chore[vortex-array]: add integer casting benchmark

The motivation is to compare an upcoming perf improvement.

Signed-off-by: Alfonso Subiotto Marques <alfonso.subiotto@polarsignals.com>

Author: asubiotto

vortex-array/Cargo.toml

@@ -96,6 +96,10 @@ serde_json = { workspace = true }
 serde_test = { workspace = true }
 vortex-array = { path = ".", features = ["_test-harness", "table-display"] }
 
+[[bench]]
+name = "cast_primitive"
+harness = false
+
 [[bench]]
 name = "search_sorted"
 harness = false

vortex-array/benches/cast_primitive.rs

@@ -0,0 +1,37 @@
+// SPDX-License-Identifier: Apache-2.0
+// SPDX-FileCopyrightText: Copyright the Vortex contributors
+
+use divan::Bencher;
+use rand::prelude::*;
+use vortex_array::IntoArray;
+use vortex_array::arrays::PrimitiveArray;
+use vortex_array::builtins::ArrayBuiltins;
+use vortex_array::dtype::DType;
+use vortex_array::dtype::Nullability;
+use vortex_array::dtype::PType;
+
+fn main() {
+    divan::main();
+}
+
+const N: usize = 100_000;
+
+#[divan::bench]
+fn cast_u16_to_u32(bencher: Bencher) {
+    let mut rng = StdRng::seed_from_u64(42);
+    #[expect(clippy::cast_possible_truncation)]
+    let arr = PrimitiveArray::from_option_iter((0..N).map(|i| {
+        if rng.random_bool(0.5) {
+            None
+        } else {
+            Some(i as u16)
+        }
+    }))
+    .into_array();
+    bencher.with_inputs(|| arr.clone()).bench_refs(|a| {
+        #[expect(clippy::unwrap_used)]
+        a.cast(DType::Primitive(PType::U32, Nullability::Nullable))
+            .unwrap()
+            .to_canonical()
+    });
+}