interp: mark pointers in aggregate call operands as external

interp/testdata/aggregate.ll

@@ -0,0 +1,27 @@
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64--linux"
+
+declare void @externalAggregate({ ptr })
+
+@main.value = global i32 1
+@main.result = global i32 0
+
+define void @runtime.initAll() unnamed_addr {
+entry:
+  call void @main.init(ptr undef)
+  ret void
+}
+
+define internal void @main.init(ptr %context) unnamed_addr {
+entry:
+  ; The pointer is hidden inside an aggregate argument.
+  %arg = insertvalue { ptr } undef, ptr @main.value, 0
+
+  ; This call runs at runtime and may modify @main.value.
+  call void @externalAggregate({ ptr } %arg)
+
+  ; Therefore this load must also remain at runtime.
+  %value = load i32, ptr @main.value
+  store i32 %value, ptr @main.result
+  ret void
+}

interp/testdata/aggregate.out.ll

@@ -0,0 +1,15 @@
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64--linux"
+
+@main.value = global i32 1
+@main.result = local_unnamed_addr global i32 0
+
+declare void @externalAggregate({ ptr }) local_unnamed_addr
+
+define void @runtime.initAll() unnamed_addr {
+entry:
+  call void @externalAggregate({ ptr } { ptr @main.value })
+  %value = load i32, ptr @main.value, align 4
+  store i32 %value, ptr @main.result, align 4
+  ret void
+}