[BOLT][AArch64] Provide createDummyReturnFunction (#96626)

AArch64 needs this function when instrumenting statically-linked binaries.

Sample commands:
```bash
clang -Wl,-q test.c -static -o out
llvm-bolt -instrument -instrumentation-sleep-time=5 out -o out.instr
```

3 files changed, 34 insertions, 8 deletions

diff --git a/bolt/include/bolt/Core/MCPlusBuilder.h b/bolt/include/bolt/Core/MCPlusBuilder.h
index ab07f07..885d627 100644
--- a/bolt/include/bolt/Core/MCPlusBuilder.h
+++ b/bolt/include/bolt/Core/MCPlusBuilder.h
@@ -2041,9 +2041,13 @@ public:
     return InstructionListType();
   }
 
+  /// Returns a function body that contains only a return instruction. An
+  /// example usage is a workaround for the '__bolt_fini_trampoline' of
+  // Instrumentation.
   virtual InstructionListType createDummyReturnFunction(MCContext *Ctx) const {
-    llvm_unreachable("not implemented");
-    return InstructionListType();
+    InstructionListType Insts(1);
+    createReturn(Insts[0]);
+    return Insts;
   }
 
   /// This method takes an indirect call instruction and splits it up into an
diff --git a/bolt/lib/Target/X86/X86MCPlusBuilder.cpp b/bolt/lib/Target/X86/X86MCPlusBuilder.cpp
index 37136f4..e46c425 100644
--- a/bolt/lib/Target/X86/X86MCPlusBuilder.cpp
+++ b/bolt/lib/Target/X86/X86MCPlusBuilder.cpp
@@ -3241,12 +3241,6 @@ public:
     return Insts;
   }
 
-  InstructionListType createDummyReturnFunction(MCContext *Ctx) const override {
-    InstructionListType Insts(1);
-    createReturn(Insts[0]);
-    return Insts;
-  }
-
   BlocksVectorTy indirectCallPromotion(
       const MCInst &CallInst,
       const std::vector<std::pair<MCSymbol *, uint64_t>> &Targets,
diff --git a/bolt/test/AArch64/dummy-return.s b/bolt/test/AArch64/dummy-return.s
new file mode 100644
index 0000000..a446343
--- /dev/null
+++ b/bolt/test/AArch64/dummy-return.s
@@ -0,0 +1,28 @@
+# REQUIRES: system-linux,target=aarch64{{.*}}
+
+# RUN: llvm-mc -filetype=obj -triple aarch64-unknown-unknown %s -o %t.o
+# RUN: %clang %cflags %t.o -o %t.exe -Wl,-q -static
+# RUN: llvm-bolt -instrument -instrumentation-sleep-time=1 %t.exe \
+# RUN:  -o %t.instr 2>&1 | FileCheck %s
+# RUN: llvm-objdump --disassemble-symbols=__bolt_fini_trampoline %t.instr -D \
+# RUN:  | FileCheck %s -check-prefix=CHECK-ASM
+
+# CHECK: BOLT-INFO: output linked against instrumentation runtime library
+# CHECK-ASM: <__bolt_fini_trampoline>:
+# CHECK-ASM-NEXT: ret
+
+  .text
+  .align 4
+  .global _start
+  .type _start, %function
+_start:
+   bl foo
+   ret
+  .size _start, .-_start
+
+  .global foo
+  .type foo, %function
+foo:
+  mov	w0, wzr
+  ret
+  .size foo, .-foo