From 6000df9779911a241423b0ff4c6576765fdc4267 Mon Sep 17 00:00:00 2001
From: Greg Fitzgerald <greg@solana.com>
Date: Mon, 7 Jan 2019 09:38:19 -0700
Subject: [PATCH] Optimize has_duplicates() for short slices

---
 benches/runtime.rs | 14 ++++++++++++++
 src/runtime.rs     | 15 ++++++++++-----
 2 files changed, 24 insertions(+), 5 deletions(-)
 create mode 100644 benches/runtime.rs

diff --git a/benches/runtime.rs b/benches/runtime.rs
new file mode 100644
index 0000000000..ecaf4176b8
--- /dev/null
+++ b/benches/runtime.rs
@@ -0,0 +1,14 @@
+#![feature(test)]
+
+extern crate test;
+
+use solana::runtime::*;
+use test::Bencher;
+
+#[bench]
+fn bench_has_duplicates(bencher: &mut Bencher) {
+    bencher.iter(|| {
+        let data = test::black_box([1, 2, 3]);
+        assert!(!has_duplicates(&data));
+    })
+}
diff --git a/src/runtime.rs b/src/runtime.rs
index 71b145fd14..2eb63d060d 100644
--- a/src/runtime.rs
+++ b/src/runtime.rs
@@ -5,8 +5,6 @@ use solana_sdk::pubkey::Pubkey;
 use solana_sdk::system_program;
 use solana_sdk::transaction::Transaction;
 use solana_system_program;
-use std::collections::HashSet;
-use std::iter::FromIterator;
 
 /// Reasons the runtime might have rejected a transaction.
 #[derive(Debug, PartialEq, Eq, Clone)]
@@ -118,9 +116,16 @@ fn execute_instruction(
 }
 
 /// Return true if the slice has any duplicate elements
-fn has_duplicates<T: Eq + std::hash::Hash>(xs: &[T]) -> bool {
-    let xs_set: HashSet<&T> = HashSet::from_iter(xs.iter());
-    xs.len() != xs_set.len()
+pub fn has_duplicates<T: PartialEq>(xs: &[T]) -> bool {
+    // Note: This is an O(n^2) algorithm, but requires no heap allocations. The benchmark
+    // `bench_has_duplicates` in benches/runtime.rs shows that this implementation is
+    // ~50 times faster than using HashSet for very short slices.
+    for i in 1..xs.len() {
+        if xs[i..].contains(&xs[i - 1]) {
+            return true;
+        }
+    }
+    false
 }
 
 /// Get mut references to a subset of elements.