLVMWritable as PolkaVMDummyLLVMWritable;

crates/llvm-context/src/optimizer/settings/mod.rs

@@ -9,7 +9,7 @@ use itertools::Itertools;
 
 use self::size_level::SizeLevel;
 
-/// The LLVM optimizer settings.
+/// The LLVM optimizer and code-gen settings.
 #[derive(Debug, Serialize, Deserialize, Clone, Eq)]
 pub struct Settings {
     /// The middle-end optimization level.
@@ -21,8 +21,6 @@ pub struct Settings {
 
     /// Fallback to optimizing for size if the bytecode is too large.
     pub is_fallback_to_size_enabled: bool,
-    /// Whether the system request memoization is disabled.
-    pub is_system_request_memoization_disabled: bool,
 
     /// Whether the LLVM `verify each` option is enabled.
     pub is_verify_each_enabled: bool,
@@ -43,7 +41,6 @@ impl Settings {
             level_back_end,
 
             is_fallback_to_size_enabled: false,
-            is_system_request_memoization_disabled: false,
 
             is_verify_each_enabled: false,
             is_debug_logging_enabled: false,
@@ -65,7 +62,6 @@ impl Settings {
             level_back_end,
 
             is_fallback_to_size_enabled: false,
-            is_system_request_memoization_disabled: false,
 
             is_verify_each_enabled,
             is_debug_logging_enabled,
@@ -206,20 +202,10 @@ impl Settings {
         self.is_fallback_to_size_enabled = true;
     }
 
-    /// Disables the system request memoization.
-    pub fn disable_system_request_memoization(&mut self) {
-        self.is_system_request_memoization_disabled = true;
-    }
-
     /// Whether the fallback to optimizing for size is enabled.
     pub fn is_fallback_to_size_enabled(&self) -> bool {
         self.is_fallback_to_size_enabled
     }
-
-    /// Whether the system request memoization is disabled.
-    pub fn is_system_request_memoization_disabled(&self) -> bool {
-        self.is_system_request_memoization_disabled
-    }
 }
 
 impl PartialEq for Settings {

crates/llvm-context/src/polkavm/const/mod.rs

@@ -1,54 +1,20 @@
 //! The LLVM context constants.
 
-/// Runtime API methods.
-pub mod runtime_api;
-
 /// The LLVM framework version.
 pub const LLVM_VERSION: semver::Version = semver::Version::new(18, 1, 4);
 
-/// The PolkaVM version.
+/// The pointer width sized type.
-pub const ZKEVM_VERSION: semver::Version = semver::Version::new(1, 3, 2);
-
-/// The register width sized type
 pub static XLEN: usize = revive_common::BIT_LENGTH_X32;
 
 /// The heap memory pointer pointer global variable name.
 pub static GLOBAL_HEAP_MEMORY_POINTER: &str = "memory_pointer";
 
-/// The calldata pointer global variable name.
-pub static GLOBAL_CALLDATA_POINTER: &str = "ptr_calldata";
-
 /// The calldata size global variable name.
 pub static GLOBAL_CALLDATA_SIZE: &str = "calldatasize";
 
 /// The call flags global variable name.
 pub static GLOBAL_CALL_FLAGS: &str = "call_flags";
 
-/// The constant array global variable name prefix.
-pub static GLOBAL_CONST_ARRAY_PREFIX: &str = "const_array_";
-
-/// The global verbatim getter identifier prefix.
-pub static GLOBAL_VERBATIM_GETTER_PREFIX: &str = "get_global::";
-
-/// The external call data offset in the auxiliary heap.
-pub const HEAP_AUX_OFFSET_EXTERNAL_CALL: u64 = 0;
-
-/// The constructor return data offset in the auxiliary heap.
-pub const HEAP_AUX_OFFSET_CONSTRUCTOR_RETURN_DATA: u64 =
-    8 * (revive_common::BYTE_LENGTH_WORD as u64);
-
-/// The `create` method deployer signature.
-pub static DEPLOYER_SIGNATURE_CREATE: &str = "create(bytes32,bytes32,bytes)";
-
-/// The `create2` method deployer signature.
-pub static DEPLOYER_SIGNATURE_CREATE2: &str = "create2(bytes32,bytes32,bytes)";
-
-/// The absence of system call bit.
-pub const NO_SYSTEM_CALL_BIT: bool = false;
-
-/// The system call bit.
-pub const SYSTEM_CALL_BIT: bool = true;
-
 /// The deployer call header size that consists of:
 /// - bytecode hash (32 bytes)
 pub const DEPLOYER_CALL_HEADER_SIZE: usize = revive_common::BYTE_LENGTH_WORD;

crates/llvm-context/src/polkavm/const/runtime_api.rs

@@ -1,77 +0,0 @@
-//! Runtime API import and export symbols.
-
-pub mod exports {
-    /// The contract deploy export.
-    pub static CALL: &str = "call";
-
-    /// The contract call export.
-    pub static DEPLOY: &str = "deploy";
-
-    /// All exported symbols.
-    /// Useful for configuring common attributes and linkage.
-    pub static EXPORTS: [&str; 2] = [CALL, DEPLOY];
-}
-
-pub mod imports {
-    pub static ADDRESS: &str = "address";
-
-    pub static BALANCE: &str = "balance";
-
-    pub static BLOCK_NUMBER: &str = "block_number";
-
-    pub static CHAIN_ID: &str = "chain_id";
-
-    pub static CALL: &str = "call";
-
-    pub static CALLER: &str = "caller";
-
-    pub static CODE_SIZE: &str = "code_size";
-
-    pub static DEPOSIT_EVENT: &str = "deposit_event";
-
-    pub static GET_STORAGE: &str = "get_storage";
-
-    pub static HASH_KECCAK_256: &str = "hash_keccak_256";
-
-    pub static INPUT: &str = "input";
-
-    pub static INSTANTIATE: &str = "instantiate";
-
-    pub static NOW: &str = "now";
-
-    pub static RETURN: &str = "seal_return";
-
-    pub static RETURNDATACOPY: &str = "return_data_copy";
-
-    pub static RETURNDATASIZE: &str = "return_data_size";
-
-    pub static SET_STORAGE: &str = "set_storage";
-
-    pub static VALUE_TRANSFERRED: &str = "value_transferred";
-
-    /// All imported runtime API symbols.
-    /// Useful for configuring common attributes and linkage.
-    pub static IMPORTS: [&str; 18] = [
-        ADDRESS,
-        BALANCE,
-        BLOCK_NUMBER,
-        CALL,
-        CALLER,
-        CHAIN_ID,
-        CODE_SIZE,
-        DEPOSIT_EVENT,
-        GET_STORAGE,
-        HASH_KECCAK_256,
-        INPUT,
-        INSTANTIATE,
-        NOW,
-        RETURN,
-        RETURNDATACOPY,
-        RETURNDATASIZE,
-        SET_STORAGE,
-        VALUE_TRANSFERRED,
-    ];
-}
-
-/// PolkaVM __sbrk API symbol to extend the heap memory.
-pub static SBRK: &str = "__sbrk";

crates/llvm-context/src/polkavm/context/debug_info.rs

@@ -1,7 +1,43 @@
 //! The LLVM debug information.
 
+use std::cell::RefCell;
+
 use inkwell::debug_info::AsDIScope;
-use num::Zero;
+use inkwell::debug_info::DIScope;
+
+/// Debug info scope stack
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub struct ScopeStack<'ctx> {
+    stack: Vec<DIScope<'ctx>>,
+}
+
+// Abstract the type of the DIScope stack.
+impl<'ctx> ScopeStack<'ctx> {
+    pub fn from(item: DIScope<'ctx>) -> Self {
+        Self { stack: vec![item] }
+    }
+
+    /// Return the top of the scope stack, or None if the stack is empty.
+    pub fn top(&self) -> Option<DIScope<'ctx>> {
+        self.stack.last().copied()
+    }
+
+    /// Push a scope onto the stack.
+    pub fn push(&mut self, scope: DIScope<'ctx>) {
+        self.stack.push(scope)
+    }
+
+    /// Pop the scope at the top of the stack and return it.
+    /// Return None if the stack is empty.
+    pub fn pop(&mut self) -> Option<DIScope<'ctx>> {
+        self.stack.pop()
+    }
+
+    /// Return the number of scopes on the stack.
+    pub fn len(&self) -> usize {
+        self.stack.len()
+    }
+}
 
 /// The LLVM debug information.
 pub struct DebugInfo<'ctx> {
@@ -9,6 +45,8 @@ pub struct DebugInfo<'ctx> {
     compile_unit: inkwell::debug_info::DICompileUnit<'ctx>,
     /// The debug info builder.
     builder: inkwell::debug_info::DebugInfoBuilder<'ctx>,
+    /// Enclosing debug info scopes.
+    scope_stack: RefCell<ScopeStack<'ctx>>,
 }
 
 impl<'ctx> DebugInfo<'ctx> {
@@ -35,19 +73,43 @@ impl<'ctx> DebugInfo<'ctx> {
         Self {
             compile_unit,
             builder,
+            scope_stack: RefCell::new(ScopeStack::from(compile_unit.as_debug_info_scope())),
         }
     }
 
+    /// Prepare an LLVM-IR module for debug-info generation
+    pub fn initialize_module(
+        &self,
+        llvm: &'ctx inkwell::context::Context,
+        module: &inkwell::module::Module<'ctx>,
+    ) {
+        let debug_metadata_value = llvm
+            .i32_type()
+            .const_int(inkwell::debug_info::debug_metadata_version() as u64, false);
+        module.add_basic_value_flag(
+            "Debug Info Version",
+            inkwell::module::FlagBehavior::Warning,
+            debug_metadata_value,
+        );
+        self.push_scope(self.compilation_unit().get_file().as_debug_info_scope());
+    }
+
+    /// Finalize debug-info for an LLVM-IR module.
+    pub fn finalize_module(&self) {
+        self.builder().finalize()
+    }
+
     /// Creates a function info.
     pub fn create_function(
         &self,
         name: &str,
     ) -> anyhow::Result<inkwell::debug_info::DISubprogram<'ctx>> {
+        let flags = inkwell::debug_info::DIFlagsConstants::ZERO;
         let subroutine_type = self.builder.create_subroutine_type(
             self.compile_unit.get_file(),
-            Some(self.create_type(revive_common::BIT_LENGTH_FIELD)?),
+            Some(self.create_word_type(Some(flags))?.as_type()),
             &[],
-            inkwell::debug_info::DIFlags::zero(),
+            flags,
         );
 
         let function = self.builder.create_function(
@@ -60,7 +122,7 @@ impl<'ctx> DebugInfo<'ctx> {
             true,
             false,
             1,
-            inkwell::debug_info::DIFlags::zero(),
+            flags,
             false,
         );
 
@@ -74,24 +136,55 @@ impl<'ctx> DebugInfo<'ctx> {
         Ok(function)
     }
 
-    /// Creates a primitive type info.
+    /// Creates primitive integer type debug-info.
-    pub fn create_type(
+    pub fn create_primitive_type(
         &self,
         bit_length: usize,
-    ) -> anyhow::Result<inkwell::debug_info::DIType<'ctx>> {
+        flags: Option<inkwell::debug_info::DIFlags>,
+    ) -> anyhow::Result<inkwell::debug_info::DIBasicType<'ctx>> {
+        let di_flags = flags.unwrap_or(inkwell::debug_info::DIFlagsConstants::ZERO);
+        let di_encoding: u32 = 0;
+        let type_name = String::from("U") + bit_length.to_string().as_str();
         self.builder
-            .create_basic_type(
+            .create_basic_type(type_name.as_str(), bit_length as u64, di_encoding, di_flags)
-                "U256",
-                bit_length as u64,
-                0,
-                inkwell::debug_info::DIFlags::zero(),
-            )
-            .map(|basic_type| basic_type.as_type())
             .map_err(|error| anyhow::anyhow!("Debug info error: {}", error))
     }
 
-    /// Finalizes the builder.
+    /// Returns the debug-info model of word-sized integer types.
-    pub fn finalize(&self) {
+    pub fn create_word_type(
-        self.builder.finalize();
+        &self,
+        flags: Option<inkwell::debug_info::DIFlags>,
+    ) -> anyhow::Result<inkwell::debug_info::DIBasicType<'ctx>> {
+        self.create_primitive_type(revive_common::BIT_LENGTH_WORD, flags)
+    }
+
+    /// Return the DIBuilder.
+    pub fn builder(&self) -> &inkwell::debug_info::DebugInfoBuilder<'ctx> {
+        &self.builder
+    }
+
+    /// Return the compilation unit. {
+    pub fn compilation_unit(&self) -> &inkwell::debug_info::DICompileUnit<'ctx> {
+        &self.compile_unit
+    }
+
+    /// Push a debug-info scope onto the stack.
+    pub fn push_scope(&self, scope: DIScope<'ctx>) {
+        self.scope_stack.borrow_mut().push(scope)
+    }
+
+    /// Pop the top of the debug-info scope stack and return it.
+    pub fn pop_scope(&self) -> Option<DIScope<'ctx>> {
+        self.scope_stack.borrow_mut().pop()
+    }
+
+    /// Return the top of the debug-info scope stack.
+    pub fn top_scope(&self) -> Option<DIScope<'ctx>> {
+        self.scope_stack.borrow().top()
+    }
+
+    /// Return the number of debug-info scopes on the scope stack.
+    pub fn num_scopes(&self) -> usize {
+        self.scope_stack.borrow().len()
     }
 }

crates/llvm-context/src/polkavm/context/evmla_data.rs

@@ -1,27 +0,0 @@
-//! The LLVM IR generator EVM legacy assembly data.
-
-use crate::polkavm::context::argument::Argument;
-
-/// The LLVM IR generator EVM legacy assembly data.
-/// Describes some data that is only relevant to the EVM legacy assembly.
-#[derive(Debug, Clone)]
-pub struct EVMLAData<'ctx> {
-    /// The Solidity compiler version.
-    /// Some instruction behave differenly depending on the version.
-    pub version: semver::Version,
-    /// The static stack allocated for the current function.
-    pub stack: Vec<Argument<'ctx>>,
-}
-
-impl<'ctx> EVMLAData<'ctx> {
-    /// The default stack size.
-    pub const DEFAULT_STACK_SIZE: usize = 64;
-
-    /// A shortcut constructor.
-    pub fn new(version: semver::Version) -> Self {
-        Self {
-            version,
-            stack: Vec::with_capacity(Self::DEFAULT_STACK_SIZE),
-        }
-    }
-}

crates/llvm-context/src/polkavm/context/function/block/evmla_data/key.rs

@@ -1,34 +0,0 @@
-//! The LLVM IR generator function block key.
-
-use crate::polkavm::context::code_type::CodeType;
-
-/// The LLVM IR generator function block key.
-/// Is only relevant to the EVM legacy assembly.
-#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)]
-pub struct Key {
-    /// The block code type.
-    pub code_type: CodeType,
-    /// The block tag.
-    pub tag: num::BigUint,
-}
-
-impl Key {
-    /// A shortcut constructor.
-    pub fn new(code_type: CodeType, tag: num::BigUint) -> Self {
-        Self { code_type, tag }
-    }
-}
-
-impl std::fmt::Display for Key {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(
-            f,
-            "{}_{}",
-            match self.code_type {
-                CodeType::Deploy => "dt",
-                CodeType::Runtime => "rt",
-            },
-            self.tag
-        )
-    }
-}

crates/llvm-context/src/polkavm/context/function/block/evmla_data/mod.rs

@@ -1,18 +0,0 @@
-//! The LLVM function block EVM legacy assembly data.
-
-pub mod key;
-
-/// The LLVM function block EVM legacy assembly data.
-/// Describes some data that is only relevant to the EVM legacy assembly.
-#[derive(Debug, Clone)]
-pub struct EVMLAData {
-    /// The initial hashes of the allowed stack states.
-    pub stack_hashes: Vec<md5::Digest>,
-}
-
-impl EVMLAData {
-    /// A shortcut constructor.
-    pub fn new(stack_hashes: Vec<md5::Digest>) -> Self {
-        Self { stack_hashes }
-    }
-}

crates/llvm-context/src/polkavm/context/function/block/mod.rs

@@ -1,52 +0,0 @@
-//! The LLVM IR generator function block.
-
-pub mod evmla_data;
-
-use self::evmla_data::EVMLAData;
-
-/// The LLVM IR generator function block.
-#[derive(Debug, Clone)]
-pub struct Block<'ctx> {
-    /// The inner block.
-    inner: inkwell::basic_block::BasicBlock<'ctx>,
-    /// The EVM legacy assembly compiler data.
-    evmla_data: Option<EVMLAData>,
-}
-
-impl<'ctx> Block<'ctx> {
-    /// A shortcut constructor.
-    pub fn new(inner: inkwell::basic_block::BasicBlock<'ctx>) -> Self {
-        Self {
-            inner,
-            evmla_data: None,
-        }
-    }
-
-    /// Sets the EVM legacy assembly data.
-    pub fn set_evmla_data(&mut self, data: EVMLAData) {
-        self.evmla_data = Some(data);
-    }
-
-    /// The LLVM object reference.
-    pub fn inner(&self) -> inkwell::basic_block::BasicBlock<'ctx> {
-        self.inner
-    }
-
-    /// Returns the EVM data reference.
-    /// # Panics
-    /// If the EVM data has not been initialized.
-    pub fn evm(&self) -> &EVMLAData {
-        self.evmla_data
-            .as_ref()
-            .expect("The EVM data must have been initialized")
-    }
-
-    /// Returns the EVM data mutable reference.
-    /// # Panics
-    /// If the EVM data has not been initialized.
-    pub fn evm_mut(&mut self) -> &mut EVMLAData {
-        self.evmla_data
-            .as_mut()
-            .expect("The EVM data must have been initialized")
-    }
-}