Move mutexes into shared data structure

Author: maneatingape

src/year2016/day05.rs

@@ -13,6 +13,7 @@ struct Shared {
     prefix: String,
     done: AtomicBool,
     counter: AtomicU32,
+    mutex: Mutex<Exclusive>,
 }
 
 struct Exclusive {
@@ -25,24 +26,24 @@ pub fn parse(input: &str) -> Vec<u32> {
         prefix: input.trim().to_owned(),
         done: AtomicBool::new(false),
         counter: AtomicU32::new(1000),
+        mutex: Mutex::new(Exclusive { found: vec![], mask: 0 }),
     };
-    let mutex = Mutex::new(Exclusive { found: vec![], mask: 0 });
 
     // Handle the first 999 numbers specially as the number of digits varies.
     for n in 1..1000 {
         let (mut buffer, size) = format_string(&shared.prefix, n);
-        check_hash(&mut buffer, size, n, &shared, &mutex);
+        check_hash(&mut buffer, size, n, &shared);
     }
 
     // Use as many cores as possible to parallelize the remaining search.
     spawn(|| {
         #[cfg(not(feature = "simd"))]
-        worker(&shared, &mutex);
+        worker(&shared);
         #[cfg(feature = "simd")]
-        simd::worker(&shared, &mutex);
+        simd::worker(&shared);
     });
 
-    let mut found = mutex.into_inner().unwrap().found;
+    let mut found = shared.mutex.into_inner().unwrap().found;
     found.sort_unstable();
     found.iter().map(|&(_, n)| n).collect()
 }
@@ -79,11 +80,11 @@ fn format_string(prefix: &str, n: u32) -> ([u8; 64], usize) {
     (buffer, size)
 }
 
-fn check_hash(buffer: &mut [u8], size: usize, n: u32, shared: &Shared, mutex: &Mutex<Exclusive>) {
+fn check_hash(buffer: &mut [u8], size: usize, n: u32, shared: &Shared) {
     let (result, ..) = hash(buffer, size);
 
     if result & 0xfffff000 == 0 {
-        let mut exclusive = mutex.lock().unwrap();
+        let mut exclusive = shared.mutex.lock().unwrap();
 
         exclusive.found.push((n, result));
         exclusive.mask |= 1 << (result >> 8);
@@ -95,7 +96,7 @@ fn check_hash(buffer: &mut [u8], size: usize, n: u32, shared: &Shared, mutex: &M
 }
 
 #[cfg(not(feature = "simd"))]
-fn worker(shared: &Shared, mutex: &Mutex<Exclusive>) {
+fn worker(shared: &Shared) {
     while !shared.done.load(Ordering::Relaxed) {
         let offset = shared.counter.fetch_add(1000, Ordering::Relaxed);
         let (mut buffer, size) = format_string(&shared.prefix, offset);
@@ -106,7 +107,7 @@ fn worker(shared: &Shared, mutex: &Mutex<Exclusive>) {
             buffer[size - 2] = b'0' + ((n / 10) % 10) as u8;
             buffer[size - 1] = b'0' + (n % 10) as u8;
 
-            check_hash(&mut buffer, size, offset + n, shared, mutex);
+            check_hash(&mut buffer, size, offset + n, shared);
         }
     }
 }
@@ -124,7 +125,6 @@ mod simd {
         start: u32,
         offset: u32,
         shared: &Shared,
-        mutex: &Mutex<Exclusive>,
     ) where
         LaneCount<N>: SupportedLaneCount,
     {
@@ -140,7 +140,7 @@ mod simd {
 
         for i in 0..N {
             if result[i] & 0xfffff000 == 0 {
-                let mut exclusive = mutex.lock().unwrap();
+                let mut exclusive = shared.mutex.lock().unwrap();
 
                 exclusive.found.push((start + offset + i as u32, result[i]));
                 exclusive.mask |= 1 << (result[i] >> 8);
@@ -152,17 +152,17 @@ mod simd {
         }
     }
 
-    pub(super) fn worker(shared: &Shared, mutex: &Mutex<Exclusive>) {
+    pub(super) fn worker(shared: &Shared) {
         while !shared.done.load(Ordering::Relaxed) {
             let start = shared.counter.fetch_add(1000, Ordering::Relaxed);
             let (prefix, size) = format_string(&shared.prefix, start);
             let mut buffers = [prefix; 32];
 
             for offset in (0..992).step_by(32) {
-                check_hash_simd::<32>(&mut buffers, size, start, offset, shared, mutex);
+                check_hash_simd::<32>(&mut buffers, size, start, offset, shared);
             }
 
-            check_hash_simd::<8>(&mut buffers, size, start, 992, shared, mutex);
+            check_hash_simd::<8>(&mut buffers, size, start, 992, shared);
         }
     }
 }

src/year2016/day14.rs

@@ -11,8 +11,10 @@ use std::sync::Mutex;
 /// Atomics can be safely shared between threads.
 struct Shared<'a> {
     input: &'a str,
+    part_two: bool,
     done: AtomicBool,
     counter: AtomicI32,
+    mutex: Mutex<Exclusive>,
 }
 
 /// Regular data structures need to be protected by a mutex.
@@ -38,20 +40,25 @@ pub fn part2(input: &str) -> i32 {
 
 /// Find the first 64 keys that sastify the rules.
 fn generate_pad(input: &str, part_two: bool) -> i32 {
-    let shared = Shared { input, done: AtomicBool::new(false), counter: AtomicI32::new(0) };
     let exclusive =
         Exclusive { threes: BTreeMap::new(), fives: BTreeMap::new(), found: BTreeSet::new() };
-    let mutex = Mutex::new(exclusive);
+    let shared = Shared {
+        input,
+        part_two,
+        done: AtomicBool::new(false),
+        counter: AtomicI32::new(0),
+        mutex: Mutex::new(exclusive),
+    };
 
     // Use as many cores as possible to parallelize the search.
-    spawn(|| worker(&shared, &mutex, part_two));
+    spawn(|| worker(&shared));
 
-    let exclusive = mutex.into_inner().unwrap();
+    let exclusive = shared.mutex.into_inner().unwrap();
     *exclusive.found.iter().nth(63).unwrap()
 }
 
 #[cfg(not(feature = "simd"))]
-fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
+fn worker(shared: &Shared<'_>) {
     while !shared.done.load(Ordering::Relaxed) {
         // Get the next key to check.
         let n = shared.counter.fetch_add(1, Ordering::Relaxed);
@@ -60,7 +67,7 @@ fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
         let (mut buffer, size) = format_string(shared.input, n);
         let mut result = hash(&mut buffer, size);
 
-        if part_two {
+        if shared.part_two {
             for _ in 0..2016 {
                 buffer[0..8].copy_from_slice(&to_ascii(result.0));
                 buffer[8..16].copy_from_slice(&to_ascii(result.1));
@@ -70,14 +77,14 @@ fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
             }
         }
 
-        check(shared, mutex, n, result);
+        check(shared, n, result);
     }
 }
 
 /// Use SIMD to compute hashes in parallel in blocks of 32.
 #[cfg(feature = "simd")]
 #[allow(clippy::needless_range_loop)]
-fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
+fn worker(shared: &Shared<'_>) {
     let mut result = ([0; 32], [0; 32], [0; 32], [0; 32]);
     let mut buffers = [[0; 64]; 32];
 
@@ -96,7 +103,7 @@ fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
             result.3[i] = d;
         }
 
-        if part_two {
+        if shared.part_two {
             for _ in 0..2016 {
                 for i in 0..32 {
                     buffers[i][0..8].copy_from_slice(&to_ascii(result.0[i]));
@@ -110,13 +117,13 @@ fn worker(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, part_two: bool) {
 
         for i in 0..32 {
             let hash = (result.0[i], result.1[i], result.2[i], result.3[i]);
-            check(shared, mutex, start + i as i32, hash);
+            check(shared, start + i as i32, hash);
         }
     }
 }
 
 /// Check for sequences of 3 or 5 consecutive matching digits.
-fn check(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, n: i32, hash: (u32, u32, u32, u32)) {
+fn check(shared: &Shared<'_>, n: i32, hash: (u32, u32, u32, u32)) {
     let (a, b, c, d) = hash;
 
     let mut prev = u32::MAX;
@@ -147,7 +154,7 @@ fn check(shared: &Shared<'_>, mutex: &Mutex<Exclusive>, n: i32, hash: (u32, u32,
     }
 
     if three != 0 || five != 0 {
-        let mut exclusive = mutex.lock().unwrap();
+        let mut exclusive = shared.mutex.lock().unwrap();
         let mut candidates = Vec::new();
 
         // Compare against all 5 digit sequences.

src/year2017/day14.rs

@@ -13,6 +13,7 @@ use std::sync::Mutex;
 pub struct Shared {
     prefix: String,
     counter: AtomicUsize,
+    mutex: Mutex<Exclusive>,
 }
 
 /// Regular data structures need to be protected by a mutex.
@@ -22,14 +23,16 @@ struct Exclusive {
 
 /// Parallelize the hashing as each row is independent.
 pub fn parse(input: &str) -> Vec<u8> {
-    let shared = Shared { prefix: input.trim().to_owned(), counter: AtomicUsize::new(0) };
-    let exclusive = Exclusive { grid: vec![0; 0x4000] };
-    let mutex = Mutex::new(exclusive);
+    let shared = Shared {
+        prefix: input.trim().to_owned(),
+        counter: AtomicUsize::new(0),
+        mutex: Mutex::new(Exclusive { grid: vec![0; 0x4000] }),
+    };
 
     // Use as many cores as possible to parallelize the hashing.
-    spawn(|| worker(&shared, &mutex));
+    spawn(|| worker(&shared));
 
-    mutex.into_inner().unwrap().grid
+    shared.mutex.into_inner().unwrap().grid
 }
 
 pub fn part1(input: &[u8]) -> u32 {
@@ -53,7 +56,7 @@ pub fn part2(input: &[u8]) -> u32 {
 
 /// Each worker thread chooses the next available index then computes the hash and patches the
 /// final vec with the result.
-fn worker(shared: &Shared, mutex: &Mutex<Exclusive>) {
+fn worker(shared: &Shared) {
     loop {
         let index = shared.counter.fetch_add(1, Ordering::Relaxed);
         if index >= 128 {
@@ -64,7 +67,7 @@ fn worker(shared: &Shared, mutex: &Mutex<Exclusive>) {
         let start = index * 128;
         let end = start + 128;
 
-        let mut exclusive = mutex.lock().unwrap();
+        let mut exclusive = shared.mutex.lock().unwrap();
         exclusive.grid[start..end].copy_from_slice(&row);
     }
 }

src/year2018/day11.rs

@@ -16,6 +16,11 @@ pub struct Result {
     power: i32,
 }
 
+struct Shared {
+    sat: Vec<i32>,
+    mutex: Mutex<Vec<Result>>,
+}
+
 pub fn parse(input: &str) -> Vec<Result> {
     let grid_serial_number: i32 = input.signed();
 
@@ -45,9 +50,9 @@ pub fn parse(input: &str) -> Vec<Result> {
     // * 2, 6, 10, ..
     // * 3, 7, 11, ..
     // * 4, 8, 12, ..
-    let mutex = Mutex::new(Vec::new());
-    spawn_batches((1..301).collect(), |batch| worker(batch, &sat, &mutex));
-    mutex.into_inner().unwrap()
+    let shared = Shared { sat, mutex: Mutex::new(Vec::new()) };
+    spawn_batches((1..301).collect(), |batch| worker(&shared, batch));
+    shared.mutex.into_inner().unwrap()
 }
 
 pub fn part1(input: &[Result]) -> String {
@@ -60,16 +65,16 @@ pub fn part2(input: &[Result]) -> String {
     format!("{x},{y},{size}")
 }
 
-fn worker(batch: Vec<usize>, sat: &[i32], mutex: &Mutex<Vec<Result>>) {
+fn worker(shared: &Shared, batch: Vec<usize>) {
     let result: Vec<_> = batch
         .into_iter()
         .map(|size| {
-            let (power, x, y) = square(sat, size);
+            let (power, x, y) = square(&shared.sat, size);
             Result { x, y, size, power }
         })
         .collect();
 
-    mutex.lock().unwrap().extend(result);
+    shared.mutex.lock().unwrap().extend(result);
 }
 
 /// Find the (x,y) coordinates and max power for a square of the specified size.

src/year2021/day18.rs

@@ -28,7 +28,7 @@
 //! `2i + 1`, right child at index `2i + 2` and parent at index `i / 2`. As leaf nodes are
 //! always greater than or equal to zero, `-1` is used as a special sentinel value for non-leaf nodes.
 use crate::util::thread::*;
-use std::sync::Mutex;
+use std::sync::atomic::{AtomicI32, Ordering};
 
 type Snailfish = [i32; 63];
 
@@ -85,21 +85,20 @@ pub fn part2(input: &[Snailfish]) -> i32 {
 
     // Use as many cores as possible to parallelize the calculation,
     // breaking the work into roughly equally size batches.
-    let mutex = Mutex::new(0);
-    spawn_batches(pairs, |batch| worker(&batch, &mutex));
-    mutex.into_inner().unwrap()
+    let shared = AtomicI32::new(0);
+    spawn_batches(pairs, |batch| worker(&shared, &batch));
+    shared.load(Ordering::Relaxed)
 }
 
 /// Pair addition is independent so we can parallelize across multiple threads.
-fn worker(batch: &[(&Snailfish, &Snailfish)], mutex: &Mutex<i32>) {
+fn worker(shared: &AtomicI32, batch: &[(&Snailfish, &Snailfish)]) {
     let mut partial = 0;
 
     for (a, b) in batch {
         partial = partial.max(magnitude(&mut add(a, b)));
     }
 
-    let mut result = mutex.lock().unwrap();
-    *result = result.max(partial);
+    shared.fetch_max(partial, Ordering::Relaxed);
 }
 
 /// Add two snailfish numbers.

src/year2022/day11.rs

@@ -62,6 +62,11 @@ pub enum Operation {
 type Pair = (usize, u64);
 type Business = [u64; 8];
 
+struct Shared<'a> {
+    monkeys: &'a [Monkey],
+    mutex: Mutex<Exclusive>,
+}
+
 struct Exclusive {
     pairs: Vec<Pair>,
     business: Business,
@@ -125,30 +130,28 @@ fn sequential(monkeys: &[Monkey], pairs: Vec<Pair>) -> Business {
 
 /// Play 10,000 rounds adjusting the worry level modulo the product of all the monkey's test values.
 fn parallel(monkeys: &[Monkey], pairs: Vec<Pair>) -> Business {
-    let business = [0; 8];
-    let exclusive = Exclusive { pairs, business };
-    let mutex = Mutex::new(exclusive);
+    let shared = Shared { monkeys, mutex: Mutex::new(Exclusive { pairs, business: [0; 8] }) };
 
     // Use as many cores as possible to parallelize the calculation.
-    spawn(|| worker(monkeys, &mutex));
+    spawn(|| worker(&shared));
 
-    mutex.into_inner().unwrap().business
+    shared.mutex.into_inner().unwrap().business
 }
 
 /// Multiple worker functions are executed in parallel, one per thread.
-fn worker(monkeys: &[Monkey], mutex: &Mutex<Exclusive>) {
-    let product: u64 = monkeys.iter().map(|m| m.test).product();
+fn worker(shared: &Shared<'_>) {
+    let product: u64 = shared.monkeys.iter().map(|m| m.test).product();
 
     loop {
         // Take an item from the queue until empty, using the mutex to allow access
         // to a single thread at a time.
-        let Some(pair) = mutex.lock().unwrap().pairs.pop() else {
+        let Some(pair) = shared.mutex.lock().unwrap().pairs.pop() else {
             break;
         };
 
-        let extra = play(monkeys, 10000, |x| x % product, pair);
+        let extra = play(shared.monkeys, 10000, |x| x % product, pair);
 
-        let mut exclusive = mutex.lock().unwrap();
+        let mut exclusive = shared.mutex.lock().unwrap();
         exclusive.business.iter_mut().enumerate().for_each(|(i, b)| *b += extra[i]);
     }
 }