[SPIR-V] Fix asdouble issue in SPIRV codegen to correctly generate OpBitCast instruction.

[luciechoi]

Generate OpBitCast instruction for pointer cast operation if the element type is different.

The HLSL for the unit test is

StructuredBuffer<uint2> In : register(t0);

RWStructuredBuffer<double2> Out : register(u2);


[numthreads(1,1,1)]
void main() {
  Out[0] = asdouble(In[0], In[1]);
}

Resolves #153513

[llvmbot]

@llvm/pr-subscribers-backend-spir-v

Author: Lucie Choi (luciechoi)

Changes

Generate OpBitCast instruction for pointer cast operation if the element type is different.

The HLSL for the unit test is

StructuredBuffer<uint2> In : register(t0);

RWStructuredBuffer<double2> Out : register(u2);


[numthreads(1,1,1)]
void main() {
  Out[0] = asdouble(In[0], In[1]);
}

Resolves #153513

---

Full diff: https://github.com/llvm/llvm-project/pull/161891.diff

2 Files Affected:

• (modified) llvm/lib/Target/SPIRV/SPIRVLegalizePointerCast.cpp (+24-1)
• (added) llvm/test/CodeGen/SPIRV/pointers/ptrcast-bitcast.ll (+29)

diff --git a/llvm/lib/Target/SPIRV/SPIRVLegalizePointerCast.cpp b/llvm/lib/Target/SPIRV/SPIRVLegalizePointerCast.cpp
index 75055072573b3..ebd957c42762c 100644
--- a/llvm/lib/Target/SPIRV/SPIRVLegalizePointerCast.cpp
+++ b/llvm/lib/Target/SPIRV/SPIRVLegalizePointerCast.cpp
@@ -188,8 +188,31 @@ class SPIRVLegalizePointerCast : public FunctionPass {
     FixedVectorType *SrcType = cast<FixedVectorType>(Src->getType());
     FixedVectorType *DstType =
         cast<FixedVectorType>(GR->findDeducedElementType(Dst));
-    assert(DstType->getNumElements() >= SrcType->getNumElements());
+    auto dstNumElements = DstType->getNumElements();
+    auto srcNumElements = SrcType->getNumElements();
+
+    // if the element type differs, it is a bitcast.
+    if (DstType->getElementType() != SrcType->getElementType()) {
+      // Support bitcast between vectors of different sizes only if
+      // the total bitwidth is the same.
+      auto dstBitWidth =
+          DstType->getElementType()->getScalarSizeInBits() * dstNumElements;
+      auto srcBitWidth =
+          SrcType->getElementType()->getScalarSizeInBits() * srcNumElements;
+      assert(dstBitWidth == srcBitWidth &&
+             "Unsupported bitcast between vectors of different sizes.");
+
+      Src =
+          B.CreateIntrinsic(Intrinsic::spv_bitcast, {DstType, SrcType}, {Src});
+      buildAssignType(B, DstType, Src);
+      SrcType = DstType;
+
+      StoreInst *SI = B.CreateStore(Src, Dst);
+      SI->setAlignment(Alignment);
+      return SI;
+    }
 
+    assert(DstType->getNumElements() >= SrcType->getNumElements());
     LoadInst *LI = B.CreateLoad(DstType, Dst);
     LI->setAlignment(Alignment);
     Value *OldValues = LI;
diff --git a/llvm/test/CodeGen/SPIRV/pointers/ptrcast-bitcast.ll b/llvm/test/CodeGen/SPIRV/pointers/ptrcast-bitcast.ll
new file mode 100644
index 0000000000000..d9374833887f0
--- /dev/null
+++ b/llvm/test/CodeGen/SPIRV/pointers/ptrcast-bitcast.ll
@@ -0,0 +1,29 @@
+; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv-unknown-vulkan-compute %s -o - | FileCheck %s --match-full-lines
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv-unknown-vulkan %s -o - -filetype=obj | spirv-val %}
+
+; CHECK-DAG:        %[[#uint:]] = OpTypeInt 32 0
+; CHECK-DAG:     %[[#v2_uint:]] = OpTypeVector %[[#uint]] 2
+; CHECK-DAG:      %[[#double:]] = OpTypeFloat 64
+; CHECK-DAG:   %[[#v2_double:]] = OpTypeVector %[[#double]] 2
+; CHECK-DAG:     %[[#v4_uint:]] = OpTypeVector %[[#uint]] 4
+@.str = private unnamed_addr constant [3 x i8] c"In\00", align 1
+@.str.2 = private unnamed_addr constant [4 x i8] c"Out\00", align 1
+
+
+define void @main() local_unnamed_addr #0 {
+entry:
+  %0 = tail call target("spirv.VulkanBuffer", [0 x <2 x i32>], 12, 0) @llvm.spv.resource.handlefrombinding.tspirv.VulkanBuffer_a0v2i32_12_0t(i32 0, i32 0, i32 1, i32 0, ptr nonnull @.str)
+  %1 = tail call target("spirv.VulkanBuffer", [0 x <2 x double>], 12, 1) @llvm.spv.resource.handlefrombinding.tspirv.VulkanBuffer_a0v2f64_12_1t(i32 0, i32 2, i32 1, i32 0, ptr nonnull @.str.2)
+  %2 = tail call noundef align 8 dereferenceable(8) ptr addrspace(11) @llvm.spv.resource.getpointer.p11.tspirv.VulkanBuffer_a0v2i32_12_0t(target("spirv.VulkanBuffer", [0 x <2 x i32>], 12, 0) %0, i32 0)
+  %3 = load <2 x i32>, ptr addrspace(11) %2, align 8
+  %4 = tail call noundef align 8 dereferenceable(8) ptr addrspace(11) @llvm.spv.resource.getpointer.p11.tspirv.VulkanBuffer_a0v2i32_12_0t(target("spirv.VulkanBuffer", [0 x <2 x i32>], 12, 0) %0, i32 1)
+  %5 = load <2 x i32>, ptr addrspace(11) %4, align 8
+; CHECK: %[[#tmp:]] = OpVectorShuffle %[[#v4_uint]] {{%[0-9]+}} {{%[0-9]+}} 0 2 1 3
+  %6 = shufflevector <2 x i32> %3, <2 x i32> %5, <4 x i32> <i32 0, i32 2, i32 1, i32 3>
+; CHECK: %[[#access:]] = OpAccessChain {{.*}}
+  %7 = tail call noundef align 16 dereferenceable(16) ptr addrspace(11) @llvm.spv.resource.getpointer.p11.tspirv.VulkanBuffer_a0v2f64_12_1t(target("spirv.VulkanBuffer", [0 x <2 x double>], 12, 1) %1, i32 0)
+; CHECK: %[[#bitcast:]] = OpBitcast %[[#v2_double]] %[[#tmp]]
+; CHECK: OpStore %[[#access]] %[[#bitcast]] Aligned 16
+  store <4 x i32> %6, ptr addrspace(11) %7, align 16
+  ret void
+}

[luciechoi]

The final SPIRV for this function had lots of verbose OpCopyObject, so I left some variables unmatched/ignored.

        %36 = OpFunction %2 DontInline %3       ; -- Begin function main
        %1 = OpLabel
        %37 = OpVariable %23 Function %32
        %38 = OpVariable %22 Function %33
        %39 = OpCopyObject %16 %34
        %40 = OpCopyObject %13 %35
        %41 = OpAccessChain %10 %39 %24 %24
        %42 = OpLoad %9 %41 Aligned 8
        %43 = OpAccessChain %10 %39 %24 %25
        %44 = OpLoad %9 %43 Aligned 8
        %45 = OpVectorShuffle %8 %42 %44 0 2 1 3
        %46 = OpAccessChain %6 %40 %24 %24
        %47 = OpBitcast %5 %45
        %48 = OpCopyObject %13 %35
        OpStore %46 %47 Aligned 16
        OpReturn
        OpFunctionEnd
                                        ; -- End function

[luciechoi]

@s-perron