Implement `-fvk-use-dx-layout` (#4912)

* Implement `-fvk-use-dx-layout`

Fixes: #4126

Changes:
* Added fvk-use-dx-layout
* Modified `HLSLConstantBufferLayoutRulesImpl` for correctness (ex: Array is always 16 byte aligned)
* Added kFXCShaderResourceLayoutRulesFamilyImpl and kFXCConstantBufferLayoutRulesFamilyImpl to handle fvk-use-dx-layout
* Added `ConstantBufferLayoutRules` to manage constant buffer rules
* Added `alignCompositeElementOfNonAggregate`/`alignCompositeElementOfAggregate` to handle forced alignment of composites for ConstantBuffers
* `StructuredBuffer` rules are mostly equal to `scalar` layout, not much was needed to be changed to support this behavior.

* seperate legacy constant buffer and how Slang does constant-buffer normally

* undo an addition

* remove accidental test

* Address review and fix

Address review and remove GLSL support since GLSL requires a seperate legalization (need to linearlize structs like with `legalizeMetalIR` to assign explicit offsets)

* comments

* remove aggregate and non-aggregate logic

We don't need this distinction for the logic

---------

Co-authored-by: Yong He <yonghe@outlook.com>

1 files changed, 46 insertions, 0 deletions

diff --git a/tests/spirv/cbuffer-dx-layout-3.slang b/tests/spirv/cbuffer-dx-layout-3.slang
new file mode 100644
index 000000000..e05cd25f1
--- /dev/null
+++ b/tests/spirv/cbuffer-dx-layout-3.slang
@@ -0,0 +1,46 @@
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX(filecheck-buffer=CHECK):-vk -compute -xslang -fvk-use-dx-layout -emit-spirv-directly
+//TEST:SIMPLE(filecheck=SPIRV): -target spirv -entry computeMain -stage compute -fvk-use-dx-layout 
+//TEST_INPUT:cbuffer(data=[1 0 0 0   2.0 3.0 4.0 0   5.0 6.0 7.0 0   8.0 9.0 10.0 11]):name=Test
+
+//SPIRV: ArrayStride 16
+
+cbuffer Test
+{
+//SPIRV: Offset 0
+    uint v0;
+
+//SPIRV: Offset 16
+// matrix always start on a new register
+    float3x3 v1;
+//SPIRV: Offset 60
+// Non-matrix can pack with a partially filled register
+    uint v2;
+};
+
+//TEST_INPUT:ubuffer(data=[0], stride=4):out,name outputBuffer
+RWStructuredBuffer<int> outputBuffer;
+
+__generic<T : IArithmetic, let N : int>
+bool comp(vector<T,N> v1, vector<T,N> v2)
+{
+    for (uint i = 0; i < N; i++)
+        if (v1[i] != v2[i])
+            return false;
+
+    return true;
+}
+
+[shader("compute")]
+[numthreads(2, 2, 1)]
+void computeMain()
+{
+    // CHECK: 64
+
+    outputBuffer[0] = (true
+            && v0 == 1
+            && comp(v1[0], float3(2, 3, 4))
+            && comp(v1[1], float3(5, 6, 7))
+            && comp(v1[2], float3(8, 9, 10))
+            && v2 == 11
+        ) ? 100 : 0;
+}
\ No newline at end of file