[BOLT] Improve constant island disassembly

[maksfb]

• Add label that identifies constant island.
• Support cases where the island is located after the function.

[llvmbot]

@llvm/pr-subscribers-bolt

Author: Maksim Panchenko (maksfb)

Changes

• Add label that identifies constant island.
• Support cases where the island is located after the function.

---

Full diff: https://github.com/llvm/llvm-project/pull/127971.diff

2 Files Affected:

• (modified) bolt/lib/Core/BinaryFunction.cpp (+6-1)
• (modified) bolt/test/AArch64/data-in-code.s (+8-2)

diff --git a/bolt/lib/Core/BinaryFunction.cpp b/bolt/lib/Core/BinaryFunction.cpp
index bc45caf3ec8b7..1e427b2df11cf 100644
--- a/bolt/lib/Core/BinaryFunction.cpp
+++ b/bolt/lib/Core/BinaryFunction.cpp
@@ -498,6 +498,11 @@ void BinaryFunction::print(raw_ostream &OS, std::string Annotation) {
     if (!IslandOffset)
       return;
 
+    // Print label if it exists at this offset.
+    if (const BinaryData *BD =
+            BC.getBinaryDataAtAddress(getAddress() + *IslandOffset))
+      OS << BD->getName() << ":\n";
+
     const size_t IslandSize = getSizeOfDataInCodeAt(*IslandOffset);
     BC.printData(OS, BC.extractData(getAddress() + *IslandOffset, IslandSize),
                  *IslandOffset);
@@ -1066,7 +1071,7 @@ size_t BinaryFunction::getSizeOfDataInCodeAt(uint64_t Offset) const {
   auto Iter = Islands->CodeOffsets.upper_bound(Offset);
   if (Iter != Islands->CodeOffsets.end())
     return *Iter - Offset;
-  return getSize() - Offset;
+  return getMaxSize() - Offset;
 }
 
 std::optional<uint64_t>
diff --git a/bolt/test/AArch64/data-in-code.s b/bolt/test/AArch64/data-in-code.s
index 8d3179a0c3350..1df5d4568542f 100644
--- a/bolt/test/AArch64/data-in-code.s
+++ b/bolt/test/AArch64/data-in-code.s
@@ -7,7 +7,8 @@
 ## Check disassembly of BOLT input.
 # RUN: llvm-objdump %t.exe -d | FileCheck %s
 
-# RUN: llvm-bolt %t.exe -o %t.bolt --print-disasm | FileCheck %s
+# RUN: llvm-bolt %t.exe -o %t.bolt --print-disasm \
+# RUN:   | FileCheck %s --check-prefixes CHECK,CHECK-BOLT-ONLY
 
 .text
 .balign 4
@@ -16,16 +17,21 @@
 .type _start, %function
 _start:
   mov x0, #0x0
+  ldr x1, .L1
   .word 0x4f82e010
   ret
+.size _start, .-_start
+.L1:
   .byte 0x0, 0xff, 0x42
 # CHECK-LABEL: _start
 # CHECK:        mov x0, #0x0
+# CHECK-NEXT:   ldr x1
+# CHECK-BOLT-ONLY-SAME: ISLANDat[[ADDR:]]
 # CHECK-NEXT:   .word 0x4f82e010
 # CHECK-NEXT:   ret
+# CHECK-BOLT-ONLY-NEXT: ISLANDat[[ADDR]]
 # CHECK-NEXT:   .short 0xff00
 # CHECK-NEXT:   .byte 0x42
-.size _start, .-_start
 
 ## Force relocation mode.
   .reloc 0, R_AARCH64_NONE

[dcci]

LGTM