1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
|
//TEST:SIMPLE(filecheck=SPIRV): -target spirv -emit-spirv-directly -profile cs_6_2 -entry computeMain -line-directive-mode none -fvk-use-dx-layout
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUFFER):-slang -compute -dx12 -Xslang... -Xdxc -fvk-use-dx-layout -Xdxc -enable-16bit-types -X. -output-using-type
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUFFER):-slang -compute -vk -emit-spirv-directly -Xslang... -fvk-use-dx-layout -X. -output-using-type
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUFFER):-cuda -compute -output-using-type
//TEST:REFLECTION(filecheck=REFLECT):-stage compute -entry computeMain -target spirv -profile cs_6_2 -no-codegen -line-directive-mode none -fvk-use-dx-layout
// dxc: -T cs_6_2 -E computeMain -spirv -fvk-use-dx-layout -enable-16bit-types
//TEST_INPUT:ubuffer(stride=4, count=24):out,name=outputBuffer
RWStructuredBuffer<float> outputBuffer;
//TEST_INPUT:set Constants.v0=1.0
//TEST_INPUT:set Constants.v1={2.0,3.0,4.0}
//TEST_INPUT:set Constants.v2=0x4500
//TEST_INPUT:set Constants.v3={0x4600,0x4700,0x4800}
//TEST_INPUT:set Constants.v4=0x4880
//TEST_INPUT:set Constants.v5={0x4900,0x4980,0x4a00}
//TEST_INPUT:set Constants.v6=0x4a80
//TEST_INPUT:set Constants.v7={0x4b00,0x4b80,0x4c00,0x4c40}
//TEST_INPUT:set Constants.v8={0x4c80,0x4cc0,0x4d00}
//TEST_INPUT:set Constants.v9=0x4d40
//TEST_INPUT:set Constants.v10={22.0,23.0,24.0}
// Checks cbuffer packing rule cases involving 3-element vectors.
// HLSL aligns at the element size, while GLSL std140 and std430
// align to the size of a 4-element vector of the element type.
// Checks cases that are not compatible with std140 and std430.
cbuffer Constants
{
// float3 v1 should be packed in with float v0.
float v0;
float3 v1;
// v2,v3,v4,v5 should all be packed together
float16_t v2;
vector<float16_t, 3> v3;
float16_t v4;
vector<float16_t, 3> v5;
float16_t v6;
vector<float16_t, 4> v7;
vector<float16_t, 3> v8;
// There should be a 2-byte gap between v4 and v5.
float16_t v9;
float3 v10;
};
// SPIRV: OpMemberDecorate {{.*}} 0 Offset 0
// SPIRV: OpMemberDecorate {{.*}} 1 Offset 4
// SPIRV: OpMemberDecorate {{.*}} 2 Offset 16
// SPIRV: OpMemberDecorate {{.*}} 3 Offset 18
// SPIRV: OpMemberDecorate {{.*}} 4 Offset 24
// SPIRV: OpMemberDecorate {{.*}} 5 Offset 26
// SPIRV: OpMemberDecorate {{.*}} 6 Offset 32
// SPIRV: OpMemberDecorate {{.*}} 7 Offset 34
// SPIRV: OpMemberDecorate {{.*}} 8 Offset 42
// SPIRV: OpMemberDecorate {{.*}} 9 Offset 48
// SPIRV: OpMemberDecorate {{.*}} 10 Offset 52
// REFLECT: "name": "v0",
// REFLECT: "offset": 0
// REFLECT: "name": "v1",
// REFLECT: "offset": 4
// REFLECT: "name": "v2",
// REFLECT: "offset": 16
// REFLECT: "name": "v3",
// REFLECT: "offset": 18
// REFLECT: "name": "v4",
// REFLECT: "offset": 24
// REFLECT: "name": "v5",
// REFLECT: "offset": 26
// REFLECT: "name": "v6",
// REFLECT: "offset": 32
// REFLECT: "name": "v7",
// REFLECT: "offset": 34
// REFLECT: "name": "v8",
// REFLECT: "offset": 42
// REFLECT: "name": "v9",
// REFLECT: "offset": 48
// REFLECT: "name": "v10",
// REFLECT: "offset": 52
[numthreads(1, 1, 1)]
void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
{
int i = 0;
outputBuffer[i++] = v0;
outputBuffer[i++] = v1.x;
outputBuffer[i++] = v1.y;
outputBuffer[i++] = v1.z;
outputBuffer[i++] = v2;
outputBuffer[i++] = v3.x;
outputBuffer[i++] = v3.y;
outputBuffer[i++] = v3.z;
outputBuffer[i++] = v4;
outputBuffer[i++] = v5.x;
outputBuffer[i++] = v5.y;
outputBuffer[i++] = v5.z;
outputBuffer[i++] = v6;
outputBuffer[i++] = v7.x;
outputBuffer[i++] = v7.y;
outputBuffer[i++] = v7.z;
outputBuffer[i++] = v7.w;
outputBuffer[i++] = v8.x;
outputBuffer[i++] = v8.y;
outputBuffer[i++] = v8.z;
outputBuffer[i++] = v9;
outputBuffer[i++] = v10.x;
outputBuffer[i++] = v10.y;
outputBuffer[i++] = v10.z;
// BUFFER: 1
// BUFFER-NEXT: 2
// BUFFER-NEXT: 3
// BUFFER-NEXT: 4
// BUFFER-NEXT: 5
// BUFFER-NEXT: 6
// BUFFER-NEXT: 7
// BUFFER-NEXT: 8
// BUFFER-NEXT: 9
// BUFFER-NEXT: 10
// BUFFER-NEXT: 11
// BUFFER-NEXT: 12
// BUFFER-NEXT: 13
// BUFFER-NEXT: 14
// BUFFER-NEXT: 15
// BUFFER-NEXT: 16
// BUFFER-NEXT: 17
// BUFFER-NEXT: 18
// BUFFER-NEXT: 19
// BUFFER-NEXT: 20
// BUFFER-NEXT: 21
// BUFFER-NEXT: 22
// BUFFER-NEXT: 23
// BUFFER-NEXT: 24
}
|
