Move double-precision tests to DoublePrecision class (microsoft#8157)

Addresses mmerecki's comment on PR microsoft#8134: several tests operating on
double-precision values remained in `DxilConf_SM69_Vectorized_Core` but
belong in `DxilConf_SM69_Vectorized_Double`.

## Changes

**Moved from `DxilConf_SM69_Vectorized_Core` to
`DxilConf_SM69_Vectorized_Double`:**
- `CastToFloat64_*` (9 variants: HLSLBool_t, HLSLHalf_t, float,
uint16_t, uint32_t, uint64_t, int16_t, int32_t, int64_t)
- `AsDouble_uint32_t`
- `AsUint_SplitDouble_double`

**Technical change:**
- Removed tests from `DxilConf_SM69_Vectorized_Core` class
- Added same tests to `DxilConf_SM69_Vectorized_Double` class using
`HLK_TEST_DOUBLE` macro
- Tests are now properly associated with
`Device.Graphics.D3D12.DXILCore.ShaderModel69.DoublePrecision`
specification
- Tests organized in sections: "Explicit Cast to Double (from various
types)" and "AsType (double precision)"

Test names and OpType template parameters remain unchanged.

<!-- START COPILOT ORIGINAL PROMPT -->



<details>

<summary>Original prompt</summary>

> Create a follow-up PR in microsoft/DirectXShaderCompiler to address
mmerecki's comment on PR microsoft#8134
(microsoft#8134 (comment)).
> 
> Problem:
> PR microsoft#8134 split long vector execution tests into separate classes based
on HLK requirements, introducing DxilConf_SM69_Vectorized_Core and
DxilConf_SM69_Vectorized_Double test classes in
tools/clang/unittests/HLSLExec/LongVectors.cpp. mmerecki noted that
several tests still live under DxilConf_SM69_Vectorized_Core but appear
to be double-precision related and should be moved to
DxilConf_SM69_Vectorized_Double.
> 
> The tests listed by mmerecki:
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_HLSLBool_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_HLSLHalf_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_float
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_uint16_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_uint32_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_uint64_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_int16_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_int32_t
> - DxilConf_SM69_Vectorized_Core::CastToFloat64_int64_t
> - DxilConf_SM69_Vectorized_Core::AsDouble_uint32_t
> - DxilConf_SM69_Vectorized_Core::AsUint_SplitDouble_double
> 
> Expected behavior / fix:
> - Update tools/clang/unittests/HLSLExec/LongVectors.cpp so that these
tests are associated with the DoublePrecision HLK specification.
> - Concretely, remove the above tests from
DxilConf_SM69_Vectorized_Core and add equivalent tests to
DxilConf_SM69_Vectorized_Double.
> - Prefer using HLK_TEST_DOUBLE for the tests being moved so they pick
up the "Device.Graphics.D3D12.DXILCore.ShaderModel69.DoublePrecision"
specification.
> - Ensure the resulting test method names remain stable/consistent and
compilation passes.
> - Validate that moved tests still call the correct OpType values
(OpType::CastToFloat64, OpType::AsDouble, OpType::AsUint_SplitDouble)
with appropriate template type arguments.
> 
> Notes / pointers:
> - Relevant file: tools/clang/unittests/HLSLExec/LongVectors.cpp (see
locations around DxilConf_SM69_Vectorized_Core and
DxilConf_SM69_Vectorized_Double classes).
> - DxilConf_SM69_Vectorized_Double class defines HLK_TEST_DOUBLE and
already contains many double tests.
> 
> Deliverables:
> - A new PR with changes only necessary to re-home these tests from
Core to Double.
> - Include a concise PR description referencing mmerecki's comment and
why the tests belong in the DoublePrecision group.
> 
> 


</details>



<!-- START COPILOT CODING AGENT SUFFIX -->

*This pull request was created from Copilot chat.*
>

<!-- START COPILOT CODING AGENT TIPS -->
---

✨ Let Copilot coding agent [set things up for
you](https://github.com/microsoft/DirectXShaderCompiler/issues/new?title=✨+Set+up+Copilot+instructions&body=Configure%20instructions%20for%20this%20repository%20as%20documented%20in%20%5BBest%20practices%20for%20Copilot%20coding%20agent%20in%20your%20repository%5D%28https://gh.io/copilot-coding-agent-tips%29%2E%0A%0A%3COnboard%20this%20repo%3E&assignees=copilot)
— coding agent works faster and does higher quality work when set up for
your repo.

---------

Co-authored-by: copilot-swe-agent[bot] <[email protected]>
Co-authored-by: alsepkow <[email protected]>

Author: Copilot

tools/clang/unittests/HLSLExec/LongVectors.cpp

@@ -2140,7 +2140,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, HLSLBool_t);
   HLK_TEST(CastToFloat16, HLSLBool_t);
   HLK_TEST(CastToFloat32, HLSLBool_t);
-  HLK_TEST(CastToFloat64, HLSLBool_t);
 
   HLK_TEST(CastToBool, HLSLHalf_t);
   HLK_TEST(CastToInt16, HLSLHalf_t);
@@ -2150,7 +2149,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint32_FromFP, HLSLHalf_t);
   HLK_TEST(CastToUint64_FromFP, HLSLHalf_t);
   HLK_TEST(CastToFloat32, HLSLHalf_t);
-  HLK_TEST(CastToFloat64, HLSLHalf_t);
 
   HLK_TEST(CastToBool, float);
   HLK_TEST(CastToInt16, float);
@@ -2160,7 +2158,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint32_FromFP, float);
   HLK_TEST(CastToUint64_FromFP, float);
   HLK_TEST(CastToFloat16, float);
-  HLK_TEST(CastToFloat64, float);
 
   HLK_TEST(CastToBool, uint16_t);
   HLK_TEST(CastToInt16, uint16_t);
@@ -2170,7 +2167,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, uint16_t);
   HLK_TEST(CastToFloat16, uint16_t);
   HLK_TEST(CastToFloat32, uint16_t);
-  HLK_TEST(CastToFloat64, uint16_t);
 
   HLK_TEST(CastToBool, uint32_t);
   HLK_TEST(CastToInt16, uint32_t);
@@ -2180,7 +2176,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, uint32_t);
   HLK_TEST(CastToFloat16, uint32_t);
   HLK_TEST(CastToFloat32, uint32_t);
-  HLK_TEST(CastToFloat64, uint32_t);
 
   HLK_TEST(CastToBool, uint64_t);
   HLK_TEST(CastToInt16, uint64_t);
@@ -2190,7 +2185,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint32, uint64_t);
   HLK_TEST(CastToFloat16, uint64_t);
   HLK_TEST(CastToFloat32, uint64_t);
-  HLK_TEST(CastToFloat64, uint64_t);
 
   HLK_TEST(CastToBool, int16_t);
   HLK_TEST(CastToInt32, int16_t);
@@ -2200,7 +2194,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, int16_t);
   HLK_TEST(CastToFloat16, int16_t);
   HLK_TEST(CastToFloat32, int16_t);
-  HLK_TEST(CastToFloat64, int16_t);
 
   HLK_TEST(CastToBool, int32_t);
   HLK_TEST(CastToInt16, int32_t);
@@ -2210,7 +2203,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, int32_t);
   HLK_TEST(CastToFloat16, int32_t);
   HLK_TEST(CastToFloat32, int32_t);
-  HLK_TEST(CastToFloat64, int32_t);
 
   HLK_TEST(CastToBool, int64_t);
   HLK_TEST(CastToInt16, int64_t);
@@ -2220,7 +2212,6 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(CastToUint64, int64_t);
   HLK_TEST(CastToFloat16, int64_t);
   HLK_TEST(CastToFloat32, int64_t);
-  HLK_TEST(CastToFloat64, int64_t);
 
   // Trigonometric
   HLK_TEST(Acos, HLSLHalf_t);
@@ -2255,11 +2246,9 @@ class DxilConf_SM69_Vectorized_Core : public TestClassCommon {
   HLK_TEST(AsFloat, uint32_t);
   HLK_TEST(AsInt, uint32_t);
   HLK_TEST(AsUint, uint32_t);
-  HLK_TEST(AsDouble, uint32_t);
   HLK_TEST(AsFloat16, HLSLHalf_t);
   HLK_TEST(AsInt16, HLSLHalf_t);
   HLK_TEST(AsUint16, HLSLHalf_t);
-  HLK_TEST(AsUint_SplitDouble, double);
 
   // Unary Math
   HLK_TEST(Abs, int16_t);
@@ -2802,6 +2791,21 @@ class DxilConf_SM69_Vectorized_Double : public TestClassCommon {
   HLK_TEST_DOUBLE(CastToFloat16, double);
   HLK_TEST_DOUBLE(CastToFloat32, double);
 
+  // Explicit Cast to Double (from various types)
+  HLK_TEST_DOUBLE(CastToFloat64, HLSLBool_t);
+  HLK_TEST_DOUBLE(CastToFloat64, HLSLHalf_t);
+  HLK_TEST_DOUBLE(CastToFloat64, float);
+  HLK_TEST_DOUBLE(CastToFloat64, uint16_t);
+  HLK_TEST_DOUBLE(CastToFloat64, uint32_t);
+  HLK_TEST_DOUBLE(CastToFloat64, uint64_t);
+  HLK_TEST_DOUBLE(CastToFloat64, int16_t);
+  HLK_TEST_DOUBLE(CastToFloat64, int32_t);
+  HLK_TEST_DOUBLE(CastToFloat64, int64_t);
+
+  // AsType (double precision)
+  HLK_TEST_DOUBLE(AsDouble, uint32_t);
+  HLK_TEST_DOUBLE(AsUint_SplitDouble, double);
+
   // Unary Math
   HLK_TEST_DOUBLE(Abs, double);
   HLK_TEST_DOUBLE(Sign, double);