mirror of
https://github.com/encounter/dawn-cmake.git
synced 2025-12-14 07:36:15 +00:00
tint/transform: Inline HLSL uniform / storage buffers
Change the DecomposeMemoryAccess to behave more like the DirectVariableAccess transform, in that it'll inline the access of buffer variable into the load / store helper functions, instead of passing the array down. This avoids large array copies observed with FXC, which can have *severe* performance costs. Fixed: tint:1819 Change-Id: I52eb3f908813f72ab9da446743e24a2637158309 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/121460 Kokoro: Kokoro <noreply+kokoro@google.com> Auto-Submit: Ben Clayton <bclayton@google.com> Reviewed-by: James Price <jrprice@google.com> Commit-Queue: James Price <jrprice@google.com>
This commit is contained in:
committed by
Dawn LUCI CQ
parent
7052cb57ff
commit
1a1b5278d5
@@ -8,20 +8,20 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -31,8 +31,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x2 l_a_i = tint_symbol_1(a, (16u * uint(p_a_i_save)));
|
||||
const float2x2 l_a[4] = a_load(0u);
|
||||
const float2x2 l_a_i = a_load_1((16u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((16u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const float2 l_a_i_i = asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy));
|
||||
|
||||
@@ -8,20 +8,20 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -31,8 +31,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x2 l_a_i = tint_symbol_1(a, (16u * uint(p_a_i_save)));
|
||||
const float2x2 l_a[4] = a_load(0u);
|
||||
const float2x2 l_a_i = a_load_1((16u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((16u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const float2 l_a_i_i = asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy));
|
||||
|
||||
@@ -2,20 +2,20 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[4];
|
||||
};
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = a_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x2 l_a_i = tint_symbol_1(a, 32u);
|
||||
const float2x2 l_a[4] = a_load(0u);
|
||||
const float2x2 l_a_i = a_load_1(32u);
|
||||
const float2 l_a_i_i = asfloat(a[2].zw);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,20 +2,20 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[4];
|
||||
};
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = a_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x2 l_a_i = tint_symbol_1(a, 32u);
|
||||
const float2x2 l_a[4] = a_load(0u);
|
||||
const float2x2 l_a_i = a_load_1(32u);
|
||||
const float2 l_a_i_i = asfloat(a[2].zw);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,17 +2,17 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[4];
|
||||
};
|
||||
|
||||
float2x2 tint_symbol(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x2 t = transpose(tint_symbol(u, 32u));
|
||||
const float2x2 t = transpose(u_load(32u));
|
||||
const float l = length(asfloat(u[0].zw).yx);
|
||||
const float a = abs(asfloat(u[0].zw).yx.x);
|
||||
return;
|
||||
|
||||
@@ -2,17 +2,17 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[4];
|
||||
};
|
||||
|
||||
float2x2 tint_symbol(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x2 t = transpose(tint_symbol(u, 32u));
|
||||
const float2x2 t = transpose(u_load(32u));
|
||||
const float l = length(asfloat(u[0].zw).yx);
|
||||
const float a = abs(asfloat(u[0].zw).yx.x);
|
||||
return;
|
||||
|
||||
@@ -14,20 +14,20 @@ void c(float2 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -35,8 +35,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 16u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(16u));
|
||||
c(asfloat(u[1].xy).yx);
|
||||
d(asfloat(u[1].xy).yx.x);
|
||||
return;
|
||||
|
||||
@@ -14,20 +14,20 @@ void c(float2 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -35,8 +35,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 16u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(16u));
|
||||
c(asfloat(u[1].xy).yx);
|
||||
d(asfloat(u[1].xy).yx.x);
|
||||
return;
|
||||
|
||||
@@ -3,20 +3,20 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x2 p[4] = (float2x2[4])0;
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -24,8 +24,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 32u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(32u);
|
||||
p[1][0] = asfloat(u[0].zw).yx;
|
||||
p[1][0].x = asfloat(u[0].z);
|
||||
return;
|
||||
|
||||
@@ -3,20 +3,20 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x2 p[4] = (float2x2[4])0;
|
||||
|
||||
float2x2 tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -24,8 +24,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 32u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(32u);
|
||||
p[1][0] = asfloat(u[0].zw).yx;
|
||||
p[1][0].x = asfloat(u[0].z);
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x2 value) {
|
||||
buffer.Store2((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store2((offset + 8u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x2 value) {
|
||||
s.Store2((offset + 0u), asuint(value[0u]));
|
||||
s.Store2((offset + 8u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x2 value[4]) {
|
||||
void s_store(uint offset, float2x2 value[4]) {
|
||||
float2x2 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 16u)), array_1[i]);
|
||||
s_store_1((offset + (i * 16u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_4(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 16u, tint_symbol_4(u, 32u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(16u, u_load_1(32u));
|
||||
s.Store2(16u, asuint(asfloat(u[0].zw).yx));
|
||||
s.Store(16u, asuint(asfloat(u[0].z)));
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x2 value) {
|
||||
buffer.Store2((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store2((offset + 8u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x2 value) {
|
||||
s.Store2((offset + 0u), asuint(value[0u]));
|
||||
s.Store2((offset + 8u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x2 value[4]) {
|
||||
void s_store(uint offset, float2x2 value[4]) {
|
||||
float2x2 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 16u)), array_1[i]);
|
||||
s_store_1((offset + (i * 16u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x2 tint_symbol_4(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 16u, tint_symbol_4(u, 32u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(16u, u_load_1(32u));
|
||||
s.Store2(16u, asuint(asfloat(u[0].zw).yx));
|
||||
s.Store(16u, asuint(asfloat(u[0].z)));
|
||||
return;
|
||||
|
||||
@@ -7,20 +7,20 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x2 tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 32u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(32u);
|
||||
w[1][0] = asfloat(u[0].zw).yx;
|
||||
w[1][0].x = asfloat(u[0].z);
|
||||
}
|
||||
|
||||
@@ -7,20 +7,20 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x2 tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
float2x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
return float2x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)));
|
||||
}
|
||||
|
||||
typedef float2x2 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[4], uint offset) {
|
||||
typedef float2x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x2 arr[4] = (float2x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 32u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(32u);
|
||||
w[1][0] = asfloat(u[0].zw).yx;
|
||||
w[1][0].x = asfloat(u[0].z);
|
||||
}
|
||||
|
||||
@@ -8,26 +8,26 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -37,8 +37,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const matrix<float16_t, 2, 3> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 2, 3> l_a_i = tint_symbol_1(a, (16u * uint(p_a_i_save)));
|
||||
const matrix<float16_t, 2, 3> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 2, 3> l_a_i = a_load_1((16u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((16u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_5 = a[scalar_offset_2 / 4];
|
||||
uint2 ubo_load_4 = ((scalar_offset_2 & 2) ? ubo_load_5.zw : ubo_load_5.xy);
|
||||
|
||||
@@ -2,26 +2,26 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = a_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 2, 3> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 2, 3> l_a_i = tint_symbol_1(a, 32u);
|
||||
const matrix<float16_t, 2, 3> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 2, 3> l_a_i = a_load_1(32u);
|
||||
uint2 ubo_load_4 = a[2].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -2,14 +2,14 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
@@ -18,7 +18,7 @@ matrix<float16_t, 2, 3> tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 3, 2> t = transpose(tint_symbol(u, 32u));
|
||||
const matrix<float16_t, 3, 2> t = transpose(u_load(32u));
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -14,26 +14,26 @@ void c(vector<float16_t, 3> v) {
|
||||
void d(float16_t f_1) {
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -41,8 +41,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 16u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(16u));
|
||||
uint2 ubo_load_4 = u[1].xy;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -3,26 +3,26 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static matrix<float16_t, 2, 3> p[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 32u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(32u);
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -3,40 +3,40 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 2, 3> value) {
|
||||
buffer.Store<vector<float16_t, 3> >((offset + 0u), value[0u]);
|
||||
buffer.Store<vector<float16_t, 3> >((offset + 8u), value[1u]);
|
||||
void s_store_1(uint offset, matrix<float16_t, 2, 3> value) {
|
||||
s.Store<vector<float16_t, 3> >((offset + 0u), value[0u]);
|
||||
s.Store<vector<float16_t, 3> >((offset + 8u), value[1u]);
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 2, 3> value[4]) {
|
||||
void s_store(uint offset, matrix<float16_t, 2, 3> value[4]) {
|
||||
matrix<float16_t, 2, 3> array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 16u)), array_1[i]);
|
||||
s_store_1((offset + (i * 16u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_4(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -44,8 +44,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 16u, tint_symbol_4(u, 32u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(16u, u_load_1(32u));
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -7,26 +7,26 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 3> tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 3> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16);
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16);
|
||||
return matrix<float16_t, 2, 3>(vector<float16_t, 3>(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector<float16_t, 3>(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 3> tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 3> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 3> arr[4] = (matrix<float16_t, 2, 3>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -40,8 +40,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 32u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(32u);
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16);
|
||||
|
||||
@@ -8,18 +8,18 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x3 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float2x3 l_a[4] = a_load(0u);
|
||||
const float2x3 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((32u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float3 l_a_i_i = asfloat(a[scalar_offset_2 / 4].xyz);
|
||||
return;
|
||||
|
||||
@@ -8,18 +8,18 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x3 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float2x3 l_a[4] = a_load(0u);
|
||||
const float2x3 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((32u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float3 l_a_i_i = asfloat(a[scalar_offset_2 / 4].xyz);
|
||||
return;
|
||||
|
||||
@@ -2,18 +2,18 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -21,8 +21,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x3 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float2x3 l_a[4] = a_load(0u);
|
||||
const float2x3 l_a_i = a_load_1(64u);
|
||||
const float3 l_a_i_i = asfloat(a[5].xyz);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,18 +2,18 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -21,8 +21,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x3 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float2x3 l_a[4] = a_load(0u);
|
||||
const float2x3 l_a_i = a_load_1(64u);
|
||||
const float3 l_a_i_i = asfloat(a[5].xyz);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,15 +2,15 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float2x3 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x2 t = transpose(tint_symbol(u, 64u));
|
||||
const float3x2 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[1].xyz).zxy);
|
||||
const float a = abs(asfloat(u[1].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -2,15 +2,15 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float2x3 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x2 t = transpose(tint_symbol(u, 64u));
|
||||
const float3x2 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[1].xyz).zxy);
|
||||
const float a = abs(asfloat(u[1].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -14,18 +14,18 @@ void c(float3 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2].xyz).zxy);
|
||||
d(asfloat(u[2].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -14,18 +14,18 @@ void c(float3 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2].xyz).zxy);
|
||||
d(asfloat(u[2].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -3,18 +3,18 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x3 p[4] = (float2x3[4])0;
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 64u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(64u);
|
||||
p[1][0] = asfloat(u[1].xyz).zxy;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,18 +3,18 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x3 p[4] = (float2x3[4])0;
|
||||
|
||||
float2x3 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 64u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(64u);
|
||||
p[1][0] = asfloat(u[1].xyz).zxy;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,32 +3,32 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x3 value) {
|
||||
buffer.Store3((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store3((offset + 16u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x3 value) {
|
||||
s.Store3((offset + 0u), asuint(value[0u]));
|
||||
s.Store3((offset + 16u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x3 value[4]) {
|
||||
void s_store(uint offset, float2x3 value[4]) {
|
||||
float2x3 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 32u)), array_1[i]);
|
||||
s_store_1((offset + (i * 32u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_4(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -36,8 +36,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 32u, tint_symbol_4(u, 64u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(32u, u_load_1(64u));
|
||||
s.Store3(32u, asuint(asfloat(u[1].xyz).zxy));
|
||||
s.Store(32u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -3,32 +3,32 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x3 value) {
|
||||
buffer.Store3((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store3((offset + 16u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x3 value) {
|
||||
s.Store3((offset + 0u), asuint(value[0u]));
|
||||
s.Store3((offset + 16u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x3 value[4]) {
|
||||
void s_store(uint offset, float2x3 value[4]) {
|
||||
float2x3 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 32u)), array_1[i]);
|
||||
s_store_1((offset + (i * 32u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x3 tint_symbol_4(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -36,8 +36,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 32u, tint_symbol_4(u, 64u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(32u, u_load_1(64u));
|
||||
s.Store3(32u, asuint(asfloat(u[1].xyz).zxy));
|
||||
s.Store(32u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -7,18 +7,18 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x3 tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -32,8 +32,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 64u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(64u);
|
||||
w[1][0] = asfloat(u[1].xyz).zxy;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -7,18 +7,18 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x3 tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
float2x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz));
|
||||
return float2x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float2x3 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[8], uint offset) {
|
||||
typedef float2x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x3 arr[4] = (float2x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -32,8 +32,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 64u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(64u);
|
||||
w[1][0] = asfloat(u[1].xyz).zxy;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -8,26 +8,26 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -37,8 +37,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const matrix<float16_t, 2, 4> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 2, 4> l_a_i = tint_symbol_1(a, (16u * uint(p_a_i_save)));
|
||||
const matrix<float16_t, 2, 4> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 2, 4> l_a_i = a_load_1((16u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((16u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_5 = a[scalar_offset_2 / 4];
|
||||
uint2 ubo_load_4 = ((scalar_offset_2 & 2) ? ubo_load_5.zw : ubo_load_5.xy);
|
||||
|
||||
@@ -2,26 +2,26 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = a_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 2, 4> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 2, 4> l_a_i = tint_symbol_1(a, 32u);
|
||||
const matrix<float16_t, 2, 4> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 2, 4> l_a_i = a_load_1(32u);
|
||||
uint2 ubo_load_4 = a[2].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -2,14 +2,14 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
@@ -18,7 +18,7 @@ matrix<float16_t, 2, 4> tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 4, 2> t = transpose(tint_symbol(u, 32u));
|
||||
const matrix<float16_t, 4, 2> t = transpose(u_load(32u));
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -14,26 +14,26 @@ void c(vector<float16_t, 4> v) {
|
||||
void d(float16_t f_1) {
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -41,8 +41,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 16u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(16u));
|
||||
uint2 ubo_load_4 = u[1].xy;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -3,26 +3,26 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static matrix<float16_t, 2, 4> p[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 32u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(32u);
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -3,40 +3,40 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 2, 4> value) {
|
||||
buffer.Store<vector<float16_t, 4> >((offset + 0u), value[0u]);
|
||||
buffer.Store<vector<float16_t, 4> >((offset + 8u), value[1u]);
|
||||
void s_store_1(uint offset, matrix<float16_t, 2, 4> value) {
|
||||
s.Store<vector<float16_t, 4> >((offset + 0u), value[0u]);
|
||||
s.Store<vector<float16_t, 4> >((offset + 8u), value[1u]);
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 2, 4> value[4]) {
|
||||
void s_store(uint offset, matrix<float16_t, 2, 4> value[4]) {
|
||||
matrix<float16_t, 2, 4> array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 16u)), array_1[i]);
|
||||
s_store_1((offset + (i * 16u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_4(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -44,8 +44,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 16u, tint_symbol_4(u, 32u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(16u, u_load_1(32u));
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -7,26 +7,26 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
matrix<float16_t, 2, 4> tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 2, 4> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset / 4];
|
||||
uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy);
|
||||
vector<float16_t, 2> ubo_load_xz = vector<float16_t, 2>(f16tof32(ubo_load & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_yw = vector<float16_t, 2>(f16tof32(ubo_load >> 16));
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_1 / 4];
|
||||
uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy);
|
||||
vector<float16_t, 2> ubo_load_2_xz = vector<float16_t, 2>(f16tof32(ubo_load_2 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_2_yw = vector<float16_t, 2>(f16tof32(ubo_load_2 >> 16));
|
||||
return matrix<float16_t, 2, 4>(vector<float16_t, 4>(ubo_load_xz[0], ubo_load_yw[0], ubo_load_xz[1], ubo_load_yw[1]), vector<float16_t, 4>(ubo_load_2_xz[0], ubo_load_2_yw[0], ubo_load_2_xz[1], ubo_load_2_yw[1]));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 2, 4> tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 2, 4> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 2, 4> arr[4] = (matrix<float16_t, 2, 4>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -40,8 +40,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 32u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(32u);
|
||||
uint2 ubo_load_4 = u[0].zw;
|
||||
vector<float16_t, 2> ubo_load_4_xz = vector<float16_t, 2>(f16tof32(ubo_load_4 & 0xFFFF));
|
||||
vector<float16_t, 2> ubo_load_4_yw = vector<float16_t, 2>(f16tof32(ubo_load_4 >> 16));
|
||||
|
||||
@@ -8,18 +8,18 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x4 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float2x4 l_a[4] = a_load(0u);
|
||||
const float2x4 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((32u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float4 l_a_i_i = asfloat(a[scalar_offset_2 / 4]);
|
||||
return;
|
||||
|
||||
@@ -8,18 +8,18 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -29,8 +29,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float2x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x4 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float2x4 l_a[4] = a_load(0u);
|
||||
const float2x4 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_2 = (((32u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float4 l_a_i_i = asfloat(a[scalar_offset_2 / 4]);
|
||||
return;
|
||||
|
||||
@@ -2,18 +2,18 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -21,8 +21,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x4 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float2x4 l_a[4] = a_load(0u);
|
||||
const float2x4 l_a_i = a_load_1(64u);
|
||||
const float4 l_a_i_i = asfloat(a[5]);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,18 +2,18 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -21,8 +21,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float2x4 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float2x4 l_a[4] = a_load(0u);
|
||||
const float2x4 l_a_i = a_load_1(64u);
|
||||
const float4 l_a_i_i = asfloat(a[5]);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,15 +2,15 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float2x4 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x2 t = transpose(tint_symbol(u, 64u));
|
||||
const float4x2 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[1]).ywxz);
|
||||
const float a = abs(asfloat(u[1]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -2,15 +2,15 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float2x4 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x2 t = transpose(tint_symbol(u, 64u));
|
||||
const float4x2 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[1]).ywxz);
|
||||
const float a = abs(asfloat(u[1]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -14,18 +14,18 @@ void c(float4 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2]).ywxz);
|
||||
d(asfloat(u[2]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -14,18 +14,18 @@ void c(float4 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2]).ywxz);
|
||||
d(asfloat(u[2]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -3,18 +3,18 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x4 p[4] = (float2x4[4])0;
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 64u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(64u);
|
||||
p[1][0] = asfloat(u[1]).ywxz;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,18 +3,18 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float2x4 p[4] = (float2x4[4])0;
|
||||
|
||||
float2x4 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 64u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(64u);
|
||||
p[1][0] = asfloat(u[1]).ywxz;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,32 +3,32 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x4 value) {
|
||||
buffer.Store4((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store4((offset + 16u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x4 value) {
|
||||
s.Store4((offset + 0u), asuint(value[0u]));
|
||||
s.Store4((offset + 16u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x4 value[4]) {
|
||||
void s_store(uint offset, float2x4 value[4]) {
|
||||
float2x4 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 32u)), array_1[i]);
|
||||
s_store_1((offset + (i * 32u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_4(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -36,8 +36,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 32u, tint_symbol_4(u, 64u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(32u, u_load_1(64u));
|
||||
s.Store4(32u, asuint(asfloat(u[1]).ywxz));
|
||||
s.Store(32u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -3,32 +3,32 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float2x4 value) {
|
||||
buffer.Store4((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store4((offset + 16u), asuint(value[1u]));
|
||||
void s_store_1(uint offset, float2x4 value) {
|
||||
s.Store4((offset + 0u), asuint(value[0u]));
|
||||
s.Store4((offset + 16u), asuint(value[1u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float2x4 value[4]) {
|
||||
void s_store(uint offset, float2x4 value[4]) {
|
||||
float2x4 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 32u)), array_1[i]);
|
||||
s_store_1((offset + (i * 32u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float2x4 tint_symbol_4(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -36,8 +36,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 32u, tint_symbol_4(u, 64u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(32u, u_load_1(64u));
|
||||
s.Store4(32u, asuint(asfloat(u[1]).ywxz));
|
||||
s.Store(32u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -7,18 +7,18 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x4 tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -32,8 +32,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 64u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(64u);
|
||||
w[1][0] = asfloat(u[1]).ywxz;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -7,18 +7,18 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float2x4 tint_symbol_3(uint4 buffer[8], uint offset) {
|
||||
float2x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
return float2x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]));
|
||||
return float2x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]));
|
||||
}
|
||||
|
||||
typedef float2x4 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[8], uint offset) {
|
||||
typedef float2x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float2x4 arr[4] = (float2x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -32,8 +32,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 64u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(64u);
|
||||
w[1][0] = asfloat(u[1]).ywxz;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -8,19 +8,19 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz), asfloat(a[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float3x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x3 l_a_i = tint_symbol_1(a, (48u * uint(p_a_i_save)));
|
||||
const float3x3 l_a[4] = a_load(0u);
|
||||
const float3x3 l_a_i = a_load_1((48u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_3 = (((48u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float3 l_a_i_i = asfloat(a[scalar_offset_3 / 4].xyz);
|
||||
return;
|
||||
|
||||
@@ -8,19 +8,19 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz), asfloat(a[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float3x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x3 l_a_i = tint_symbol_1(a, (48u * uint(p_a_i_save)));
|
||||
const float3x3 l_a[4] = a_load(0u);
|
||||
const float3x3 l_a_i = a_load_1((48u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_3 = (((48u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float3 l_a_i_i = asfloat(a[scalar_offset_3 / 4].xyz);
|
||||
return;
|
||||
|
||||
@@ -2,19 +2,19 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[12];
|
||||
};
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz), asfloat(a[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = a_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x3 l_a_i = tint_symbol_1(a, 96u);
|
||||
const float3x3 l_a[4] = a_load(0u);
|
||||
const float3x3 l_a_i = a_load_1(96u);
|
||||
const float3 l_a_i_i = asfloat(a[7].xyz);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,19 +2,19 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[12];
|
||||
};
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(a[scalar_offset / 4].xyz), asfloat(a[scalar_offset_1 / 4].xyz), asfloat(a[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = a_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x3 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x3 l_a_i = tint_symbol_1(a, 96u);
|
||||
const float3x3 l_a[4] = a_load(0u);
|
||||
const float3x3 l_a_i = a_load_1(96u);
|
||||
const float3 l_a_i_i = asfloat(a[7].xyz);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,16 +2,16 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[12];
|
||||
};
|
||||
|
||||
float3x3 tint_symbol(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x3 t = transpose(tint_symbol(u, 96u));
|
||||
const float3x3 t = transpose(u_load(96u));
|
||||
const float l = length(asfloat(u[1].xyz).zxy);
|
||||
const float a = abs(asfloat(u[1].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -2,16 +2,16 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[12];
|
||||
};
|
||||
|
||||
float3x3 tint_symbol(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x3 t = transpose(tint_symbol(u, 96u));
|
||||
const float3x3 t = transpose(u_load(96u));
|
||||
const float l = length(asfloat(u[1].xyz).zxy);
|
||||
const float a = abs(asfloat(u[1].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -14,19 +14,19 @@ void c(float3 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 48u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(48u));
|
||||
c(asfloat(u[3].xyz).zxy);
|
||||
d(asfloat(u[3].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -14,19 +14,19 @@ void c(float3 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 48u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(48u));
|
||||
c(asfloat(u[3].xyz).zxy);
|
||||
d(asfloat(u[3].xyz).zxy.x);
|
||||
return;
|
||||
|
||||
@@ -3,19 +3,19 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float3x3 p[4] = (float3x3[4])0;
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 96u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(96u);
|
||||
p[1][0] = asfloat(u[1].xyz).zxy;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,19 +3,19 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float3x3 p[4] = (float3x3[4])0;
|
||||
|
||||
float3x3 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 96u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(96u);
|
||||
p[1][0] = asfloat(u[1].xyz).zxy;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float3x3 value) {
|
||||
buffer.Store3((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store3((offset + 16u), asuint(value[1u]));
|
||||
buffer.Store3((offset + 32u), asuint(value[2u]));
|
||||
void s_store_1(uint offset, float3x3 value) {
|
||||
s.Store3((offset + 0u), asuint(value[0u]));
|
||||
s.Store3((offset + 16u), asuint(value[1u]));
|
||||
s.Store3((offset + 32u), asuint(value[2u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float3x3 value[4]) {
|
||||
void s_store(uint offset, float3x3 value[4]) {
|
||||
float3x3 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 48u)), array_1[i]);
|
||||
s_store_1((offset + (i * 48u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_4(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 48u, tint_symbol_4(u, 96u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(48u, u_load_1(96u));
|
||||
s.Store3(48u, asuint(asfloat(u[1].xyz).zxy));
|
||||
s.Store(48u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float3x3 value) {
|
||||
buffer.Store3((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store3((offset + 16u), asuint(value[1u]));
|
||||
buffer.Store3((offset + 32u), asuint(value[2u]));
|
||||
void s_store_1(uint offset, float3x3 value) {
|
||||
s.Store3((offset + 0u), asuint(value[0u]));
|
||||
s.Store3((offset + 16u), asuint(value[1u]));
|
||||
s.Store3((offset + 32u), asuint(value[2u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float3x3 value[4]) {
|
||||
void s_store(uint offset, float3x3 value[4]) {
|
||||
float3x3 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 48u)), array_1[i]);
|
||||
s_store_1((offset + (i * 48u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float3x3 tint_symbol_4(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 48u, tint_symbol_4(u, 96u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(48u, u_load_1(96u));
|
||||
s.Store3(48u, asuint(asfloat(u[1].xyz).zxy));
|
||||
s.Store(48u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -7,19 +7,19 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float3x3 tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 96u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(96u);
|
||||
w[1][0] = asfloat(u[1].xyz).zxy;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -7,19 +7,19 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float3x3 tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
float3x3 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
|
||||
return float3x3(asfloat(u[scalar_offset / 4].xyz), asfloat(u[scalar_offset_1 / 4].xyz), asfloat(u[scalar_offset_2 / 4].xyz));
|
||||
}
|
||||
|
||||
typedef float3x3 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[12], uint offset) {
|
||||
typedef float3x3 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x3 arr[4] = (float3x3[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 96u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(96u);
|
||||
w[1][0] = asfloat(u[1].xyz).zxy;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -8,19 +8,19 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]), asfloat(a[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float3x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x4 l_a_i = tint_symbol_1(a, (48u * uint(p_a_i_save)));
|
||||
const float3x4 l_a[4] = a_load(0u);
|
||||
const float3x4 l_a_i = a_load_1((48u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_3 = (((48u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float4 l_a_i_i = asfloat(a[scalar_offset_3 / 4]);
|
||||
return;
|
||||
|
||||
@@ -8,19 +8,19 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]), asfloat(a[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -30,8 +30,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float3x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x4 l_a_i = tint_symbol_1(a, (48u * uint(p_a_i_save)));
|
||||
const float3x4 l_a[4] = a_load(0u);
|
||||
const float3x4 l_a_i = a_load_1((48u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_3 = (((48u * uint(p_a_i_save)) + (16u * uint(p_a_i_i_save)))) / 4;
|
||||
const float4 l_a_i_i = asfloat(a[scalar_offset_3 / 4]);
|
||||
return;
|
||||
|
||||
@@ -2,19 +2,19 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[12];
|
||||
};
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]), asfloat(a[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = a_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x4 l_a_i = tint_symbol_1(a, 96u);
|
||||
const float3x4 l_a[4] = a_load(0u);
|
||||
const float3x4 l_a_i = a_load_1(96u);
|
||||
const float4 l_a_i_i = asfloat(a[7]);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,19 +2,19 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[12];
|
||||
};
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(a[scalar_offset / 4]), asfloat(a[scalar_offset_1 / 4]), asfloat(a[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = a_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -22,8 +22,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float3x4 l_a[4] = tint_symbol(a, 0u);
|
||||
const float3x4 l_a_i = tint_symbol_1(a, 96u);
|
||||
const float3x4 l_a[4] = a_load(0u);
|
||||
const float3x4 l_a_i = a_load_1(96u);
|
||||
const float4 l_a_i_i = asfloat(a[7]);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,16 +2,16 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[12];
|
||||
};
|
||||
|
||||
float3x4 tint_symbol(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x3 t = transpose(tint_symbol(u, 96u));
|
||||
const float4x3 t = transpose(u_load(96u));
|
||||
const float l = length(asfloat(u[1]).ywxz);
|
||||
const float a = abs(asfloat(u[1]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -2,16 +2,16 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[12];
|
||||
};
|
||||
|
||||
float3x4 tint_symbol(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x3 t = transpose(tint_symbol(u, 96u));
|
||||
const float4x3 t = transpose(u_load(96u));
|
||||
const float l = length(asfloat(u[1]).ywxz);
|
||||
const float a = abs(asfloat(u[1]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -14,19 +14,19 @@ void c(float4 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 48u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(48u));
|
||||
c(asfloat(u[3]).ywxz);
|
||||
d(asfloat(u[3]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -14,19 +14,19 @@ void c(float4 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -34,8 +34,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 48u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(48u));
|
||||
c(asfloat(u[3]).ywxz);
|
||||
d(asfloat(u[3]).ywxz.x);
|
||||
return;
|
||||
|
||||
@@ -3,19 +3,19 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float3x4 p[4] = (float3x4[4])0;
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 96u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(96u);
|
||||
p[1][0] = asfloat(u[1]).ywxz;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,19 +3,19 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float3x4 p[4] = (float3x4[4])0;
|
||||
|
||||
float3x4 tint_symbol_1(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 48u)));
|
||||
arr[i] = u_load_1((offset + (i * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -23,8 +23,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 96u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(96u);
|
||||
p[1][0] = asfloat(u[1]).ywxz;
|
||||
p[1][0].x = asfloat(u[1].x);
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float3x4 value) {
|
||||
buffer.Store4((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store4((offset + 16u), asuint(value[1u]));
|
||||
buffer.Store4((offset + 32u), asuint(value[2u]));
|
||||
void s_store_1(uint offset, float3x4 value) {
|
||||
s.Store4((offset + 0u), asuint(value[0u]));
|
||||
s.Store4((offset + 16u), asuint(value[1u]));
|
||||
s.Store4((offset + 32u), asuint(value[2u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float3x4 value[4]) {
|
||||
void s_store(uint offset, float3x4 value[4]) {
|
||||
float3x4 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 48u)), array_1[i]);
|
||||
s_store_1((offset + (i * 48u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_4(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 48u, tint_symbol_4(u, 96u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(48u, u_load_1(96u));
|
||||
s.Store4(48u, asuint(asfloat(u[1]).ywxz));
|
||||
s.Store(48u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -3,34 +3,34 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, float3x4 value) {
|
||||
buffer.Store4((offset + 0u), asuint(value[0u]));
|
||||
buffer.Store4((offset + 16u), asuint(value[1u]));
|
||||
buffer.Store4((offset + 32u), asuint(value[2u]));
|
||||
void s_store_1(uint offset, float3x4 value) {
|
||||
s.Store4((offset + 0u), asuint(value[0u]));
|
||||
s.Store4((offset + 16u), asuint(value[1u]));
|
||||
s.Store4((offset + 32u), asuint(value[2u]));
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, float3x4 value[4]) {
|
||||
void s_store(uint offset, float3x4 value[4]) {
|
||||
float3x4 array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 48u)), array_1[i]);
|
||||
s_store_1((offset + (i * 48u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
float3x4 tint_symbol_4(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 48u, tint_symbol_4(u, 96u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(48u, u_load_1(96u));
|
||||
s.Store4(48u, asuint(asfloat(u[1]).ywxz));
|
||||
s.Store(48u, asuint(asfloat(u[1].x)));
|
||||
return;
|
||||
|
||||
@@ -7,19 +7,19 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float3x4 tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 96u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(96u);
|
||||
w[1][0] = asfloat(u[1]).ywxz;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -7,19 +7,19 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
float3x4 tint_symbol_3(uint4 buffer[12], uint offset) {
|
||||
float3x4 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
const uint scalar_offset_1 = ((offset + 16u)) / 4;
|
||||
const uint scalar_offset_2 = ((offset + 32u)) / 4;
|
||||
return float3x4(asfloat(buffer[scalar_offset / 4]), asfloat(buffer[scalar_offset_1 / 4]), asfloat(buffer[scalar_offset_2 / 4]));
|
||||
return float3x4(asfloat(u[scalar_offset / 4]), asfloat(u[scalar_offset_1 / 4]), asfloat(u[scalar_offset_2 / 4]));
|
||||
}
|
||||
|
||||
typedef float3x4 tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[12], uint offset) {
|
||||
typedef float3x4 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float3x4 arr[4] = (float3x4[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 48u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 48u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -33,8 +33,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 96u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(96u);
|
||||
w[1][0] = asfloat(u[1]).ywxz;
|
||||
w[1][0].x = asfloat(u[1].x);
|
||||
}
|
||||
|
||||
@@ -8,24 +8,24 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = a[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = a[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = a[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = a[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -35,8 +35,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const matrix<float16_t, 4, 2> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 4, 2> l_a_i = tint_symbol_1(a, (16u * uint(p_a_i_save)));
|
||||
const matrix<float16_t, 4, 2> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 4, 2> l_a_i = a_load_1((16u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_4 = (((16u * uint(p_a_i_save)) + (4u * uint(p_a_i_i_save)))) / 4;
|
||||
uint ubo_load_4 = a[scalar_offset_4 / 4][scalar_offset_4 % 4];
|
||||
const vector<float16_t, 2> l_a_i_i = vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16)));
|
||||
|
||||
@@ -2,24 +2,24 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = a[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = a[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = a[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = a[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = a_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -27,8 +27,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 4, 2> l_a[4] = tint_symbol(a, 0u);
|
||||
const matrix<float16_t, 4, 2> l_a_i = tint_symbol_1(a, 32u);
|
||||
const matrix<float16_t, 4, 2> l_a[4] = a_load(0u);
|
||||
const matrix<float16_t, 4, 2> l_a_i = a_load_1(32u);
|
||||
uint ubo_load_4 = a[2].y;
|
||||
const vector<float16_t, 2> l_a_i_i = vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16)));
|
||||
return;
|
||||
|
||||
@@ -2,21 +2,21 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[4];
|
||||
};
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = u[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = u[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = u[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = u[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const matrix<float16_t, 2, 4> t = transpose(tint_symbol(u, 32u));
|
||||
const matrix<float16_t, 2, 4> t = transpose(u_load(32u));
|
||||
uint ubo_load_4 = u[0].y;
|
||||
const float16_t l = length(vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16))).yx);
|
||||
uint ubo_load_5 = u[0].y;
|
||||
|
||||
@@ -14,24 +14,24 @@ void c(vector<float16_t, 2> v) {
|
||||
void d(float16_t f_1) {
|
||||
}
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = u[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = u[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = u[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = u[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -39,8 +39,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 16u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(16u));
|
||||
uint ubo_load_4 = u[1].x;
|
||||
c(vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16))).yx);
|
||||
uint ubo_load_5 = u[1].x;
|
||||
|
||||
@@ -3,24 +3,24 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static matrix<float16_t, 4, 2> p[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_1(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = u[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = u[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = u[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = u[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 16u)));
|
||||
arr[i] = u_load_1((offset + (i * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -28,8 +28,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 32u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(32u);
|
||||
uint ubo_load_4 = u[0].y;
|
||||
p[1][0] = vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16))).yx;
|
||||
p[1][0].x = float16_t(f16tof32(((u[0].y) & 0xFFFF)));
|
||||
|
||||
@@ -3,40 +3,40 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
RWByteAddressBuffer s : register(u1, space0);
|
||||
|
||||
void tint_symbol_1(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 4, 2> value) {
|
||||
buffer.Store<vector<float16_t, 2> >((offset + 0u), value[0u]);
|
||||
buffer.Store<vector<float16_t, 2> >((offset + 4u), value[1u]);
|
||||
buffer.Store<vector<float16_t, 2> >((offset + 8u), value[2u]);
|
||||
buffer.Store<vector<float16_t, 2> >((offset + 12u), value[3u]);
|
||||
void s_store_1(uint offset, matrix<float16_t, 4, 2> value) {
|
||||
s.Store<vector<float16_t, 2> >((offset + 0u), value[0u]);
|
||||
s.Store<vector<float16_t, 2> >((offset + 4u), value[1u]);
|
||||
s.Store<vector<float16_t, 2> >((offset + 8u), value[2u]);
|
||||
s.Store<vector<float16_t, 2> >((offset + 12u), value[3u]);
|
||||
}
|
||||
|
||||
void tint_symbol(RWByteAddressBuffer buffer, uint offset, matrix<float16_t, 4, 2> value[4]) {
|
||||
void s_store(uint offset, matrix<float16_t, 4, 2> value[4]) {
|
||||
matrix<float16_t, 4, 2> array_1[4] = value;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
tint_symbol_1(buffer, (offset + (i * 16u)), array_1[i]);
|
||||
s_store_1((offset + (i * 16u)), array_1[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_4(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = u[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = u[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = u[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = u[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_3_ret[4];
|
||||
tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_4(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -44,8 +44,8 @@ tint_symbol_3_ret tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
tint_symbol(s, 0u, tint_symbol_3(u, 0u));
|
||||
tint_symbol_1(s, 16u, tint_symbol_4(u, 32u));
|
||||
s_store(0u, u_load(0u));
|
||||
s_store_1(16u, u_load_1(32u));
|
||||
uint ubo_load_4 = u[0].y;
|
||||
s.Store<vector<float16_t, 2> >(16u, vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16))).yx);
|
||||
s.Store<float16_t>(16u, float16_t(f16tof32(((u[0].y) & 0xFFFF))));
|
||||
|
||||
@@ -7,24 +7,24 @@ struct tint_symbol_1 {
|
||||
uint local_invocation_index : SV_GroupIndex;
|
||||
};
|
||||
|
||||
matrix<float16_t, 4, 2> tint_symbol_3(uint4 buffer[4], uint offset) {
|
||||
matrix<float16_t, 4, 2> u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint ubo_load = buffer[scalar_offset / 4][scalar_offset % 4];
|
||||
uint ubo_load = u[scalar_offset / 4][scalar_offset % 4];
|
||||
const uint scalar_offset_1 = ((offset + 4u)) / 4;
|
||||
uint ubo_load_1 = buffer[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
uint ubo_load_1 = u[scalar_offset_1 / 4][scalar_offset_1 % 4];
|
||||
const uint scalar_offset_2 = ((offset + 8u)) / 4;
|
||||
uint ubo_load_2 = buffer[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
uint ubo_load_2 = u[scalar_offset_2 / 4][scalar_offset_2 % 4];
|
||||
const uint scalar_offset_3 = ((offset + 12u)) / 4;
|
||||
uint ubo_load_3 = buffer[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
uint ubo_load_3 = u[scalar_offset_3 / 4][scalar_offset_3 % 4];
|
||||
return matrix<float16_t, 4, 2>(vector<float16_t, 2>(float16_t(f16tof32(ubo_load & 0xFFFF)), float16_t(f16tof32(ubo_load >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_1 & 0xFFFF)), float16_t(f16tof32(ubo_load_1 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_2 & 0xFFFF)), float16_t(f16tof32(ubo_load_2 >> 16))), vector<float16_t, 2>(float16_t(f16tof32(ubo_load_3 & 0xFFFF)), float16_t(f16tof32(ubo_load_3 >> 16))));
|
||||
}
|
||||
|
||||
typedef matrix<float16_t, 4, 2> tint_symbol_2_ret[4];
|
||||
tint_symbol_2_ret tint_symbol_2(uint4 buffer[4], uint offset) {
|
||||
typedef matrix<float16_t, 4, 2> u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
matrix<float16_t, 4, 2> arr[4] = (matrix<float16_t, 4, 2>[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_3(buffer, (offset + (i_1 * 16u)));
|
||||
arr[i_1] = u_load_1((offset + (i_1 * 16u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -38,8 +38,8 @@ void f_inner(uint local_invocation_index) {
|
||||
}
|
||||
}
|
||||
GroupMemoryBarrierWithGroupSync();
|
||||
w = tint_symbol_2(u, 0u);
|
||||
w[1] = tint_symbol_3(u, 32u);
|
||||
w = u_load(0u);
|
||||
w[1] = u_load_1(32u);
|
||||
uint ubo_load_4 = u[0].y;
|
||||
w[1][0] = vector<float16_t, 2>(float16_t(f16tof32(ubo_load_4 & 0xFFFF)), float16_t(f16tof32(ubo_load_4 >> 16))).yx;
|
||||
w[1][0].x = float16_t(f16tof32(((u[0].y) & 0xFFFF)));
|
||||
|
||||
@@ -8,24 +8,24 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -35,8 +35,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float4x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float4x2 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float4x2 l_a[4] = a_load(0u);
|
||||
const float4x2 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_4 = (((32u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_4 = a[scalar_offset_4 / 4];
|
||||
const float2 l_a_i_i = asfloat(((scalar_offset_4 & 2) ? ubo_load_4.zw : ubo_load_4.xy));
|
||||
|
||||
@@ -8,24 +8,24 @@ int i() {
|
||||
return counter;
|
||||
}
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i_1 = 0u; (i_1 < 4u); i_1 = (i_1 + 1u)) {
|
||||
arr[i_1] = tint_symbol_1(buffer, (offset + (i_1 * 32u)));
|
||||
arr[i_1] = a_load_1((offset + (i_1 * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -35,8 +35,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
void f() {
|
||||
const int p_a_i_save = i();
|
||||
const int p_a_i_i_save = i();
|
||||
const float4x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float4x2 l_a_i = tint_symbol_1(a, (32u * uint(p_a_i_save)));
|
||||
const float4x2 l_a[4] = a_load(0u);
|
||||
const float4x2 l_a_i = a_load_1((32u * uint(p_a_i_save)));
|
||||
const uint scalar_offset_4 = (((32u * uint(p_a_i_save)) + (8u * uint(p_a_i_i_save)))) / 4;
|
||||
uint4 ubo_load_4 = a[scalar_offset_4 / 4];
|
||||
const float2 l_a_i_i = asfloat(((scalar_offset_4 & 2) ? ubo_load_4.zw : ubo_load_4.xy));
|
||||
|
||||
@@ -2,24 +2,24 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -27,8 +27,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float4x2 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float4x2 l_a[4] = a_load(0u);
|
||||
const float4x2 l_a_i = a_load_1(64u);
|
||||
const float2 l_a_i_i = asfloat(a[4].zw);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,24 +2,24 @@ cbuffer cbuffer_a : register(b0, space0) {
|
||||
uint4 a[8];
|
||||
};
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 a_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = a[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = a[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = a[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = a[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 a_load_ret[4];
|
||||
a_load_ret a_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = a_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -27,8 +27,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float4x2 l_a[4] = tint_symbol(a, 0u);
|
||||
const float4x2 l_a_i = tint_symbol_1(a, 64u);
|
||||
const float4x2 l_a[4] = a_load(0u);
|
||||
const float4x2 l_a_i = a_load_1(64u);
|
||||
const float2 l_a_i_i = asfloat(a[4].zw);
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -2,21 +2,21 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float4x2 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float4x2 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = u[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x4 t = transpose(tint_symbol(u, 64u));
|
||||
const float2x4 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[0].zw).yx);
|
||||
const float a = abs(asfloat(u[0].zw).yx.x);
|
||||
return;
|
||||
|
||||
@@ -2,21 +2,21 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
uint4 u[8];
|
||||
};
|
||||
|
||||
float4x2 tint_symbol(uint4 buffer[8], uint offset) {
|
||||
float4x2 u_load(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = u[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
const float2x4 t = transpose(tint_symbol(u, 64u));
|
||||
const float2x4 t = transpose(u_load(64u));
|
||||
const float l = length(asfloat(u[0].zw).yx);
|
||||
const float a = abs(asfloat(u[0].zw).yx.x);
|
||||
return;
|
||||
|
||||
@@ -14,24 +14,24 @@ void c(float2 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = u[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -39,8 +39,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2].xy).yx);
|
||||
d(asfloat(u[2].xy).yx.x);
|
||||
return;
|
||||
|
||||
@@ -14,24 +14,24 @@ void c(float2 v) {
|
||||
void d(float f_1) {
|
||||
}
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = u[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -39,8 +39,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
a(tint_symbol(u, 0u));
|
||||
b(tint_symbol_1(u, 32u));
|
||||
a(u_load(0u));
|
||||
b(u_load_1(32u));
|
||||
c(asfloat(u[2].xy).yx);
|
||||
d(asfloat(u[2].xy).yx.x);
|
||||
return;
|
||||
|
||||
@@ -3,24 +3,24 @@ cbuffer cbuffer_u : register(b0, space0) {
|
||||
};
|
||||
static float4x2 p[4] = (float4x2[4])0;
|
||||
|
||||
float4x2 tint_symbol_1(uint4 buffer[8], uint offset) {
|
||||
float4x2 u_load_1(uint offset) {
|
||||
const uint scalar_offset = ((offset + 0u)) / 4;
|
||||
uint4 ubo_load = buffer[scalar_offset / 4];
|
||||
uint4 ubo_load = u[scalar_offset / 4];
|
||||
const uint scalar_offset_1 = ((offset + 8u)) / 4;
|
||||
uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
|
||||
uint4 ubo_load_1 = u[scalar_offset_1 / 4];
|
||||
const uint scalar_offset_2 = ((offset + 16u)) / 4;
|
||||
uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
|
||||
uint4 ubo_load_2 = u[scalar_offset_2 / 4];
|
||||
const uint scalar_offset_3 = ((offset + 24u)) / 4;
|
||||
uint4 ubo_load_3 = buffer[scalar_offset_3 / 4];
|
||||
uint4 ubo_load_3 = u[scalar_offset_3 / 4];
|
||||
return float4x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)), asfloat(((scalar_offset_3 & 2) ? ubo_load_3.zw : ubo_load_3.xy)));
|
||||
}
|
||||
|
||||
typedef float4x2 tint_symbol_ret[4];
|
||||
tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
typedef float4x2 u_load_ret[4];
|
||||
u_load_ret u_load(uint offset) {
|
||||
float4x2 arr[4] = (float4x2[4])0;
|
||||
{
|
||||
for(uint i = 0u; (i < 4u); i = (i + 1u)) {
|
||||
arr[i] = tint_symbol_1(buffer, (offset + (i * 32u)));
|
||||
arr[i] = u_load_1((offset + (i * 32u)));
|
||||
}
|
||||
}
|
||||
return arr;
|
||||
@@ -28,8 +28,8 @@ tint_symbol_ret tint_symbol(uint4 buffer[8], uint offset) {
|
||||
|
||||
[numthreads(1, 1, 1)]
|
||||
void f() {
|
||||
p = tint_symbol(u, 0u);
|
||||
p[1] = tint_symbol_1(u, 64u);
|
||||
p = u_load(0u);
|
||||
p[1] = u_load_1(64u);
|
||||
p[1][0] = asfloat(u[0].zw).yx;
|
||||
p[1][0].x = asfloat(u[0].z);
|
||||
return;
|
||||
|
||||
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user