Fix parsing of dwarfdump output to be more resilient (#20777)

To generate source maps, we run llvm-dwarfdump against wasm file and
parse its plain-text output heavily relied on regex matches.

Previously, in order to retrieve comp dir from debug_info section, we
search regex patterns assuming that the output has entries DW_AT_stmt_list
and DW_AT_comp_dir consecutively in this order, but I noticed that
assumption doesn't hold in some cases. With this change, it can parse
comp_dir regardless of the order of DW_* entries.

Author: kon72

AUTHORS

@@ -596,3 +596,4 @@ a license to everyone to use it as detailed in LICENSE.)
 * 郑苏波 (Super Zheng) <superzheng@tencent.com>
 * James Hu <jameshu2022@gmail.com>
 * Jerry Zhuang <jerry.zhuang@jwzg.com>
+* Taisei Kon <kinsei0916@gmail.com>

test/other/wasm_sourcemap_extract_comp_dir_map/bar.wasm.dump

@@ -0,0 +1,71 @@
+bar.wasm:	file format WASM
+
+.debug_info contents:
+0x00000000: Compile Unit: length = 0x00000129, format = DWARF32, version = 0x0004, abbr_offset = 0x0000, addr_size = 0x04 (next unit at 0x0000012d)
+
+0x0000000b: DW_TAG_compile_unit
+              DW_AT_stmt_list	(0x00000000)
+              DW_AT_ranges	(0x00000000
+                 [0x00000002, 0x0000000a)
+                 [0x0000000b, 0x00000015)
+                 [0x00000016, 0x00000030)
+                 [0x00000031, 0x00000039))
+              DW_AT_name	("system/lib/compiler-rt/stack_ops.S")
+              DW_AT_comp_dir	("/emsdk/emscripten")
+              DW_AT_producer	("clang version 18.0.0 (https://github.com/llvm/llvm-project 269685545e439ad050b67740533c59f965cae955)")
+              DW_AT_language	(DW_LANG_Mips_Assembler)
+
+.debug_line contents:
+debug_line[0x00000000]
+Line table prologue:
+    total_length: 0x00000091
+          format: DWARF32
+         version: 4
+ prologue_length: 0x00000043
+ min_inst_length: 1
+max_ops_per_inst: 1
+ default_is_stmt: 1
+       line_base: -5
+      line_range: 14
+     opcode_base: 13
+standard_opcode_lengths[DW_LNS_copy] = 0
+standard_opcode_lengths[DW_LNS_advance_pc] = 1
+standard_opcode_lengths[DW_LNS_advance_line] = 1
+standard_opcode_lengths[DW_LNS_set_file] = 1
+standard_opcode_lengths[DW_LNS_set_column] = 1
+standard_opcode_lengths[DW_LNS_negate_stmt] = 0
+standard_opcode_lengths[DW_LNS_set_basic_block] = 0
+standard_opcode_lengths[DW_LNS_const_add_pc] = 0
+standard_opcode_lengths[DW_LNS_fixed_advance_pc] = 1
+standard_opcode_lengths[DW_LNS_set_prologue_end] = 0
+standard_opcode_lengths[DW_LNS_set_epilogue_begin] = 0
+standard_opcode_lengths[DW_LNS_set_isa] = 1
+include_directories[  1] = "system/lib/compiler-rt"
+file_names[  1]:
+           name: "stack_ops.S"
+      dir_index: 1
+       mod_time: 0x00000000
+         length: 0x00000000
+
+Address            Line   Column File   ISA Discriminator OpIndex Flags
+------------------ ------ ------ ------ --- ------------- ------- -------------
+0x0000000000000002     18      0      1   0             0       0  is_stmt
+0x0000000000000009     19      0      1   0             0       0  is_stmt
+0x000000000000000a     19      0      1   0             0       0  is_stmt end_sequence
+0x000000000000000b     23      0      1   0             0       0  is_stmt
+0x000000000000000e     24      0      1   0             0       0  is_stmt
+0x0000000000000014     25      0      1   0             0       0  is_stmt
+0x0000000000000015     25      0      1   0             0       0  is_stmt end_sequence
+0x0000000000000019     30      0      1   0             0       0  is_stmt
+0x000000000000001f     32      0      1   0             0       0  is_stmt
+0x0000000000000021     34      0      1   0             0       0  is_stmt
+0x0000000000000022     36      0      1   0             0       0  is_stmt
+0x0000000000000024     37      0      1   0             0       0  is_stmt
+0x0000000000000025     38      0      1   0             0       0  is_stmt
+0x0000000000000027     39      0      1   0             0       0  is_stmt
+0x000000000000002d     40      0      1   0             0       0  is_stmt
+0x000000000000002f     41      0      1   0             0       0  is_stmt
+0x0000000000000030     41      0      1   0             0       0  is_stmt end_sequence
+0x0000000000000031     45      0      1   0             0       0  is_stmt
+0x0000000000000038     46      0      1   0             0       0  is_stmt
+0x0000000000000039     46      0      1   0             0       0  is_stmt end_sequence

test/other/wasm_sourcemap_extract_comp_dir_map/foo.wasm.dump

@@ -0,0 +1,51 @@
+foo.wasm:	file format WASM
+
+.debug_info contents:
+0x00000000: Compile Unit: length = 0x0000003a, format = DWARF32, version = 0x0004, abbr_offset = 0x0000, addr_size = 0x04 (next unit at 0x0000003e)
+
+0x0000000b: DW_TAG_compile_unit
+              DW_AT_producer	("clang version 18.0.0 (https://github.com/llvm/llvm-project 269685545e439ad050b67740533c59f965cae955)")
+              DW_AT_language	(DW_LANG_C11)
+              DW_AT_name	("system/lib/compiler-rt/__trap.c")
+              DW_AT_stmt_list	(0x00000000)
+              DW_AT_comp_dir	("/emsdk/emscripten")
+              DW_AT_low_pc	(0x00000002)
+              DW_AT_high_pc	(0x00000006)
+
+.debug_line contents:
+debug_line[0x00000000]
+Line table prologue:
+    total_length: 0x00000059
+          format: DWARF32
+         version: 4
+ prologue_length: 0x00000040
+ min_inst_length: 1
+max_ops_per_inst: 1
+ default_is_stmt: 1
+       line_base: -5
+      line_range: 14
+     opcode_base: 13
+standard_opcode_lengths[DW_LNS_copy] = 0
+standard_opcode_lengths[DW_LNS_advance_pc] = 1
+standard_opcode_lengths[DW_LNS_advance_line] = 1
+standard_opcode_lengths[DW_LNS_set_file] = 1
+standard_opcode_lengths[DW_LNS_set_column] = 1
+standard_opcode_lengths[DW_LNS_negate_stmt] = 0
+standard_opcode_lengths[DW_LNS_set_basic_block] = 0
+standard_opcode_lengths[DW_LNS_const_add_pc] = 0
+standard_opcode_lengths[DW_LNS_fixed_advance_pc] = 1
+standard_opcode_lengths[DW_LNS_set_prologue_end] = 0
+standard_opcode_lengths[DW_LNS_set_epilogue_begin] = 0
+standard_opcode_lengths[DW_LNS_set_isa] = 1
+include_directories[  1] = "system/lib/compiler-rt"
+file_names[  1]:
+           name: "__trap.c"
+      dir_index: 1
+       mod_time: 0x00000000
+         length: 0x00000000
+
+Address            Line   Column File   ISA Discriminator OpIndex Flags
+------------------ ------ ------ ------ --- ------------- ------- -------------
+0x0000000000000003      2      3      1   0             0       0  is_stmt prologue_end
+0x0000000000000004      3      1      1   0             0       0  is_stmt
+0x0000000000000006      3      1      1   0             0       0  is_stmt end_sequence

test/test_other.py

@@ -9,6 +9,7 @@
 from functools import wraps
 import glob
 import gzip
+import importlib
 import itertools
 import json
 import os
@@ -9589,6 +9590,24 @@ def test(infile, source_map_added_dir=''):
     ensure_dir('inner')
     test('inner/a.cpp', 'inner')
 
+  def test_wasm_sourcemap_extract_comp_dir_map(self):
+    wasm_sourcemap = importlib.import_module('tools.wasm-sourcemap')
+
+    def test(dump_file):
+      dwarfdump_output = read_file(
+          test_file(
+              os.path.join('other/wasm_sourcemap_extract_comp_dir_map',
+                           dump_file)))
+      map_stmt_list_to_comp_dir = wasm_sourcemap.extract_comp_dir_map(
+          dwarfdump_output)
+      self.assertEqual(map_stmt_list_to_comp_dir,
+                       {'0x00000000': '/emsdk/emscripten'})
+
+    # Make sure we can extract the compilation directories no matter what the
+    # order of `DW_AT_*` attributes is.
+    test('foo.wasm.dump')
+    test('bar.wasm.dump')
+
   def test_emsymbolizer(self):
     def check_dwarf_loc_info(address, funcs, locs):
       out = self.run_process(

tools/wasm-sourcemap.py

@@ -178,6 +178,19 @@ def remove_dead_entries(entries):
     block_start = cur_entry
 
 
+def extract_comp_dir_map(text):
+  map_stmt_list_to_comp_dir = {}
+  chunks = re.split(r"0x[0-9a-f]*: DW_TAG_compile_unit", text)
+  for chunk in chunks[1:]:
+    stmt_list_match = re.search(r"DW_AT_stmt_list\s+\((0x[0-9a-f]*)\)", chunk)
+    if stmt_list_match is not None:
+      stmt_list = stmt_list_match.group(1)
+      comp_dir_match = re.search(r"DW_AT_comp_dir\s+\(\"([^\"]+)\"\)", chunk)
+      comp_dir = comp_dir_match.group(1) if comp_dir_match is not None else ''
+      map_stmt_list_to_comp_dir[stmt_list] = comp_dir
+  return map_stmt_list_to_comp_dir
+
+
 def read_dwarf_entries(wasm, options):
   if options.dwarfdump_output:
     output = Path(options.dwarfdump_output).read_bytes()
@@ -198,14 +211,9 @@ def read_dwarf_entries(wasm, options):
 
   entries = []
   debug_line_chunks = re.split(r"debug_line\[(0x[0-9a-f]*)\]", output.decode('utf-8'))
-  maybe_debug_info_content = debug_line_chunks[0]
-  for i in range(1, len(debug_line_chunks), 2):
-    stmt_list = debug_line_chunks[i]
-    comp_dir_match = re.search(r"DW_AT_stmt_list\s+\(" + stmt_list + r"\)\s+" +
-                               r"DW_AT_comp_dir\s+\(\"([^\"]+)", maybe_debug_info_content)
-    comp_dir = comp_dir_match.group(1) if comp_dir_match is not None else ""
-
-    line_chunk = debug_line_chunks[i + 1]
+  map_stmt_list_to_comp_dir = extract_comp_dir_map(debug_line_chunks[0])
+  for stmt_list, line_chunk in zip(debug_line_chunks[1::2], debug_line_chunks[2::2]):
+    comp_dir = map_stmt_list_to_comp_dir.get(stmt_list, '')
 
     # include_directories[  1] = "/Users/yury/Work/junk/sqlite-playground/src"
     # file_names[  1]: