Auto merge of #83 - edre:bench, r=Amanieu

Add the std hasher and different key distributions to benchmarks.

Also remove the i32/i64 duplication, as those differences were in the noise compared to the other dimensions.

I would have liked to reduce some of the macro duplication, but it looks like that's still not easy.
rust-lang/rust#29599

benches/bench.rs

@@ -1,150 +1,234 @@
+// This benchmark suite contains some benchmarks along a set of dimensions:
+//   Hasher: std default (SipHash) and crate default (FxHash).
+//   Int key distribution: low bit heavy, top bit heavy, and random.
+//   Task: basic functionality: insert, insert_erase, lookup, lookup_fail, iter
 #![feature(test)]
 
 extern crate test;
 
-use std::hash::Hash;
 use test::{black_box, Bencher};
 
+use hashbrown::hash_map::DefaultHashBuilder;
 use hashbrown::HashMap;
-//use rustc_hash::FxHashMap as HashMap;
-//use std::collections::HashMap;
+use std::collections::hash_map::RandomState;
 
-fn new_map<K: Eq + Hash, V>() -> HashMap<K, V> {
-    HashMap::default()
-}
-
-#[bench]
-fn insert_i32(b: &mut Bencher) {
-    b.iter(|| {
-        let mut m: HashMap<i32, i32> = new_map();
-        for i in 1..1001 {
-            m.insert(i, i);
-        }
-        black_box(m);
-    })
-}
+const SIZE: usize = 1000;
 
-#[bench]
-fn insert_i64(b: &mut Bencher) {
-    b.iter(|| {
-        let mut m: HashMap<i64, i64> = new_map();
-        for i in 1..1001 {
-            m.insert(i, i);
-        }
-        black_box(m);
-    })
-}
+// The default hashmap when using this crate directly.
+type FxHashMap<K, V> = HashMap<K, V, DefaultHashBuilder>;
+// This uses the hashmap from this crate with the default hasher of the stdlib.
+type StdHashMap<K, V> = HashMap<K, V, RandomState>;
 
-#[bench]
-fn insert_erase_i32(b: &mut Bencher) {
-    b.iter(|| {
-        let mut m: HashMap<i32, i32> = new_map();
-        for i in 1..1001 {
-            m.insert(i, i);
-        }
-        black_box(&mut m);
-        for i in 1..1001 {
-            m.remove(&i);
-        }
-        black_box(m);
-    })
+// A random key iterator.
+#[derive(Clone, Copy)]
+struct RandomKeys {
+    remaining: usize,
+    state: usize,
 }
 
-#[bench]
-fn insert_erase_i64(b: &mut Bencher) {
-    b.iter(|| {
-        let mut m: HashMap<i64, i64> = new_map();
-        for i in 1..1001 {
-            m.insert(i, i);
-        }
-        black_box(&mut m);
-        for i in 1..1001 {
-            m.remove(&i);
+impl RandomKeys {
+    fn new(size: usize) -> Self {
+        RandomKeys {
+            remaining: size,
+            state: 1,
         }
-        black_box(m);
-    })
-}
-
-#[bench]
-fn lookup_i32(b: &mut Bencher) {
-    let mut m: HashMap<i32, i32> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
     }
 
-    b.iter(|| {
-        for i in 1..1001 {
-            black_box(m.get(&i));
+    // Produce a different set of random values.
+    fn new2(size: usize) -> Self {
+        RandomKeys {
+            remaining: size,
+            state: 2,
         }
-    })
+    }
 }
 
-#[bench]
-fn lookup_i64(b: &mut Bencher) {
-    let mut m: HashMap<i64, i64> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
+impl Iterator for RandomKeys {
+    type Item = usize;
+    fn next(&mut self) -> Option<usize> {
+        if self.remaining == 0 {
+            None
+        } else {
+            self.remaining -= 1;
+            // Multiply by some 32 bit prime.
+            self.state = self.state.wrapping_mul(3787392781);
+            // Mix in to the bottom bits which are constant mod powers of 2.
+            Some(self.state ^ (self.state >> 4))
+        }
     }
+}
 
-    b.iter(|| {
-        for i in 1..1001 {
-            black_box(m.get(&i));
+macro_rules! bench_insert {
+    ($name:ident, $maptype:ident, $keydist:expr) => {
+        #[bench]
+        fn $name(b: &mut Bencher) {
+            b.iter(|| {
+                let mut m = $maptype::default();
+                for i in $keydist {
+                    m.insert(i, i);
+                }
+                black_box(m);
+            })
         }
-    })
+    };
 }
 
-#[bench]
-fn lookup_fail_i32(b: &mut Bencher) {
-    let mut m: HashMap<i32, i32> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
-    }
-
-    b.iter(|| {
-        for i in 1001..2001 {
-            black_box(m.get(&i));
+bench_insert!(insert_fx_serial, FxHashMap, 0..SIZE);
+bench_insert!(insert_std_serial, StdHashMap, 0..SIZE);
+bench_insert!(
+    insert_fx_highbits,
+    FxHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_insert!(
+    insert_std_highbits,
+    StdHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_insert!(insert_fx_random, FxHashMap, RandomKeys::new(SIZE));
+bench_insert!(insert_std_random, StdHashMap, RandomKeys::new(SIZE));
+
+macro_rules! bench_insert_erase {
+    ($name:ident, $maptype:ident, $keydist:expr) => {
+        #[bench]
+        fn $name(b: &mut Bencher) {
+            b.iter(|| {
+                let mut m = $maptype::default();
+                for i in $keydist {
+                    m.insert(i, i);
+                }
+                black_box(&mut m);
+                for i in $keydist {
+                    m.remove(&i);
+                }
+                black_box(m);
+            })
         }
-    })
+    };
 }
 
-#[bench]
-fn lookup_fail_i64(b: &mut Bencher) {
-    let mut m: HashMap<i64, i64> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
-    }
-
-    b.iter(|| {
-        for i in 1001..2001 {
-            black_box(m.get(&i));
+bench_insert_erase!(insert_erase_fx_serial, FxHashMap, 0..SIZE);
+bench_insert_erase!(insert_erase_std_serial, StdHashMap, 0..SIZE);
+bench_insert_erase!(
+    insert_erase_fx_highbits,
+    FxHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_insert_erase!(
+    insert_erase_std_highbits,
+    StdHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_insert_erase!(insert_erase_fx_random, FxHashMap, RandomKeys::new(SIZE));
+bench_insert_erase!(insert_erase_std_random, StdHashMap, RandomKeys::new(SIZE));
+
+macro_rules! bench_lookup {
+    ($name:ident, $maptype:ident, $keydist:expr) => {
+        #[bench]
+        fn $name(b: &mut Bencher) {
+            let mut m = $maptype::default();
+            for i in $keydist {
+                m.insert(i, i);
+            }
+
+            b.iter(|| {
+                for i in $keydist {
+                    black_box(m.get(&i));
+                }
+            })
         }
-    })
+    };
 }
 
-#[bench]
-fn iter_i32(b: &mut Bencher) {
-    let mut m: HashMap<i32, i32> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
-    }
-
-    b.iter(|| {
-        for i in &m {
-            black_box(i);
+bench_lookup!(lookup_fx_serial, FxHashMap, 0..SIZE);
+bench_lookup!(lookup_std_serial, StdHashMap, 0..SIZE);
+bench_lookup!(
+    lookup_fx_highbits,
+    FxHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_lookup!(
+    lookup_std_highbits,
+    StdHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_lookup!(lookup_fx_random, FxHashMap, RandomKeys::new(SIZE));
+bench_lookup!(lookup_std_random, StdHashMap, RandomKeys::new(SIZE));
+
+macro_rules! bench_lookup_fail {
+    ($name:ident, $maptype:ident, $keydist:expr, $keydist2:expr) => {
+        #[bench]
+        fn $name(b: &mut Bencher) {
+            let mut m = $maptype::default();
+            for i in $keydist {
+                m.insert(i, i);
+            }
+
+            b.iter(|| {
+                for i in $keydist2 {
+                    black_box(m.get(&i));
+                }
+            })
         }
-    })
+    };
 }
 
-#[bench]
-fn iter_i64(b: &mut Bencher) {
-    let mut m: HashMap<i64, i64> = new_map();
-    for i in 1..1001 {
-        m.insert(i, i);
-    }
-
-    b.iter(|| {
-        for i in &m {
-            black_box(i);
+bench_lookup_fail!(lookup_fail_fx_serial, FxHashMap, 0..SIZE, SIZE..SIZE * 2);
+bench_lookup_fail!(lookup_fail_std_serial, StdHashMap, 0..SIZE, SIZE..SIZE * 2);
+bench_lookup_fail!(
+    lookup_fail_fx_highbits,
+    FxHashMap,
+    (0..SIZE).map(usize::swap_bytes),
+    (SIZE..SIZE * 2).map(usize::swap_bytes)
+);
+bench_lookup_fail!(
+    lookup_fail_std_highbits,
+    StdHashMap,
+    (0..SIZE).map(usize::swap_bytes),
+    (SIZE..SIZE * 2).map(usize::swap_bytes)
+);
+bench_lookup_fail!(
+    lookup_fail_fx_random,
+    FxHashMap,
+    RandomKeys::new(SIZE),
+    RandomKeys::new2(SIZE)
+);
+bench_lookup_fail!(
+    lookup_fail_std_random,
+    StdHashMap,
+    RandomKeys::new(SIZE),
+    RandomKeys::new2(SIZE)
+);
+
+macro_rules! bench_iter {
+    ($name:ident, $maptype:ident, $keydist:expr) => {
+        #[bench]
+        fn $name(b: &mut Bencher) {
+            let mut m = $maptype::default();
+            for i in $keydist {
+                m.insert(i, i);
+            }
+
+            b.iter(|| {
+                for i in &m {
+                    black_box(i);
+                }
+            })
         }
-    })
+    };
 }
+
+bench_iter!(iter_fx_serial, FxHashMap, 0..SIZE);
+bench_iter!(iter_std_serial, StdHashMap, 0..SIZE);
+bench_iter!(
+    iter_fx_highbits,
+    FxHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_iter!(
+    iter_std_highbits,
+    StdHashMap,
+    (0..SIZE).map(usize::swap_bytes)
+);
+bench_iter!(iter_fx_random, FxHashMap, RandomKeys::new(SIZE));
+bench_iter!(iter_std_random, StdHashMap, RandomKeys::new(SIZE));