mirror of
https://github.com/italicsjenga/vello.git
synced 2024-10-17 23:11:30 +11:00
d948126c16
Make max workgroup size 256 and respect LG_WG_FACTOR. Because the monoid scans only support a height of 2, this will reduce the maximum scene complexity we can render. But it also increases compatibility. Supporting larger scans is a TODO.
95 lines
2.5 KiB
HLSL
Generated
95 lines
2.5 KiB
HLSL
Generated
struct DrawMonoid
|
|
{
|
|
uint path_ix;
|
|
uint clip_ix;
|
|
};
|
|
|
|
static const uint3 gl_WorkGroupSize = uint3(256u, 1u, 1u);
|
|
|
|
static const DrawMonoid _18 = { 0u, 0u };
|
|
|
|
RWByteAddressBuffer _57 : register(u0, space0);
|
|
|
|
static uint3 gl_LocalInvocationID;
|
|
static uint3 gl_GlobalInvocationID;
|
|
struct SPIRV_Cross_Input
|
|
{
|
|
uint3 gl_LocalInvocationID : SV_GroupThreadID;
|
|
uint3 gl_GlobalInvocationID : SV_DispatchThreadID;
|
|
};
|
|
|
|
groupshared DrawMonoid sh_scratch[256];
|
|
|
|
DrawMonoid combine_tag_monoid(DrawMonoid a, DrawMonoid b)
|
|
{
|
|
DrawMonoid c;
|
|
c.path_ix = a.path_ix + b.path_ix;
|
|
c.clip_ix = a.clip_ix + b.clip_ix;
|
|
return c;
|
|
}
|
|
|
|
DrawMonoid tag_monoid_identity()
|
|
{
|
|
return _18;
|
|
}
|
|
|
|
void comp_main()
|
|
{
|
|
uint ix = gl_GlobalInvocationID.x * 8u;
|
|
DrawMonoid _61;
|
|
_61.path_ix = _57.Load(ix * 8 + 0);
|
|
_61.clip_ix = _57.Load(ix * 8 + 4);
|
|
DrawMonoid local[8];
|
|
local[0].path_ix = _61.path_ix;
|
|
local[0].clip_ix = _61.clip_ix;
|
|
DrawMonoid param_1;
|
|
for (uint i = 1u; i < 8u; i++)
|
|
{
|
|
DrawMonoid param = local[i - 1u];
|
|
DrawMonoid _88;
|
|
_88.path_ix = _57.Load((ix + i) * 8 + 0);
|
|
_88.clip_ix = _57.Load((ix + i) * 8 + 4);
|
|
param_1.path_ix = _88.path_ix;
|
|
param_1.clip_ix = _88.clip_ix;
|
|
local[i] = combine_tag_monoid(param, param_1);
|
|
}
|
|
DrawMonoid agg = local[7];
|
|
sh_scratch[gl_LocalInvocationID.x] = agg;
|
|
for (uint i_1 = 0u; i_1 < 8u; i_1++)
|
|
{
|
|
GroupMemoryBarrierWithGroupSync();
|
|
if (gl_LocalInvocationID.x >= (1u << i_1))
|
|
{
|
|
DrawMonoid other = sh_scratch[gl_LocalInvocationID.x - (1u << i_1)];
|
|
DrawMonoid param_2 = other;
|
|
DrawMonoid param_3 = agg;
|
|
agg = combine_tag_monoid(param_2, param_3);
|
|
}
|
|
GroupMemoryBarrierWithGroupSync();
|
|
sh_scratch[gl_LocalInvocationID.x] = agg;
|
|
}
|
|
GroupMemoryBarrierWithGroupSync();
|
|
DrawMonoid row = tag_monoid_identity();
|
|
if (gl_LocalInvocationID.x > 0u)
|
|
{
|
|
row = sh_scratch[gl_LocalInvocationID.x - 1u];
|
|
}
|
|
for (uint i_2 = 0u; i_2 < 8u; i_2++)
|
|
{
|
|
DrawMonoid param_4 = row;
|
|
DrawMonoid param_5 = local[i_2];
|
|
DrawMonoid m = combine_tag_monoid(param_4, param_5);
|
|
uint _177 = ix + i_2;
|
|
_57.Store(_177 * 8 + 0, m.path_ix);
|
|
_57.Store(_177 * 8 + 4, m.clip_ix);
|
|
}
|
|
}
|
|
|
|
[numthreads(256, 1, 1)]
|
|
void main(SPIRV_Cross_Input stage_input)
|
|
{
|
|
gl_LocalInvocationID = stage_input.gl_LocalInvocationID;
|
|
gl_GlobalInvocationID = stage_input.gl_GlobalInvocationID;
|
|
comp_main();
|
|
}
|