dawn-cmake/test/tint/expressions/binary/mul/mat3x2-vec3/f32.wgsl.expected.fxc.hlsl
Ben Clayton 1a1b5278d5 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>
2023-02-24 17:16:55 +00:00

19 lines
697 B
HLSL

cbuffer cbuffer_data : register(b0, space0) {
uint4 data[3];
};
float3x2 data_load(uint offset) {
const uint scalar_offset = ((offset + 0u)) / 4;
uint4 ubo_load = data[scalar_offset / 4];
const uint scalar_offset_1 = ((offset + 8u)) / 4;
uint4 ubo_load_1 = data[scalar_offset_1 / 4];
const uint scalar_offset_2 = ((offset + 16u)) / 4;
uint4 ubo_load_2 = data[scalar_offset_2 / 4];
return float3x2(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)));
}
void main() {
const float2 x = mul(asfloat(data[2].xyz), data_load(0u));
return;
}