summaryrefslogtreecommitdiffstats
path: root/tests/cuda
diff options
context:
space:
mode:
authorjsmall-nvidia <jsmall@nvidia.com>2021-06-10 14:57:09 -0400
committerGitHub <noreply@github.com>2021-06-10 11:57:09 -0700
commit37e8917d10626b519470f2e34625f0efe741352f (patch)
tree4e8e51bd63ebc03fcdf0c893b675906cf507d73c /tests/cuda
parent0d9bd79e8fd4d57e1a723ca6b6a45efec2b42872 (diff)
CUDA layout corner cases/testing (#1881)
* #include an absolute path didn't work - because paths were taken to always be relative. * Add support for sizeOf/alignOf/offsetOf to stdlib. Add $G intrinsic expansion that works of the generic parameters not the param type * Test cuda layout. * Fix CUDA layout issues. Fix reflection to handle other built in types. Fix __offsetOf * Tests of reflection and layout as reported directly from CUDA. * Comment about use of aligned size as size. * Fix warning from VS. * Check alignment is pow2. * Small improvements to alignment calcs. * Tab to spaces. * Fix alignment pointer sizes on 32 bit OS for CUDA. * Fix CUDA reflection on 32 bit.
Diffstat (limited to 'tests/cuda')
-rw-r--r--tests/cuda/cuda-array-layout.slang32
-rw-r--r--tests/cuda/cuda-array-layout.slang.expected.txt9
-rw-r--r--tests/cuda/cuda-layout.slang24
-rw-r--r--tests/cuda/cuda-layout.slang.expected.txt33
-rw-r--r--tests/cuda/cuda-reflection.slang28
-rw-r--r--tests/cuda/cuda-reflection.slang.expected250
6 files changed, 376 insertions, 0 deletions
diff --git a/tests/cuda/cuda-array-layout.slang b/tests/cuda/cuda-array-layout.slang
new file mode 100644
index 000000000..7fee3b192
--- /dev/null
+++ b/tests/cuda/cuda-array-layout.slang
@@ -0,0 +1,32 @@
+//TEST(compute):COMPARE_COMPUTE_EX:-cuda -compute -output-using-type -shaderobj
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0 0 0 0 0], stride=4):out,name=outputBuffer
+RWStructuredBuffer<int> outputBuffer : register(u0);
+
+struct PadLadenStruct
+{
+ double a;
+ uint8_t b;
+};
+
+// This is to check if the last half can be inserted 'inside' the spare padding of a. It should not be
+struct StructWithArray
+{
+ PadLadenStruct a[1];
+ uint8_t b;
+
+ matrix<half, 3, 3> c;
+ uint8_t d;
+};
+
+[numthreads(1, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+ StructWithArray s;
+ outputBuffer[0] = __sizeOf(s);
+
+ outputBuffer[1] = __offsetOf(s, s.a);
+ outputBuffer[2] = __offsetOf(s, s.b);
+ outputBuffer[3] = __offsetOf(s, s.c);
+ outputBuffer[4] = __offsetOf(s, s.d);
+}
diff --git a/tests/cuda/cuda-array-layout.slang.expected.txt b/tests/cuda/cuda-array-layout.slang.expected.txt
new file mode 100644
index 000000000..bc3e8bd6c
--- /dev/null
+++ b/tests/cuda/cuda-array-layout.slang.expected.txt
@@ -0,0 +1,9 @@
+type: int32_t
+48
+0
+16
+20
+44
+0
+0
+0
diff --git a/tests/cuda/cuda-layout.slang b/tests/cuda/cuda-layout.slang
new file mode 100644
index 000000000..725bf798e
--- /dev/null
+++ b/tests/cuda/cuda-layout.slang
@@ -0,0 +1,24 @@
+//TEST(compute):COMPARE_COMPUTE_EX:-cuda -compute -output-using-type -shaderobj
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0], stride=4):out,name=outputBuffer
+RWStructuredBuffer<int> outputBuffer : register(u0);
+
+#define WRITE_TYPE_ALIGN(base, type) \
+ outputBuffer[base * 4 + 0] = __alignOf<type>(); \
+ outputBuffer[base * 4 + 1] = __alignOf<vector<type, 2> >(); \
+ outputBuffer[base * 4 + 2] = __alignOf<vector<type, 3> >(); \
+ outputBuffer[base * 4 + 3] = __alignOf<vector<type, 4> >();
+
+
+[numthreads(1, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+ WRITE_TYPE_ALIGN(0, uint8_t)
+ WRITE_TYPE_ALIGN(1, uint16_t)
+ WRITE_TYPE_ALIGN(2, int)
+ WRITE_TYPE_ALIGN(3, int64_t)
+
+ WRITE_TYPE_ALIGN(4, half)
+ WRITE_TYPE_ALIGN(5, float)
+ WRITE_TYPE_ALIGN(6, double)
+}
diff --git a/tests/cuda/cuda-layout.slang.expected.txt b/tests/cuda/cuda-layout.slang.expected.txt
new file mode 100644
index 000000000..bf4edf064
--- /dev/null
+++ b/tests/cuda/cuda-layout.slang.expected.txt
@@ -0,0 +1,33 @@
+type: int32_t
+1
+2
+1
+4
+2
+4
+2
+8
+4
+8
+4
+16
+8
+16
+8
+16
+2
+4
+4
+4
+4
+8
+4
+16
+8
+16
+8
+16
+0
+0
+0
+0
diff --git a/tests/cuda/cuda-reflection.slang b/tests/cuda/cuda-reflection.slang
new file mode 100644
index 000000000..95bf591c9
--- /dev/null
+++ b/tests/cuda/cuda-reflection.slang
@@ -0,0 +1,28 @@
+// cuda-reflection.slang
+
+//TEST:REFLECTION:-stage compute -entry main -target cuda
+
+struct PadLadenStruct
+{
+ double a;
+ uint8_t b;
+};
+
+// This is to check if the last half can be inserted 'inside' the spare padding of a. It should not be
+struct StructWithArray
+{
+ PadLadenStruct a[1];
+ uint8_t c;
+
+ matrix<half, 3, 3> d;
+ uint8_t e;
+};
+
+ConstantBuffer<StructWithArray> cb;
+RWStructuredBuffer<StructWithArray> sb;
+
+[numthreads(1, 1, 1)]
+void main(
+ uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+} \ No newline at end of file
diff --git a/tests/cuda/cuda-reflection.slang.expected b/tests/cuda/cuda-reflection.slang.expected
new file mode 100644
index 000000000..d27d99557
--- /dev/null
+++ b/tests/cuda/cuda-reflection.slang.expected
@@ -0,0 +1,250 @@
+result code = 0
+standard error = {
+}
+standard output = {
+{
+ "parameters": [
+ {
+ "name": "cb",
+ "binding": {"kind": "uniform", "offset": 0, "size": 8},
+ "type": {
+ "kind": "constantBuffer",
+ "elementType": {
+ "kind": "struct",
+ "name": "StructWithArray",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "array",
+ "elementCount": 1,
+ "elementType": {
+ "kind": "struct",
+ "name": "PadLadenStruct",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "float64"
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 8}
+ },
+ {
+ "name": "b",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 8, "size": 1}
+ }
+ ]
+ },
+ "uniformStride": 16
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 16}
+ },
+ {
+ "name": "c",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 16, "size": 1}
+ },
+ {
+ "name": "d",
+ "type": {
+ "kind": "matrix",
+ "rowCount": 3,
+ "columnCount": 3,
+ "elementType": {
+ "kind": "scalar",
+ "scalarType": "float16"
+ }
+ },
+ "binding": {"kind": "uniform", "offset": 20, "size": 24}
+ },
+ {
+ "name": "e",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 44, "size": 1}
+ }
+ ]
+ },
+ "containerVarLayout": {
+ "binding": {"kind": "uniform", "offset": 0, "size": 8}
+ },
+ "elementVarLayout": {
+ "type": {
+ "kind": "struct",
+ "name": "StructWithArray",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "array",
+ "elementCount": 1,
+ "elementType": {
+ "kind": "struct",
+ "name": "PadLadenStruct",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "float64"
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 8}
+ },
+ {
+ "name": "b",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 8, "size": 1}
+ }
+ ]
+ },
+ "uniformStride": 16
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 16}
+ },
+ {
+ "name": "c",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 16, "size": 1}
+ },
+ {
+ "name": "d",
+ "type": {
+ "kind": "matrix",
+ "rowCount": 3,
+ "columnCount": 3,
+ "elementType": {
+ "kind": "scalar",
+ "scalarType": "float16"
+ }
+ },
+ "binding": {"kind": "uniform", "offset": 20, "size": 24}
+ },
+ {
+ "name": "e",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 44, "size": 1}
+ }
+ ]
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 48}
+ }
+ }
+ },
+ {
+ "name": "sb",
+ "binding": {"kind": "uniform", "offset": 8, "size": 16},
+ "type": {
+ "kind": "resource",
+ "baseShape": "structuredBuffer",
+ "access": "readWrite",
+ "resultType": {
+ "kind": "struct",
+ "name": "StructWithArray",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "array",
+ "elementCount": 1,
+ "elementType": {
+ "kind": "struct",
+ "name": "PadLadenStruct",
+ "fields": [
+ {
+ "name": "a",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "float64"
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 8}
+ },
+ {
+ "name": "b",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 8, "size": 1}
+ }
+ ]
+ },
+ "uniformStride": 16
+ },
+ "binding": {"kind": "uniform", "offset": 0, "size": 16}
+ },
+ {
+ "name": "c",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 16, "size": 1}
+ },
+ {
+ "name": "d",
+ "type": {
+ "kind": "matrix",
+ "rowCount": 3,
+ "columnCount": 3,
+ "elementType": {
+ "kind": "scalar",
+ "scalarType": "float16"
+ }
+ },
+ "binding": {"kind": "uniform", "offset": 20, "size": 24}
+ },
+ {
+ "name": "e",
+ "type": {
+ "kind": "scalar",
+ "scalarType": "uint8"
+ },
+ "binding": {"kind": "uniform", "offset": 44, "size": 1}
+ }
+ ]
+ }
+ }
+ }
+ ],
+ "entryPoints": [
+ {
+ "name": "main",
+ "stage:": "compute",
+ "parameters": [
+ {
+ "name": "dispatchThreadID",
+ "semanticName": "SV_DISPATCHTHREADID",
+ "type": {
+ "kind": "vector",
+ "elementCount": 3,
+ "elementType": {
+ "kind": "scalar",
+ "scalarType": "uint32"
+ }
+ }
+ }
+ ],
+ "threadGroupSize": [1, 1, 1]
+ }
+ ]
+}
+}