From: Chris Duncan <chris@zoso.dev>
Date: Tue, 14 Jan 2025 21:14:27 +0000 (-0800)
Subject: Replace more m scalar operations.
X-Git-Url: https://zoso.dev/?a=commitdiff_plain;h=be638e4e0eaccd582292e635185aac1a9ed7424f;p=nano-pow.git

Replace more m scalar operations.
---

diff --git a/src/shaders/compute.wgsl b/src/shaders/compute.wgsl
index a93a421..2a9eae9 100644
--- a/src/shaders/compute.wgsl
+++ b/src/shaders/compute.wgsl
@@ -1394,11 +1394,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v2 + m0;
-	o1 = v3 + m1;
-	o1 = o1 + select(0u, 1u, o0 < v2);
-	v2 = o0;
-	v3 = o1;
+	v_23.x = v2;
+	v_23.y = v3;
+	v_23 = v_23 + vec2(m0, m1) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m0 < v_23.x);
+	v2 = v_23.x;
+	v3 = v_23.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_23.x = v2;
@@ -1640,11 +1640,12 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v6 + m6;
-	o1 = v7 + m7;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m6, m7) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m6 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
+
 
 	// d = rotr64(d ^ a, 16)
 	v_67.x = v6;
@@ -1836,11 +1837,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v2 + m0;
-	o1 = v3 + m1;
-	o1 = o1 + select(0u, 1u, o0 < v2);
-	v2 = o0;
-	v3 = o1;
+	v_23.x = v2;
+	v_23.y = v3;
+	v_23 = v_23 + vec2(m0, m1) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m0 < v_23.x);
+	v2 = v_23.x;
+	v3 = v_23.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_23.x = v2;
@@ -1937,11 +1938,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v4 + m4;
-	o1 = v5 + m5;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m4 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_45.x = v4;
@@ -2329,11 +2330,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v4 + m2;
-	o1 = v5 + m3;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m2, m3) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m2 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_45.x = v4;
@@ -2430,11 +2431,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v6 + m8;
-	o1 = v7 + m9;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m8, m9) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m8 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_67.x = v6;
@@ -2630,11 +2631,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v2 + m2;
-	o1 = v3 + m3;
-	o1 = o1 + select(0u, 1u, o0 < v2);
-	v2 = o0;
-	v3 = o1;
+	v_23.x = v2;
+	v_23.y = v3;
+	v_23 = v_23 + vec2(m2, m3) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m2 < v_23.x);
+	v2 = v_23.x;
+	v3 = v_23.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_23.x = v2;
@@ -2879,11 +2880,8 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v_01.x + m4;
-	o1 = v_01.y + m5;
-	o1 = o1 + select(0u, 1u, o0 < v_01.x);
-	v_01.x = o0;
-	v_01.y = o1;
+	v_01 = v_01 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + m4 < v_01.x);
+
 
 
 
@@ -3125,11 +3123,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v4 + m0;
-	o1 = v5 + m1;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m0, m1) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m0 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_45.x = v4;
@@ -3477,11 +3475,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v4 + m4;
-	o1 = v5 + m5;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m4 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_45.x = v4;
@@ -3968,11 +3966,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v6 + m6;
-	o1 = v7 + m7;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m6, m7) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m6 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_67.x = v6;
@@ -4071,11 +4069,7 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v_01.x + m4;
-	o1 = v_01.y + m5;
-	o1 = o1 + select(0u, 1u, o0 < v_01.x);
-	v_01.x = o0;
-	v_01.y = o1;
+	v_01 = v_01 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + m4 < v_01.x);
 
 
 
@@ -4267,11 +4261,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v4 + m0;
-	o1 = v5 + m1;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m0, m1) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m0 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_45.x = v4;
@@ -4412,11 +4406,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v6 + m6;
-	o1 = v7 + m7;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m6, m7) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m6 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_67.x = v6;
@@ -4958,11 +4952,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v2 + m2;
-	o1 = v3 + m3;
-	o1 = o1 + select(0u, 1u, o0 < v2);
-	v2 = o0;
-	v3 = o1;
+	v_23.x = v2;
+	v_23.y = v3;
+	v_23 = v_23 + vec2(m2, m3) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m2 < v_23.x);
+	v2 = v_23.x;
+	v3 = v_23.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_23.x = v2;
@@ -5156,11 +5150,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v6 + m8;
-	o1 = v7 + m9;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m8, m9) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m8 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_67.x = v6;
@@ -5501,11 +5495,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v4 + m4;
-	o1 = v5 + m5;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m4 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_45.x = v4;
@@ -5891,11 +5885,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v4 + m2;
-	o1 = v5 + m3;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m2, m3) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m2 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_45.x = v4;
@@ -5948,11 +5942,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v6 + m6;
-	o1 = v7 + m7;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m6, m7) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m6 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_67.x = v6;
@@ -6881,11 +6875,7 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v_01.x + m4;
-	o1 = v_01.y + m5;
-	o1 = o1 + select(0u, 1u, o0 < v_01.x);
-	v_01.x = o0;
-	v_01.y = o1;
+	v_01 = v_01 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + m4 < v_01.x);
 
 
 
@@ -7037,11 +7027,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v4 + m2;
-	o1 = v5 + m3;
-	o1 = o1 + select(0u, 1u, o0 < v4);
-	v4 = o0;
-	v5 = o1;
+	v_45.x = v4;
+	v_45.y = v5;
+	v_45 = v_45 + vec2(m2, m3) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_45.x + m2 < v_45.x);
+	v4 = v_45.x;
+	v5 = v_45.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_45.x = v4;
@@ -7281,11 +7271,7 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v_01.x + m4;
-	o1 = v_01.y + m5;
-	o1 = o1 + select(0u, 1u, o0 < v_01.x);
-	v_01.x = o0;
-	v_01.y = o1;
+	v_01 = v_01 + vec2(m4, m5) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + m4 < v_01.x);
 
 
 
@@ -9313,11 +9299,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+0]]
-	o0 = v2 + m0;
-	o1 = v3 + m1;
-	o1 = o1 + select(0u, 1u, o0 < v2);
-	v2 = o0;
-	v3 = o1;
+	v_23.x = v2;
+	v_23.y = v3;
+	v_23 = v_23 + vec2(m0, m1) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + m0 < v_23.x);
+	v2 = v_23.x;
+	v3 = v_23.y;
 
 	// d = rotr64(d ^ a, 32)
 	v_23.x = v2;
@@ -9559,11 +9545,11 @@ fn main(@builtin(global_invocation_id) id: vec3<u32>) {
 
 
 	// a = a + m[sigma[r][2*i+1]]
-	o0 = v6 + m6;
-	o1 = v7 + m7;
-	o1 = o1 + select(0u, 1u, o0 < v6);
-	v6 = o0;
-	v7 = o1;
+	v_67.x = v6;
+	v_67.y = v7;
+	v_67 = v_67 + vec2(m6, m7) + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + m6 < v_67.x);
+	v6 = v_67.x;
+	v7 = v_67.y;
 
 	// d = rotr64(d ^ a, 16)
 	v_67.x = v6;