utils/update_cc_test_checks.py

   1 #!/usr/bin/env python3
   2 '''A utility to update LLVM IR CHECK lines in C/C++ FileCheck test files.
   3
   4 Example RUN lines in .c/.cc test files:
   5
   6 // RUN: %clang -emit-llvm -S %s -o - -O2 | FileCheck %s
   7 // RUN: %clangxx -emit-llvm -S %s -o - -O2 | FileCheck -check-prefix=CHECK-A %s
   8
   9 Usage:
  10
  11 % utils/update_cc_test_checks.py --llvm-bin=release/bin test/a.cc
  12 % utils/update_cc_test_checks.py --c-index-test=release/bin/c-index-test \
  13   --clang=release/bin/clang /tmp/c/a.cc
  14 '''
  15
  16 import argparse
  17 import collections
  18 import distutils.spawn
  19 import os
  20 import shlex
  21 import string
  22 import subprocess
  23 import sys
  24 import re
  25 import tempfile
  26
  27 from UpdateTestChecks import asm, common
  28
  29 ADVERT = '// NOTE: Assertions have been autogenerated by '
  30
  31 CHECK_RE = re.compile(r'^\s*//\s*([^:]+?)(?:-NEXT|-NOT|-DAG|-LABEL)?:')
  32 RUN_LINE_RE = re.compile('^//\s*RUN:\s*(.*)$')
  33
  34 SUBST = {
  35     '%clang': [],
  36     '%clang_cc1': ['-cc1'],
  37     '%clangxx': ['--driver-mode=g++'],
  38 }
  39
  40 def get_line2spell_and_mangled(args, clang_args):
  41   ret = {}
  42   with tempfile.NamedTemporaryFile() as f:
  43     # TODO Make c-index-test print mangled names without circumventing through precompiled headers
  44     status = subprocess.run([args.c_index_test, '-write-pch', f.name, *clang_args],
  45                             stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
  46     if status.returncode:
  47       sys.stderr.write(status.stdout.decode())
  48       sys.exit(2)
  49     output = subprocess.check_output([args.c_index_test,
  50         '-test-print-mangle', f.name])
  51     if sys.version_info[0] > 2:
  52       output = output.decode()
  53
  54   RE = re.compile(r'^FunctionDecl=(\w+):(\d+):\d+ \(Definition\) \[mangled=([^]]+)\]')
  55   for line in output.splitlines():
  56     m = RE.match(line)
  57     if not m: continue
  58     spell, line, mangled = m.groups()
  59     if mangled == '_' + spell:
  60       # HACK for MacOS (where the mangled name includes an _ for C but the IR won't):
  61       mangled = spell
  62     # Note -test-print-mangle does not print file names so if #include is used,
  63     # the line number may come from an included file.
  64     ret[int(line)-1] = (spell, mangled)
  65   if args.verbose:
  66     for line, func_name in sorted(ret.items()):
  67       print('line {}: found function {}'.format(line+1, func_name), file=sys.stderr)
  68   return ret
  69
  70
  71 def config():
  72   parser = argparse.ArgumentParser(
  73       description=__doc__,
  74       formatter_class=argparse.RawTextHelpFormatter)
  75   parser.add_argument('-v', '--verbose', action='store_true')
  76   parser.add_argument('--llvm-bin', help='llvm $prefix/bin path')
  77   parser.add_argument('--clang',
  78                       help='"clang" executable, defaults to $llvm_bin/clang')
  79   parser.add_argument('--clang-args',
  80                       help='Space-separated extra args to clang, e.g. --clang-args=-v')
  81   parser.add_argument('--c-index-test',
  82                       help='"c-index-test" executable, defaults to $llvm_bin/c-index-test')
  83   parser.add_argument(
  84       '--functions', nargs='+', help='A list of function name regexes. '
  85       'If specified, update CHECK lines for functions matching at least one regex')
  86   parser.add_argument(
  87       '--x86_extra_scrub', action='store_true',
  88       help='Use more regex for x86 matching to reduce diffs between various subtargets')
  89   parser.add_argument('-u', '--update-only', action='store_true',
  90                       help='Only update test if it was already autogened')
  91   parser.add_argument('tests', nargs='+')
  92   args = parser.parse_args()
  93   args.clang_args = shlex.split(args.clang_args or '')
  94
  95   if args.clang is None:
  96     if args.llvm_bin is None:
  97       args.clang = 'clang'
  98     else:
  99       args.clang = os.path.join(args.llvm_bin, 'clang')
 100   if not distutils.spawn.find_executable(args.clang):
 101     print('Please specify --llvm-bin or --clang', file=sys.stderr)
 102     sys.exit(1)
 103   if args.c_index_test is None:
 104     if args.llvm_bin is None:
 105       args.c_index_test = 'c-index-test'
 106     else:
 107       args.c_index_test = os.path.join(args.llvm_bin, 'c-index-test')
 108   if not distutils.spawn.find_executable(args.c_index_test):
 109     print('Please specify --llvm-bin or --c-index-test', file=sys.stderr)
 110     sys.exit(1)
 111
 112   return args
 113
 114
 115 def get_function_body(args, filename, clang_args, prefixes, triple_in_cmd, func_dict):
 116   # TODO Clean up duplication of asm/common build_function_body_dictionary
 117   # Invoke external tool and extract function bodies.
 118   raw_tool_output = common.invoke_tool(args.clang, clang_args, filename)
 119   if '-emit-llvm' in clang_args:
 120     common.build_function_body_dictionary(
 121             common.OPT_FUNCTION_RE, common.scrub_body, [],
 122             raw_tool_output, prefixes, func_dict, args.verbose)
 123   else:
 124     print('The clang command line should include -emit-llvm as asm tests '
 125           'are discouraged in Clang testsuite.', file=sys.stderr)
 126     sys.exit(1)
 127
 128
 129 def main():
 130   args = config()
 131   script_name = os.path.basename(__file__)
 132   autogenerated_note = (ADVERT + 'utils/' + script_name)
 133
 134   for filename in args.tests:
 135     with open(filename) as f:
 136       input_lines = [l.rstrip() for l in f]
 137
 138     first_line = input_lines[0] if input_lines else ""
 139     if 'autogenerated' in first_line and script_name not in first_line:
 140       common.warn("Skipping test which wasn't autogenerated by " + script_name, filename)
 141       continue
 142
 143     if args.update_only:
 144       if not first_line or 'autogenerated' not in first_line:
 145         common.warn("Skipping test which isn't autogenerated: " + filename)
 146         continue
 147
 148     # Extract RUN lines.
 149     raw_lines = [m.group(1)
 150                  for m in [RUN_LINE_RE.match(l) for l in input_lines] if m]
 151     run_lines = [raw_lines[0]] if len(raw_lines) > 0 else []
 152     for l in raw_lines[1:]:
 153       if run_lines[-1].endswith("\\"):
 154         run_lines[-1] = run_lines[-1].rstrip("\\") + " " + l
 155       else:
 156         run_lines.append(l)
 157
 158     if args.verbose:
 159       print('Found {} RUN lines:'.format(len(run_lines)), file=sys.stderr)
 160       for l in run_lines:
 161         print('  RUN: ' + l, file=sys.stderr)
 162
 163     # Build a list of clang command lines and check prefixes from RUN lines.
 164     run_list = []
 165     line2spell_and_mangled_list = collections.defaultdict(list)
 166     for l in run_lines:
 167       commands = [cmd.strip() for cmd in l.split('|', 1)]
 168
 169       triple_in_cmd = None
 170       m = common.TRIPLE_ARG_RE.search(commands[0])
 171       if m:
 172         triple_in_cmd = m.groups()[0]
 173
 174       # Apply %clang substitution rule, replace %s by `filename`, and append args.clang_args
 175       clang_args = shlex.split(commands[0])
 176       if clang_args[0] not in SUBST:
 177         print('WARNING: Skipping non-clang RUN line: ' + l, file=sys.stderr)
 178         continue
 179       clang_args[0:1] = SUBST[clang_args[0]]
 180       clang_args = [filename if i == '%s' else i for i in clang_args] + args.clang_args
 181
 182       # Extract -check-prefix in FileCheck args
 183       filecheck_cmd = commands[-1]
 184       common.verify_filecheck_prefixes(filecheck_cmd)
 185       if not filecheck_cmd.startswith('FileCheck '):
 186         print('WARNING: Skipping non-FileChecked RUN line: ' + l, file=sys.stderr)
 187         continue
 188       check_prefixes = [item for m in common.CHECK_PREFIX_RE.finditer(filecheck_cmd)
 189                                for item in m.group(1).split(',')]
 190       if not check_prefixes:
 191         check_prefixes = ['CHECK']
 192       run_list.append((check_prefixes, clang_args, triple_in_cmd))
 193
 194     # Strip CHECK lines which are in `prefix_set`, update test file.
 195     prefix_set = set([prefix for p in run_list for prefix in p[0]])
 196     input_lines = []
 197     with open(filename, 'r+') as f:
 198       for line in f:
 199         m = CHECK_RE.match(line)
 200         if not (m and m.group(1) in prefix_set) and line != '//\n':
 201           input_lines.append(line)
 202       f.seek(0)
 203       f.writelines(input_lines)
 204       f.truncate()
 205
 206     # Execute clang, generate LLVM IR, and extract functions.
 207     func_dict = {}
 208     for p in run_list:
 209       prefixes = p[0]
 210       for prefix in prefixes:
 211         func_dict.update({prefix: dict()})
 212     for prefixes, clang_args, triple_in_cmd in run_list:
 213       if args.verbose:
 214         print('Extracted clang cmd: clang {}'.format(clang_args), file=sys.stderr)
 215         print('Extracted FileCheck prefixes: {}'.format(prefixes), file=sys.stderr)
 216
 217       get_function_body(args, filename, clang_args, prefixes, triple_in_cmd, func_dict)
 218
 219       # Invoke c-index-test to get mapping from start lines to mangled names.
 220       # Forward all clang args for now.
 221       for k, v in get_line2spell_and_mangled(args, clang_args).items():
 222         line2spell_and_mangled_list[k].append(v)
 223
 224     output_lines = [autogenerated_note]
 225     for idx, line in enumerate(input_lines):
 226       # Discard any previous script advertising.
 227       if line.startswith(ADVERT):
 228         continue
 229       if idx in line2spell_and_mangled_list:
 230         added = set()
 231         for spell, mangled in line2spell_and_mangled_list[idx]:
 232           # One line may contain multiple function declarations.
 233           # Skip if the mangled name has been added before.
 234           # The line number may come from an included file,
 235           # we simply require the spelling name to appear on the line
 236           # to exclude functions from other files.
 237           if mangled in added or spell not in line:
 238             continue
 239           if args.functions is None or any(re.search(regex, spell) for regex in args.functions):
 240             if added:
 241               output_lines.append('//')
 242             added.add(mangled)
 243             common.add_ir_checks(output_lines, '//', run_list, func_dict, mangled)
 244       output_lines.append(line.rstrip('\n'))
 245
 246     # Update the test file.
 247     with open(filename, 'w') as f:
 248       for line in output_lines:
 249         f.write(line + '\n')
 250
 251   return 0
 252
 253
 254 if __name__ == '__main__':
 255   sys.exit(main())