dawn-cmake/test/samples/compute_boids.wgsl.expected.hlsl
Antonio Maiorano 11d09f2fe7 HLSL: force FXC to never unroll loops
Emit the "[loop]" attribute on "for" and "while" so that FXC does not
attempt to unroll them. This is to work around an FXC bug where it fails
to unroll loops with gradient operations.

FXC ostensibly unrolls such loops because gradient operations require
uniform control flow, and loops that have varying iterations may
possibly not be uniform. Tint will eventually validate that control flow
is indeed uniform, so forcing FXC to avoid unrolling in these cases
should be fine.

Bug: tint:1112
Change-Id: I10077f8b62fbbb230a0003f3864c75a8fe0e1d18
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/69880
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
2021-11-18 13:50:12 +00:00

112 lines
3.3 KiB
HLSL

struct tint_symbol_1 {
float2 a_particlePos : TEXCOORD0;
float2 a_particleVel : TEXCOORD1;
float2 a_pos : TEXCOORD2;
};
struct tint_symbol_2 {
float4 value : SV_Position;
};
float4 vert_main_inner(float2 a_particlePos, float2 a_particleVel, float2 a_pos) {
float angle = -(atan2(a_particleVel.x, a_particleVel.y));
float2 pos = float2(((a_pos.x * cos(angle)) - (a_pos.y * sin(angle))), ((a_pos.x * sin(angle)) + (a_pos.y * cos(angle))));
return float4((pos + a_particlePos), 0.0f, 1.0f);
}
tint_symbol_2 vert_main(tint_symbol_1 tint_symbol) {
const float4 inner_result = vert_main_inner(tint_symbol.a_particlePos, tint_symbol.a_particleVel, tint_symbol.a_pos);
tint_symbol_2 wrapper_result = (tint_symbol_2)0;
wrapper_result.value = inner_result;
return wrapper_result;
}
struct tint_symbol_3 {
float4 value : SV_Target0;
};
float4 frag_main_inner() {
return float4(1.0f, 1.0f, 1.0f, 1.0f);
}
tint_symbol_3 frag_main() {
const float4 inner_result_1 = frag_main_inner();
tint_symbol_3 wrapper_result_1 = (tint_symbol_3)0;
wrapper_result_1.value = inner_result_1;
return wrapper_result_1;
}
cbuffer cbuffer_params : register(b0, space0) {
uint4 params[2];
};
RWByteAddressBuffer particlesA : register(u1, space0);
RWByteAddressBuffer particlesB : register(u2, space0);
struct tint_symbol_5 {
uint3 gl_GlobalInvocationID : SV_DispatchThreadID;
};
void comp_main_inner(uint3 gl_GlobalInvocationID) {
uint index = gl_GlobalInvocationID.x;
if ((index >= 5u)) {
return;
}
float2 vPos = asfloat(particlesA.Load2((16u * index)));
float2 vVel = asfloat(particlesA.Load2(((16u * index) + 8u)));
float2 cMass = float2(0.0f, 0.0f);
float2 cVel = float2(0.0f, 0.0f);
float2 colVel = float2(0.0f, 0.0f);
int cMassCount = 0;
int cVelCount = 0;
float2 pos = float2(0.0f, 0.0f);
float2 vel = float2(0.0f, 0.0f);
{
[loop] for(uint i = 0u; (i < 5u); i = (i + 1u)) {
if ((i == index)) {
continue;
}
pos = asfloat(particlesA.Load2((16u * i))).xy;
vel = asfloat(particlesA.Load2(((16u * i) + 8u))).xy;
if ((distance(pos, vPos) < asfloat(params[0].y))) {
cMass = (cMass + pos);
cMassCount = (cMassCount + 1);
}
if ((distance(pos, vPos) < asfloat(params[0].z))) {
colVel = (colVel - (pos - vPos));
}
if ((distance(pos, vPos) < asfloat(params[0].w))) {
cVel = (cVel + vel);
cVelCount = (cVelCount + 1);
}
}
}
if ((cMassCount > 0)) {
cMass = ((cMass / float2(float(cMassCount), float(cMassCount))) - vPos);
}
if ((cVelCount > 0)) {
cVel = (cVel / float2(float(cVelCount), float(cVelCount)));
}
vVel = (((vVel + (cMass * asfloat(params[1].x))) + (colVel * asfloat(params[1].y))) + (cVel * asfloat(params[1].z)));
vVel = (normalize(vVel) * clamp(length(vVel), 0.0f, 0.100000001f));
vPos = (vPos + (vVel * asfloat(params[0].x)));
if ((vPos.x < -1.0f)) {
vPos.x = 1.0f;
}
if ((vPos.x > 1.0f)) {
vPos.x = -1.0f;
}
if ((vPos.y < -1.0f)) {
vPos.y = 1.0f;
}
if ((vPos.y > 1.0f)) {
vPos.y = -1.0f;
}
particlesB.Store2((16u * index), asuint(vPos));
particlesB.Store2(((16u * index) + 8u), asuint(vVel));
}
[numthreads(1, 1, 1)]
void comp_main(tint_symbol_5 tint_symbol_4) {
comp_main_inner(tint_symbol_4.gl_GlobalInvocationID);
return;
}