From 62e2fa4d9ea23039bac5c93a3f80598e52f7d86b Mon Sep 17 00:00:00 2001
From: Chris Duncan <chris@zoso.dev>
Date: Tue, 14 Jan 2025 13:43:35 -0800
Subject: [PATCH] Replace another 32-bit rotation.

---
 src/shaders/compute.wgsl | 132 +++++++++++++++++++++------------------
 1 file changed, 72 insertions(+), 60 deletions(-)

diff --git a/src/shaders/compute.wgsl b/src/shaders/compute.wgsl
index b89fe92..48922eb 100644
--- a/src/shaders/compute.wgsl
+++ b/src/shaders/compute.wgsl
@@ -164,11 +164,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	v_23 = v_23 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m4 < v_23.x);
 
 	// d = rotr64(d ^ a, 32)
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 	v_2627.x = v26;
@@ -673,11 +674,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -1460,11 +1462,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -2255,11 +2258,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -3044,11 +3048,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -3831,11 +3836,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -4621,11 +4627,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -5406,11 +5413,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -6195,11 +6203,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -6986,11 +6995,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -7781,11 +7791,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
@@ -8564,11 +8575,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 	// d = rotr64(d ^ a, 32)
 
 
-	v_2627.x = v26;
-	v_2627.y = v27;
-	v_2627 = v_2627 ^ v_23;
-	v26 = v_2627.y;
-	v27 = v_2627.x;
+	v26 = v_2627.x;
+	v27 = v_2627.y;
+	xor = v_2627 ^ v_23;
+	v_2627 = vec2(xor.y, xor.x);
+	v26 = v_2627.x;
+	v27 = v_2627.y;
 
 	// c = c + d
 
-- 
2.34.1