[Compiler] Fix branch targets in disassembly (#379)

* fix diassembly

* update changelog

* fix exit crash

* doc

doc/changelog.md

@@ -134,4 +134,6 @@
 - Added support for creating a static bitfield where fields may not be known at compile time.
 - Added support for `(size-of <typename>)`
 - Fixed a bug where creating a stack array of 0 sized caused a compiler assertion. It is now a normal compiler error.
-- Fixed a bug where the repl history was not loaded on compiler start
+- Fixed a bug where the repl history was not loaded on compiler start
+- Branch target addresses in the disassembly generated by `md`, `asm-file` with `:disassemble` and `(declare (print-asm))` are now correct
+- Fixed a segfault when the `goal-library.gc` contains an `(exit)`

goalc/compiler/CodeGenerator.cpp

@@ -45,7 +45,7 @@ std::vector<u8> CodeGenerator::run(const TypeSystem* ts) {
     do_function(m_fe->functions().at(i).get(), i);
   }
 
-  // generate a v3 object. TODO - support for v4 "data" objects.
+  // generate a v3 object.
   return m_gen.generate_data_v3(ts).to_vector();
 }
 

goalc/compiler/compilation/CompilerControl.cpp

@@ -33,7 +33,9 @@ Val* Compiler::compile_exit(const goos::Object& form, const goos::Object& rest,
   }
   // flag for the REPL.
   m_want_exit = true;
-  m_repl->save_history();
+  if (m_repl) {
+    m_repl->save_history();
+  }
   return get_none();
 }
 

goalc/debugger/DebugInfo.cpp

@@ -1,5 +1,4 @@
 #include <utility>
-#include <vector>
 #include "DebugInfo.h"
 #include "third-party/fmt/core.h"
 
@@ -7,17 +6,8 @@ DebugInfo::DebugInfo(std::string obj_name) : m_obj_name(std::move(obj_name)) {}
 
 std::string FunctionDebugInfo::disassemble_debug_info(bool* had_failure) {
   std::string result = fmt::format("[{}]\n", name);
-  std::vector<u8> data;
-  u8 temp[128];
-  for (const auto& x : instructions) {
-    auto count = x.instruction.emit(temp);
-    for (int i = 0; i < count; i++) {
-      data.push_back(temp[i]);
-    }
-  }
-
-  result += disassemble_x86_function(data.data(), data.size(), 0x10000, 0x10000, instructions, irs,
-                                     had_failure);
+  result += disassemble_x86_function(generated_code.data(), generated_code.size(), 0x10000, 0x10000,
+                                     instructions, irs, had_failure);
 
   return result;
 }

goalc/debugger/DebugInfo.h

@@ -8,6 +8,11 @@
 #include "goalc/emitter/Instruction.h"
 #include "goalc/debugger/disassemble.h"
 
+/*!
+ * FunctionDebugInfo stores per-function debugging information.
+ * For now, it is pretty basic, but it will eventually contain stuff like stack frame info
+ * and which var is in each register at each instruction.
+ */
 struct FunctionDebugInfo {
   u32 offset_in_seg;  // not including type tag.
   u32 length;
@@ -15,7 +20,10 @@ struct FunctionDebugInfo {
   std::string name;
 
   std::vector<std::string> irs;
-  std::vector<InstructionInfo> instructions;
+  std::vector<InstructionInfo> instructions;  // contains mapping to IRs
+
+  // the actual bytes in the object file.
+  std::vector<u8> generated_code;
 
   std::string disassemble_debug_info(bool* had_failure);
 };

goalc/emitter/ObjectGenerator.cpp

@@ -95,13 +95,21 @@ ObjectFileData ObjectGenerator::generate_data_v3(const TypeSystem* ts) {
     handle_temp_static_ptr_links(seg);
   }
 
-  // actual linking?
+  // step 4, generate the link table
   for (int seg = N_SEG; seg-- > 0;) {
     emit_link_table(seg, ts);
   }
 
-  // emit header
+  // step 4.5, collect final result of code/object generation for compiler debugging disassembly
+  for (int seg = 0; seg < N_SEG; seg++) {
+    for (auto& function : m_function_data_by_seg.at(seg)) {
+      auto start = m_data_by_seg.at(seg).begin() + function.instruction_to_byte_in_data.at(0);
+      auto end = start + function.debug->length;
+      function.debug->generated_code = {start, end};
+    }
+  }
 
+  // step 5, build header and combine sections
   out.header = generate_header_v3();
   out.segment_data = std::move(m_data_by_seg);
   out.link_tables = std::move(m_link_by_seg);