[ELF] Rework debug line parsing to use llvm::Error (LLD-side)

[jhenderson]

Summary

This is a companion change to D44382. That change changes the debug line parser interface to report LLVM errors, and to improve the error/warning reporting mechanism to not use fprintf(stderr, ...).

As noted in the description on D44382, these warnings were not using the errs() stream, and so were not always being flushed properly by LLD on shutdown. To test this, I have extended the bad-debug undefined symbol message case to show that a corresponding warning is printed, if the debug line cannot be parsed.

[dblaikie]

(I will say "LTOrErr" is a slightly confusing variable name, given the
prevalence of "LTO" as a known acronym... )

[ruiu]

ELF/InputFiles.cpp

124–135

I agree with David; LTOrError is a confusing variable name.

Can you avoid using auto here?

[jhenderson]

ELF/InputFiles.cpp

124–135

It's also wrong, since it could be a LineTable and an Error! I'll fix that shortly (note: this may undergo some changes, as there's some discussion on the main review about the appropriate way of reporting the different levels of error).

[jhenderson]

Fixed use of auto and variable name. Also fixed a bug which I'd accidentally introduced - my changes to getOrParseLineTable meant that it reported an error, if an invalid offset was specified. This meant that an object with an empty or missing .debug_line section would result in a warning printed from this function whenever LLD tried to look up the debug information for reporting its own message. I've added a couple of tests to check this going forwards.

[jhenderson]

I've abandoned D44382, as I prefer its alternative D44560, so I'm also abandoning this one. See D44562 for the alternative.

ELF/InputFiles.cpp

 
 template <class ELFT> void ObjFile<ELFT>::initializeDwarf() {
   DWARFContext Dwarf(make_unique<LLDDwarfObj<ELFT>>(this));
   const DWARFObject &Obj = Dwarf.getDWARFObj();
   DwarfLine.reset(new DWARFDebugLine);
   DWARFDataExtractor LineData(Obj, Obj.getLineSection(), Config->IsLE,
                               Config->Wordsize);
 
+  const DWARFDebugLine::LineTable *LT = nullptr;
+  if (LineData.isValidOffset(0)) {
     // The second parameter is offset in .debug_line section
     // for compilation unit (CU) of interest. We have only one
     // CU (object file), so offset is always 0.
-  const DWARFDebugLine::LineTable *LT =
+    DWARFDebugLine::LineTableAndError LineTableAndErr =
         DwarfLine->getOrParseLineTable(LineData, 0, Dwarf, nullptr);
+    LT = LineTableAndErr.Table;
+    handleAllErrors(std::move(LineTableAndErr.Err),
+                    [&](DebugLineError &Err) { warn(Err.message()); },
+                    [&](ErrorInfoBase &Err) { error(Err.message()); });
+  }
   if (!LT)
     return;
 
   // Return if there is no debug information about CU available.
   if (!Dwarf.getNumCompileUnits())
     return;
 
   // Loop over variable records and insert them to VariableLoc.

test/ELF/Inputs/undef-bad-debug.s

 .section .text,"ax"
 sym:
     .quad zed6
-    
+.section .debug_line,"",@progbits
+    .long .Lunit_end - .Lunit_start # unit length
+.Lunit_start:
+    .short 4                        # version
+    .long .Lprologue_end - .Lprologue_start # prologue length
+.Lprologue_start:
+    .byte 1                         # minimum instruction length
+    .byte 1                         # maximum operatiosn per instruction
+    .byte 1                         # default is_stmt
+    .byte -5                        # line base
+    .byte 14                        # line range
+    .byte 13                        # opcode base
+    .byte 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1 # standard opcode lengths
+    .asciz "dir"                    # include directories
+    .byte 0
+    .asciz "undef-bad-debug.s"      # file names
+    .byte 1, 0, 0
+    .byte 0
+    .byte 0                         # extraneous byte
+.Lprologue_end:
+    .byte 0, 9, 2                   # DW_LNE_set_address
+    .quad sym
+    .byte 3                         # DW_LNS_advance_line
+    .byte 10
+    .byte 1                         # DW_LNS_copy
+    .byte 2                         # DW_LNS_advance_pc
+    .byte 8
+    .byte 0, 1, 1                   # DW_LNE_end_sequence
+.Lunit_end:
+
 .section .debug_info,"",@progbits
     .long   .Lcu_end - .Lcu_start   # Length of Unit
 .Lcu_start:
     .short  4                       # DWARF version number
     .long   .Lsection_abbrev        # Offset Into Abbrev. Section
     .byte   8                       # Address Size (in bytes)
     .byte   1                       # Abbrev [1] 0xb:0x79 DW_TAG_compile_unit
     .byte   2                       # Abbrev [2] 0x2a:0x15 DW_TAG_variable

test/ELF/no-line-parser-errors-if-empty-section.s

+# REQUIRES: x86
+
+# LLD uses the debug data to get information for error messages, if possible.
+# However, if the debug line section is empty, we should not attempt to parse
+# it, as that would result in errors from the parser.
+
+# RUN: llvm-mc -filetype=obj -triple=x86_64-pc-linux %s -o %t.o
+# RUN: not ld.lld %t.o -o %t.elf 2>&1 | FileCheck %s
+
+# CHECK-NOT: warning:
+# CHECK-NOT: error:
+# CHECK: error: undefined symbol: undefined
+# CHECK-NEXT: {{.*}}.o:(.text+0x1)
+# CHECK-NOT: warning:
+# CHECK-NOT: error:
+
+.globl _start
+_start:
+    callq undefined
+
+.section .debug_line,"",@progbits

test/ELF/no-line-parser-errors-if-no-section.s

+# REQUIRES: x86
+
+# LLD uses the debug data to get information for error messages, if possible.
+# However, if there is no debug line section, we should not attempt to parse
+# it, as that would result in errors from the parser.
+
+# RUN: llvm-mc -filetype=obj -triple=x86_64-pc-linux %s -o %t.o
+# RUN: not ld.lld %t.o -o %t.elf 2>&1 | FileCheck %s
+
+# CHECK-NOT: warning:
+# CHECK-NOT: error:
+# CHECK: error: undefined symbol: undefined
+# CHECK-NEXT: {{.*}}.o:(.text+0x1)
+# CHECK-NOT: warning:
+# CHECK-NOT: error:
+
+.globl _start
+_start:
+    callq undefined

test/ELF/undef.s

 # CHECK: error: undefined symbol: zed4
 # CHECK: >>> referenced by undef-debug.s:7 (dir{{/|\\}}undef-debug.s:7)
 # CHECK: >>>               {{.*}}.o:(.text.1+0x0)
 
 # CHECK: error: undefined symbol: zed5
 # CHECK: >>> referenced by undef-debug.s:11 (dir{{/|\\}}undef-debug.s:11)
 # CHECK: >>>               {{.*}}.o:(.text.2+0x0)
 
+# CHECK: warning: parsing line table prologue at 0x00000000 should have ended at 0x00000038 but it ended at 0x00000037
 # CHECK: error: undefined symbol: zed6
 # CHECK: >>> referenced by {{.*}}tmp4.o:(.text+0x0)
 
 # RUN: not ld.lld %t.o %t2.a -o %t.exe -no-demangle 2>&1 | \
 # RUN:   FileCheck -check-prefix=NO-DEMANGLE %s
 # NO-DEMANGLE: error: undefined symbol: _Z3fooi
 
 .file "undef.s"
 
   .globl _start
 _start:
   call foo
   call bar
   call zed1
   call _Z3fooi