[clang] [HLSL] Array by-value assignment (PR #109323)

[Chris B via cfe-commits]

================
@@ -0,0 +1,113 @@
+// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.3-library -emit-llvm -disable-llvm-passes -o - %s | FileCheck %s --enable-var-scope
+
+// CHECK-LABEL: define void {{.*}}arr_assign1
+// CHECK: [[Arr:%.*]] = alloca [2 x i32], align 4
+// CHECK: [[Arr2:%.*]] = alloca [2 x i32], align 4
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr]], ptr align 4 {{@.*}}, i32 8, i1 false)
+// CHECK: call void @llvm.memset.p0.i32(ptr align 4 [[Arr2]], i8 0, i32 8, i1 false)
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr]], ptr align 4 [[Arr2]], i32 8, i1 false)
+// CHECK: ret void
+void arr_assign1() {
+  int Arr[2] = {0, 1};
+  int Arr2[2] = {0, 0};
+  Arr = Arr2;
+}
+
+// CHECK-LABEL: define void {{.*}}arr_assign2
+// CHECK: [[Arr:%.*]] = alloca [2 x i32], align 4
+// CHECK: [[Arr2:%.*]] = alloca [2 x i32], align 4
+// CHECK: [[Arr3:%.*]] = alloca [2 x i32], align 4
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr]], ptr align 4 {{@.*}}, i32 8, i1 false)
+// CHECK: call void @llvm.memset.p0.i32(ptr align 4 [[Arr2]], i8 0, i32 8, i1 false)
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr3]], ptr align 4 {{@.*}}, i32 8, i1 false)
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr2]], ptr align 4 [[Arr3]], i32 8, i1 false)
+// CHECK: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[Arr]], ptr align 4 [[Arr2]], i32 8, i1 false)
+// CHECK: ret void
----------------
llvm-beanz wrote:

We should either have a `CHECK-NEXT` or `CHECK-NOT` to prevent this from succeeding if say an extra memcpy gets emitted, and maybe a `CHECK-NOT: alloca` after the last expected allocation to ensure we don't allocate any temporaries.

https://github.com/llvm/llvm-project/pull/109323