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
|
// column-major-with-row-major-operations.slang
// Metal/CPP/CUDA do not deal with packing currently, different results will occur.
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-cpu -compute -xslang -DTARGET_WITHOUT_PACKING
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-cuda -compute -xslang -DTARGET_WITHOUT_PACKING
//TEST(compute, vulkan):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-vk -compute -emit-spirv-via-glsl
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-slang -compute
//TEST(compute, vulkan):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-vk -compute
//TEST(compute):COMPARE_COMPUTE_EX(filecheck-buffer=BUF):-mtl -compute
// CPP/CUDA due to natural layout rules will recieve the following ROW matrix:
// {1,2,3}
// {0,4,5}
// {6,0,7}
// GLSL/SPIRV/HLSL/Metal with cbuffer layout rules will recieve the following ROW/COL matrix:
// {1,2,3}
// {4,5,6}
// {7,8,9}
//TEST_INPUT:cbuffer(data=[1.0 2.0 3.0 0.0 4.0 5.0 6.0 0.0 7.0 8.0 9.0 0.0]):name matrixTestCBuf1
ConstantBuffer<row_major float3x3> matrixTestCBuf1;
// CPP/Metal/CUDA due to not having memory packing will recieve the following COL matrix post-transpose:
// {1,0,6}
// {2,4,0}
// {3,5,7}
//TEST_INPUT:cbuffer(data=[1.0 2.0 3.0 0.0 4.0 5.0 6.0 0.0 7.0 8.0 9.0 0.0]):name matrixTestCBuf2
ConstantBuffer<column_major float3x3> matrixTestCBuf2;
//TEST_INPUT:cbuffer(data=[1.0 2.0 3.0 0.0 4.0 5.0 6.0 0.0]):name NeedsPadding
cbuffer NeedsPadding
{
float3 data1;
float3 data2;
};
//TEST_INPUT:ubuffer(data=[0 0 0 0], stride=4):out,name output
RWStructuredBuffer<uint> output;
bool floatCheck(float data, float valueToCheckFor)
{
return data < (valueToCheckFor + 0.001) && data > valueToCheckFor - 0.001;
}
[numthreads(1, 1, 1)]
void computeMain(uint3 tid : SV_DispatchThreadID)
{
float3x3 matrixTest1;
matrixTest1 = matrixTestCBuf1;
float3x3 matrixTest2;
matrixTest2 = matrixTestCBuf2;
output[0] = uint(true
#ifdef TARGET_WITHOUT_PACKING
&& floatCheck(matrixTest1[0][0], 1)
&& floatCheck(matrixTest1[0][1], 2)
&& floatCheck(matrixTest1[0][2], 3)
&& floatCheck(matrixTest1[1][0], 0)
&& floatCheck(matrixTest1[1][1], 4)
&& floatCheck(matrixTest1[1][2], 5)
&& floatCheck(matrixTest1[2][0], 6)
&& floatCheck(matrixTest1[2][1], 0)
&& floatCheck(matrixTest1[2][2], 7)
&& floatCheck(matrixTest2[0][0], 1)
&& floatCheck(matrixTest2[0][1], 0)
&& floatCheck(matrixTest2[0][2], 6)
&& floatCheck(matrixTest2[1][0], 2)
&& floatCheck(matrixTest2[1][1], 4)
&& floatCheck(matrixTest2[1][2], 0)
&& floatCheck(matrixTest2[2][0], 3)
&& floatCheck(matrixTest2[2][1], 5)
&& floatCheck(matrixTest2[2][2], 7)
&& floatCheck(data1[0], 1)
&& floatCheck(data1[1], 2)
&& floatCheck(data1[2], 3)
&& floatCheck(data2[0], 0)
&& floatCheck(data2[1], 4)
&& floatCheck(data2[2], 5)
#else
&& floatCheck(matrixTest1[0][0], 1)
&& floatCheck(matrixTest1[0][1], 2)
&& floatCheck(matrixTest1[0][2], 3)
&& floatCheck(matrixTest1[1][0], 4)
&& floatCheck(matrixTest1[1][1], 5)
&& floatCheck(matrixTest1[1][2], 6)
&& floatCheck(matrixTest1[2][0], 7)
&& floatCheck(matrixTest1[2][1], 8)
&& floatCheck(matrixTest1[2][2], 9)
&& floatCheck(matrixTest2[0][0], 1)
&& floatCheck(matrixTest2[0][1], 4)
&& floatCheck(matrixTest2[0][2], 7)
&& floatCheck(matrixTest2[1][0], 2)
&& floatCheck(matrixTest2[1][1], 5)
&& floatCheck(matrixTest2[1][2], 8)
&& floatCheck(matrixTest2[2][0], 3)
&& floatCheck(matrixTest2[2][1], 6)
&& floatCheck(matrixTest2[2][2], 9)
&& floatCheck(data1[0], 1)
&& floatCheck(data1[1], 2)
&& floatCheck(data1[2], 3)
&& floatCheck(data2[0], 4)
&& floatCheck(data2[1], 5)
&& floatCheck(data2[2], 6)
#endif
);
//BUF: 1
}
|
