From 16fc41cedc5f81a4ecba7bfaa9f408d41ca87f62 Mon Sep 17 00:00:00 2001
From: mat <git@matdoes.dev>
Date: Mon, 1 Jul 2024 22:36:40 +0000
Subject: [PATCH] Optimize SipHash by reordering compress instructions

---
 compiler/rustc_data_structures/src/sip128.rs | 11 ++++++-----
 library/core/src/hash/sip.rs                 | 11 ++++++-----
 2 files changed, 12 insertions(+), 10 deletions(-)

diff --git a/compiler/rustc_data_structures/src/sip128.rs b/compiler/rustc_data_structures/src/sip128.rs
index fed23df10dc..812ed410a94 100644
--- a/compiler/rustc_data_structures/src/sip128.rs
+++ b/compiler/rustc_data_structures/src/sip128.rs
@@ -70,18 +70,19 @@ macro_rules! compress {
     ($state:expr) => {{ compress!($state.v0, $state.v1, $state.v2, $state.v3) }};
     ($v0:expr, $v1:expr, $v2:expr, $v3:expr) => {{
         $v0 = $v0.wrapping_add($v1);
+        $v2 = $v2.wrapping_add($v3);
         $v1 = $v1.rotate_left(13);
         $v1 ^= $v0;
-        $v0 = $v0.rotate_left(32);
-        $v2 = $v2.wrapping_add($v3);
         $v3 = $v3.rotate_left(16);
         $v3 ^= $v2;
-        $v0 = $v0.wrapping_add($v3);
-        $v3 = $v3.rotate_left(21);
-        $v3 ^= $v0;
+        $v0 = $v0.rotate_left(32);
+
         $v2 = $v2.wrapping_add($v1);
+        $v0 = $v0.wrapping_add($v3);
         $v1 = $v1.rotate_left(17);
         $v1 ^= $v2;
+        $v3 = $v3.rotate_left(21);
+        $v3 ^= $v0;
         $v2 = $v2.rotate_left(32);
     }};
 }
diff --git a/library/core/src/hash/sip.rs b/library/core/src/hash/sip.rs
index 78a232faaf8..0d1ac64aa56 100644
--- a/library/core/src/hash/sip.rs
+++ b/library/core/src/hash/sip.rs
@@ -76,18 +76,19 @@ macro_rules! compress {
     ($state:expr) => {{ compress!($state.v0, $state.v1, $state.v2, $state.v3) }};
     ($v0:expr, $v1:expr, $v2:expr, $v3:expr) => {{
         $v0 = $v0.wrapping_add($v1);
+        $v2 = $v2.wrapping_add($v3);
         $v1 = $v1.rotate_left(13);
         $v1 ^= $v0;
-        $v0 = $v0.rotate_left(32);
-        $v2 = $v2.wrapping_add($v3);
         $v3 = $v3.rotate_left(16);
         $v3 ^= $v2;
-        $v0 = $v0.wrapping_add($v3);
-        $v3 = $v3.rotate_left(21);
-        $v3 ^= $v0;
+        $v0 = $v0.rotate_left(32);
+
         $v2 = $v2.wrapping_add($v1);
+        $v0 = $v0.wrapping_add($v3);
         $v1 = $v1.rotate_left(17);
         $v1 ^= $v2;
+        $v3 = $v3.rotate_left(21);
+        $v3 ^= $v0;
         $v2 = $v2.rotate_left(32);
     }};
 }