bench: use mimalloc as the global allocator in every benchmark binary

docs/developer-guide/benchmarking.md

@@ -182,17 +182,23 @@ trace — including cache and memory access costs. This has several implications
   [walltime instrument](https://codspeed.io/docs/instruments/walltime) or be gated with
   `#[cfg(not(codspeed))]`.
 
-### Prefer `mimalloc` for throughput benchmarks
+### Use `mimalloc` as the global allocator
 
-Throughput benchmarks should use `mimalloc` as the global allocator to reduce system allocator
-noise:
+Every benchmark binary uses `mimalloc` as its global allocator:
 
 ```rust
 use mimalloc::MiMalloc;
 #[global_allocator]
 static GLOBAL: MiMalloc = MiMalloc;
 ```
 
+The system allocator's cost depends on its state, which differs between runner images and between
+runs, so allocation inside a timed region made several simulation benchmarks flip between two
+values on pull requests that could not have affected them. `mimalloc` does the same work every
+time, and one allocator for every binary means no benchmark measures a different allocator from
+its neighbours. Add the two lines to every new benchmark file; each crate with benchmarks already
+has the `mimalloc` dev-dependency.
+
 ## SQL Benchmarks
 
 SQL benchmarks measure end-to-end query performance across different engines and file formats.

encodings/alp/Cargo.toml

@@ -30,6 +30,7 @@ vortex-session = { workspace = true }
 
 [dev-dependencies]
 divan = { workspace = true }
+mimalloc = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }
 vortex-array = { workspace = true, features = ["_test-harness"] }

encodings/alp/benches/alp_compress.rs

@@ -6,6 +6,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use rand::RngExt;
 use rand::SeedableRng as _;
 use rand::rngs::StdRng;
@@ -25,6 +26,9 @@ use vortex_buffer::Buffer;
 use vortex_buffer::buffer;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/datetime-parts/Cargo.toml

@@ -27,6 +27,7 @@ vortex-session = { workspace = true }
 
 [dev-dependencies]
 divan = { workspace = true }
+mimalloc = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }
 vortex-array = { workspace = true, features = ["_test-harness"] }

encodings/datetime-parts/benches/split_temporal.rs

@@ -6,6 +6,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use rand::RngExt;
 use rand::SeedableRng as _;
 use rand::rngs::StdRng;
@@ -19,6 +20,9 @@ use vortex_buffer::Buffer;
 use vortex_datetime_parts::split_temporal;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/decimal-byte-parts/Cargo.toml

@@ -28,6 +28,7 @@ vortex-session = { workspace = true }
 [dev-dependencies]
 divan = { workspace = true }
 hegeltest = { workspace = true }
+mimalloc = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }
 vortex-array = { path = "../../vortex-array", features = ["_test-harness"] }

encodings/decimal-byte-parts/benches/dbp_assemble.rs

@@ -8,6 +8,7 @@ mod common;
 
 use divan::Bencher;
 use divan::black_box;
+use mimalloc::MiMalloc;
 use vortex_array::dtype::DecimalType;
 use vortex_array::dtype::i256;
 use vortex_buffer::buffer;
@@ -22,6 +23,9 @@ use crate::common::cases;
 use crate::common::i128_values;
 use crate::common::i256_values;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/decimal-byte-parts/benches/dbp_split.rs

@@ -8,6 +8,7 @@ mod common;
 
 use divan::Bencher;
 use divan::black_box;
+use mimalloc::MiMalloc;
 use rand::RngExt;
 use rand::SeedableRng;
 use rand::rngs::StdRng;
@@ -22,6 +23,9 @@ use crate::common::cases;
 use crate::common::i128_values;
 use crate::common::i256_values;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/Cargo.toml

@@ -35,6 +35,7 @@ vortex-session = { workspace = true }
 [dev-dependencies]
 divan = { workspace = true }
 itertools = { workspace = true }
+mimalloc = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }
 vortex-alp = { path = "../alp" }

encodings/fastlanes/benches/bitpack_compare.rs

@@ -14,6 +14,7 @@ use std::sync::LazyLock;
 
 use divan::Bencher;
 use divan::counter::ItemsCount;
+use mimalloc::MiMalloc;
 use vortex_array::ArrayRef;
 use vortex_array::ExecutionCtx;
 use vortex_array::IntoArray;
@@ -31,6 +32,9 @@ use vortex_fastlanes::BitPackedArray;
 use vortex_fastlanes::BitPackedData;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/benches/bitpack_compare_sweep.rs

@@ -16,6 +16,7 @@ use std::sync::LazyLock;
 
 use divan::Bencher;
 use divan::counter::ItemsCount;
+use mimalloc::MiMalloc;
 use vortex_array::ArrayRef;
 use vortex_array::ExecutionCtx;
 use vortex_array::IntoArray;
@@ -35,6 +36,9 @@ use vortex_fastlanes::BitPackedArray;
 use vortex_fastlanes::BitPackedData;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/benches/bitpacking_take.rs

@@ -7,6 +7,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use rand::RngExt;
 use rand::SeedableRng;
 use rand::distr::Uniform;
@@ -22,6 +23,9 @@ use vortex_fastlanes::BitPackedArrayExt;
 use vortex_fastlanes::bitpack_compress::bitpack_to_best_bit_width;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/benches/canonicalize_bench.rs

@@ -4,6 +4,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use rand::SeedableRng;
 use rand::prelude::StdRng;
 use vortex_array::Canonical;
@@ -16,6 +17,9 @@ use vortex_error::VortexExpect;
 use vortex_fastlanes::bitpack_compress::test_harness::make_array;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/benches/cast_bitpacked.rs

@@ -15,6 +15,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use rand::RngExt;
 use rand::SeedableRng;
 use rand::prelude::StdRng;
@@ -34,6 +35,9 @@ use vortex_fastlanes::BitPackedArray;
 use vortex_fastlanes::BitPackedData;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fastlanes/benches/compute_between.rs

@@ -4,6 +4,7 @@
 
 use std::sync::LazyLock;
 
+use mimalloc::MiMalloc;
 use num_traits::NumCast;
 use rand::RngExt;
 use rand::rngs::StdRng;
@@ -20,6 +21,9 @@ use vortex_error::VortexExpect;
 use vortex_fastlanes::bitpack_compress::bitpack_to_best_bit_width;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }

encodings/fsst/Cargo.toml

@@ -32,6 +32,7 @@ _test-harness = ["dep:rand", "vortex-array/_test-harness"]
 
 [dev-dependencies]
 divan = { workspace = true }
+mimalloc = { workspace = true }
 rand = { workspace = true }
 rstest = { workspace = true }
 test-with = { workspace = true }

encodings/fsst/benches/chunked_dict_fsst_builder.rs

@@ -4,6 +4,7 @@
 use std::sync::LazyLock;
 
 use divan::Bencher;
+use mimalloc::MiMalloc;
 use vortex_array::ArrayRef;
 use vortex_array::Canonical;
 use vortex_array::IntoArray;
@@ -15,6 +16,9 @@ use vortex_error::VortexExpect;
 use vortex_fsst::test_utils::gen_dict_fsst_test_data;
 use vortex_session::VortexSession;
 
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
 fn main() {
     divan::main();
 }