feat: build with stable rust (#3881)

Author: a10y

.github/actions/setup-rust/action.yml

@@ -2,6 +2,12 @@ name: "Setup Rust"
 description: "Toolchain setup and Initial compilation"
 
 inputs:
+  toolchain:
+    description: "optional override for the toolchain version (e.g. nightly)"
+    required: false
+  components:
+    description: "optional override for the components to install for the step (e.g. clippy, rustfmt, miri)"
+    required: false
   targets:
     description: "optional targets override (e.g. wasm32-unknown-unknown)"
     required: false
@@ -22,9 +28,9 @@ runs:
       uses: dtolnay/rust-toolchain@master
       if: steps.rustup-cache.outputs.cache-hit != 'true'
       with:
-        toolchain: "${{ steps.rust-version.outputs.version }}"
+        toolchain: "${{ inputs.toolchain || steps.rust-version.outputs.version }}"
         targets: "${{inputs.targets || ''}}"
-        components: clippy, rustfmt, miri, llvm-tools-preview
+        components: "${{ inputs.components || 'clippy, rustfmt' }}"
 
     - name: Rust Dependency Cache
       uses: Swatinem/rust-cache@v2

.github/workflows/ci.yml

@@ -314,7 +314,7 @@ jobs:
       - name: Rust Tests
         if: ${{ matrix.suite == 'tests' }}
         run: |
-          cargo nextest run --locked --workspace --all-features --no-fail-fast
+          cargo +nightly nextest run --locked --workspace --all-features --no-fail-fast
       - name: Run TPC-H
         if: ${{ matrix.suite == 'tpc-h' }}
         run: |
@@ -364,14 +364,17 @@ jobs:
         with:
           submodules: "recursive"
       - uses: ./.github/actions/setup-rust
+        with:
+          toolchain: nightly
+          components: "rust-src, rustfmt, clippy, llvm-tools-preview"
       - name: Install nextest
         uses: taiki-e/install-action@v2
         with:
           tool: [email protected]
       - name: Rust Tests
         # vortex-duckdb-ext currently fails linking for cargo test targets.
         run: |
-          cargo nextest run --locked --workspace --all-features --no-fail-fast --target x86_64-unknown-linux-gnu
+          cargo +nightly nextest run --locked --workspace --all-features --no-fail-fast --target x86_64-unknown-linux-gnu
 
   build-java:
     name: "Java"
@@ -463,10 +466,12 @@ jobs:
       - uses: actions/checkout@v4
       - uses: ./.github/actions/setup-rust
         with:
+          toolchain: nightly
           targets: "wasm32-wasip1"
+          components: "rust-src"
       - name: Setup Wasmer
         uses: wasmerio/[email protected]
-      - run: cargo -Zbuild-std=panic_abort,std build --target wasm32-wasip1
+      - run: cargo +nightly -Zbuild-std=panic_abort,std build --target wasm32-wasip1
         working-directory: ./wasm-test
       - run: wasmer run ./target/wasm32-wasip1/debug/wasm-test.wasm
         working-directory: ./wasm-test
@@ -484,19 +489,14 @@ jobs:
         with:
           submodules: "recursive"
       - uses: ./.github/actions/setup-rust
+        with:
+          toolchain: nightly
+          components: "rust-src, rustfmt, clippy, miri"
       - uses: taiki-e/install-action@v2
         with:
           tool: [email protected]
-      - name: Run all tests with Miri
-        # For now, we only run Miri against known "fiddly" crates.
-        if: false
-        run: |
-          cargo miri nextest run --no-fail-fast --workspace --exclude vortex-file --exclude vortex-layout \
-          --exclude vortex-fsst --exclude vortex-array --exclude vortex-dtype --exclude vortex-expr \
-          --exclude vortex-scalar --exclude vortex-duckdb --exclude vortex-ffi --exclude bench-vortex \
-          --exclude vortex-python --exclude vortex-jni
       - name: Run Miri
-        run: cargo miri nextest run --no-fail-fast -p vortex-buffer -p vortex-ffi
+        run: cargo +nightly miri nextest run --no-fail-fast -p vortex-buffer -p vortex-ffi
 
   generated-files:
     name: "Check generated proto/fbs files are up to date"

.github/workflows/fuzz.yml

@@ -24,6 +24,8 @@ jobs:
         with:
           submodules: "recursive"
       - uses: ./.github/actions/setup-rust
+        with:
+          rust-version: nightly
       - name: Install cargo fuzz
         run: cargo install --locked cargo-fuzz
       - name: Restore corpus

Cargo.lock

@@ -50,7 +50,7 @@ keywords = ["vortex"]
 license = "Apache-2.0"
 readme = "README.md"
 repository = "https://github.com/spiraldb/vortex"
-rust-version = "1.86"
+rust-version = "1.87"
 version = "0.1.0"
 
 [workspace.dependencies]
@@ -240,6 +240,10 @@ redundant_lifetimes = "deny"
 unsafe_op_in_unsafe_fn = "deny"
 unused_lifetimes = "deny"
 unused_qualifications = "deny"
+unexpected_cfgs = { level = "deny", check-cfg = [
+    "cfg(codspeed)",
+    "cfg(vortex_nightly)",
+] }
 warnings = "deny"
 
 [workspace.lints.clippy]

Cargo.toml

@@ -106,7 +106,7 @@ impl BenchmarkDataset {
     }
 
     pub fn format_path(&self, format: Format, base_url: &Url) -> Result<Url> {
-        Ok(base_url.join(&format!("{}/", format))?)
+        Ok(base_url.join(&format!("{format}/"))?)
     }
 
     pub async fn register_tables(

bench-vortex/src/datasets/mod.rs

@@ -66,7 +66,7 @@ impl DuckDBCtx {
 
     /// Execute DuckDB queries for benchmarks using the internal connection
     pub fn execute_query(&self, query: &str) -> Result<(Duration, usize)> {
-        trace!("execute duckdb query: {}", query);
+        trace!("execute duckdb query: {query}");
         let time_instant = Instant::now();
         let result = self.connection.query(query)?;
         let query_time = time_instant.elapsed();
@@ -104,7 +104,7 @@ impl DuckDBCtx {
         // Generate and execute table registration commands
         let commands = self.generate_table_commands(&effective_url, extension, dataset, object);
         self.execute_query(&commands)?;
-        trace!("Executing table registration commands: {}", commands);
+        trace!("Executing table registration commands: {commands}");
 
         Ok(())
     }

bench-vortex/src/engines/ddb/mod.rs

@@ -286,7 +286,7 @@ impl TpcHBenchmark {
                         ChangeTag::Insert => "+",
                         ChangeTag::Equal => " ",
                     };
-                    print!("{}{}", sign, change);
+                    print!("{sign}{change}");
                 }
 
                 eprintln!("query output does not match the reference for {query_name}");

bench-vortex/src/tpch/tpch_benchmark.rs

@@ -2,10 +2,12 @@ allow-dbg-in-tests = true
 allow-expect-in-tests = true
 allow-unwrap-in-tests = true
 single-char-binding-names-threshold = 2
-# We prefer using parking_lot for improved ergonomics and performance.
-disallowed-types = ["std::collections::HashMap", "std::collections::HashSet", "std::sync::Mutex", "std::sync::RwLock"]
+disallowed-types = [
+	{ path = "std::collections::HashMap", reason = "Use the HashMap in vortex_utils::aliases for better performance" }, 
+	{ path = "std::collections::HashSet", reason = "Use the HashSet in vortex_utils::aliases for better performance" },
+	{ path = "std::sync::Mutex", reason = "Prefer using parking_lot Mutex for improved ergonomics and performance" },
+	{ path = "std::sync::RwLock", reason = "Prefer using parking_lot RwLock for improved ergonomics and performance" }]
 
 disallowed-methods = [
-    # It uses the default hasher and is very easy to just inline with a faster implementation
-    "itertools::Itertools::counts"
-    ]
+    { path = "itertools::Itertools::counts", reason = "It uses the default hasher which is slow for primitives. Just inline the loop for better performance." }
+]

clippy.toml

@@ -216,7 +216,7 @@ fn encode_chunk_unchecked<T: ALPFloat>(
 
     // encode the chunk, counting the number of patches
     let mut chunk_patch_count = 0;
-    encoded_output.extend(chunk.iter().map(|&v| {
+    encoded_output.extend_trusted(chunk.iter().map(|&v| {
         let encoded = T::encode_single_unchecked(v, exp);
         let decoded = T::decode_single(encoded, exp);
         let neq = !decoded.is_eq(v) as usize;