1#!/usr/bin/env python3
2'''A utility to update LLVM IR CHECK lines in C/C++ FileCheck test files.
3
4Example 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
9Usage:
10
11% utils/update_cc_test_checks.py --llvm-bin=release/bin test/a.cc
12% utils/update_cc_test_checks.py --clang=release/bin/clang /tmp/c/a.cc
13'''
14
15from __future__ import print_function
16
17import argparse
18import collections
19import distutils.spawn
20import json
21import os
22import re
23import shlex
24import subprocess
25import sys
26import tempfile
27
28from UpdateTestChecks import common
29
30SUBST = {
31    '%clang': [],
32    '%clang_cc1': ['-cc1'],
33    '%clangxx': ['--driver-mode=g++'],
34}
35
36def get_line2func_list(args, clang_args):
37  ret = collections.defaultdict(list)
38  # Use clang's JSON AST dump to get the mangled name
39  json_dump_args = [args.clang] + clang_args + ['-fsyntax-only', '-o', '-']
40  if '-cc1' not in json_dump_args:
41    # For tests that invoke %clang instead if %clang_cc1 we have to use
42    # -Xclang -ast-dump=json instead:
43    json_dump_args.append('-Xclang')
44  json_dump_args.append('-ast-dump=json')
45  common.debug('Running', ' '.join(json_dump_args))
46
47  popen = subprocess.Popen(json_dump_args, stdout=subprocess.PIPE,
48                           stderr=subprocess.PIPE, universal_newlines=True)
49  stdout, stderr = popen.communicate()
50  if popen.returncode != 0:
51    sys.stderr.write('Failed to run ' + ' '.join(json_dump_args) + '\n')
52    sys.stderr.write(stderr)
53    sys.stderr.write(stdout)
54    sys.exit(2)
55
56  # Parse the clang JSON and add all children of type FunctionDecl.
57  # TODO: Should we add checks for global variables being emitted?
58  def parse_clang_ast_json(node, loc, search):
59    node_kind = node['kind']
60    # Recurse for the following nodes that can contain nested function decls:
61    if node_kind in ('NamespaceDecl', 'LinkageSpecDecl', 'TranslationUnitDecl',
62                     'CXXRecordDecl', 'ClassTemplateSpecializationDecl'):
63      # Specializations must use the loc from the specialization, not the
64      # template, and search for the class's spelling as the specialization
65      # does not mention the method names in the source.
66      if node_kind == 'ClassTemplateSpecializationDecl':
67        inner_loc = node['loc']
68        inner_search = node['name']
69      else:
70        inner_loc = None
71        inner_search = None
72      if 'inner' in node:
73        for inner in node['inner']:
74          parse_clang_ast_json(inner, inner_loc, inner_search)
75    # Otherwise we ignore everything except functions:
76    if node_kind not in ('FunctionDecl', 'CXXMethodDecl', 'CXXConstructorDecl',
77                         'CXXDestructorDecl', 'CXXConversionDecl'):
78      return
79    if loc is None:
80      loc = node['loc']
81    if node.get('isImplicit') is True and node.get('storageClass') == 'extern':
82      common.debug('Skipping builtin function:', node['name'], '@', loc)
83      return
84    common.debug('Found function:', node['kind'], node['name'], '@', loc)
85    line = loc.get('line')
86    # If there is no line it is probably a builtin function -> skip
87    if line is None:
88      common.debug('Skipping function without line number:', node['name'], '@', loc)
89      return
90
91    # If there is no 'inner' object, it is a function declaration and we can
92    # skip it. However, function declarations may also contain an 'inner' list,
93    # but in that case it will only contains ParmVarDecls. If we find an entry
94    # that is not a ParmVarDecl, we know that this is a function definition.
95    has_body = False
96    if 'inner' in node:
97      for i in node['inner']:
98        if i.get('kind', 'ParmVarDecl') != 'ParmVarDecl':
99          has_body = True
100          break
101    if not has_body:
102      common.debug('Skipping function without body:', node['name'], '@', loc)
103      return
104    spell = node['name']
105    if search is None:
106      search = spell
107    mangled = node.get('mangledName', spell)
108    ret[int(line)-1].append((spell, mangled, search))
109
110  ast = json.loads(stdout)
111  if ast['kind'] != 'TranslationUnitDecl':
112    common.error('Clang AST dump JSON format changed?')
113    sys.exit(2)
114  parse_clang_ast_json(ast, None, None)
115
116  for line, funcs in sorted(ret.items()):
117    for func in funcs:
118      common.debug('line {}: found function {}'.format(line+1, func), file=sys.stderr)
119  if not ret:
120    common.warn('Did not find any functions using', ' '.join(json_dump_args))
121  return ret
122
123
124def str_to_commandline(value):
125  if not value:
126    return []
127  return shlex.split(value)
128
129
130def infer_dependent_args(args):
131  if not args.clang:
132    if not args.llvm_bin:
133      args.clang = 'clang'
134    else:
135      args.clang = os.path.join(args.llvm_bin, 'clang')
136  if not args.opt:
137    if not args.llvm_bin:
138      args.opt = 'opt'
139    else:
140      args.opt = os.path.join(args.llvm_bin, 'opt')
141
142
143def config():
144  parser = argparse.ArgumentParser(
145      description=__doc__,
146      formatter_class=argparse.RawTextHelpFormatter)
147  parser.add_argument('--llvm-bin', help='llvm $prefix/bin path')
148  parser.add_argument('--clang',
149                      help='"clang" executable, defaults to $llvm_bin/clang')
150  parser.add_argument('--clang-args', default=[], type=str_to_commandline,
151                      help='Space-separated extra args to clang, e.g. --clang-args=-v')
152  parser.add_argument('--opt',
153                      help='"opt" executable, defaults to $llvm_bin/opt')
154  parser.add_argument(
155      '--functions', nargs='+', help='A list of function name regexes. '
156      'If specified, update CHECK lines for functions matching at least one regex')
157  parser.add_argument(
158      '--x86_extra_scrub', action='store_true',
159      help='Use more regex for x86 matching to reduce diffs between various subtargets')
160  parser.add_argument('--function-signature', action='store_true',
161                      help='Keep function signature information around for the check line')
162  parser.add_argument('--check-attributes', action='store_true',
163                      help='Check "Function Attributes" for functions')
164  parser.add_argument('--check-globals', action='store_true',
165                      help='Check global entries (global variables, metadata, attribute sets, ...) for functions')
166  parser.add_argument('tests', nargs='+')
167  args = common.parse_commandline_args(parser)
168  infer_dependent_args(args)
169
170  if not distutils.spawn.find_executable(args.clang):
171    print('Please specify --llvm-bin or --clang', file=sys.stderr)
172    sys.exit(1)
173
174  # Determine the builtin includes directory so that we can update tests that
175  # depend on the builtin headers. See get_clang_builtin_include_dir() and
176  # use_clang() in llvm/utils/lit/lit/llvm/config.py.
177  try:
178    builtin_include_dir = subprocess.check_output(
179      [args.clang, '-print-file-name=include']).decode().strip()
180    SUBST['%clang_cc1'] = ['-cc1', '-internal-isystem', builtin_include_dir,
181                           '-nostdsysteminc']
182  except subprocess.CalledProcessError:
183    common.warn('Could not determine clang builtins directory, some tests '
184                'might not update correctly.')
185
186  if not distutils.spawn.find_executable(args.opt):
187    # Many uses of this tool will not need an opt binary, because it's only
188    # needed for updating a test that runs clang | opt | FileCheck. So we
189    # defer this error message until we find that opt is actually needed.
190    args.opt = None
191
192  return args, parser
193
194
195def get_function_body(builder, args, filename, clang_args, extra_commands,
196                      prefixes):
197  # TODO Clean up duplication of asm/common build_function_body_dictionary
198  # Invoke external tool and extract function bodies.
199  raw_tool_output = common.invoke_tool(args.clang, clang_args, filename)
200  for extra_command in extra_commands:
201    extra_args = shlex.split(extra_command)
202    with tempfile.NamedTemporaryFile() as f:
203      f.write(raw_tool_output.encode())
204      f.flush()
205      if extra_args[0] == 'opt':
206        if args.opt is None:
207          print(filename, 'needs to run opt. '
208                'Please specify --llvm-bin or --opt', file=sys.stderr)
209          sys.exit(1)
210        extra_args[0] = args.opt
211      raw_tool_output = common.invoke_tool(extra_args[0],
212                                           extra_args[1:], f.name)
213  if '-emit-llvm' in clang_args:
214    builder.process_run_line(
215            common.OPT_FUNCTION_RE, common.scrub_body, raw_tool_output,
216            prefixes, False)
217  else:
218    print('The clang command line should include -emit-llvm as asm tests '
219          'are discouraged in Clang testsuite.', file=sys.stderr)
220    sys.exit(1)
221
222def exec_run_line(exe):
223  popen = subprocess.Popen(exe, stdout=subprocess.PIPE, stderr=subprocess.PIPE, universal_newlines=True)
224  stdout, stderr = popen.communicate()
225  if popen.returncode != 0:
226    sys.stderr.write('Failed to run ' + ' '.join(exe) + '\n')
227    sys.stderr.write(stderr)
228    sys.stderr.write(stdout)
229    sys.exit(3)
230
231def main():
232  initial_args, parser = config()
233  script_name = os.path.basename(__file__)
234
235  for ti in common.itertests(initial_args.tests, parser, 'utils/' + script_name,
236                             comment_prefix='//', argparse_callback=infer_dependent_args):
237    # Build a list of filechecked and non-filechecked RUN lines.
238    run_list = []
239    line2func_list = collections.defaultdict(list)
240
241    subs = {
242      '%s' : ti.path,
243      '%t' : tempfile.NamedTemporaryFile().name,
244      '%S' : os.path.dirname(ti.path),
245    }
246
247    for l in ti.run_lines:
248      commands = [cmd.strip() for cmd in l.split('|')]
249
250      triple_in_cmd = None
251      m = common.TRIPLE_ARG_RE.search(commands[0])
252      if m:
253        triple_in_cmd = m.groups()[0]
254
255      # Parse executable args.
256      exec_args = shlex.split(commands[0])
257      # Execute non-clang runline.
258      if exec_args[0] not in SUBST:
259        # Do lit-like substitutions.
260        for s in subs:
261          exec_args = [i.replace(s, subs[s]) if s in i else i for i in exec_args]
262        run_list.append((None, exec_args, None, None))
263        continue
264      # This is a clang runline, apply %clang substitution rule, do lit-like substitutions,
265      # and append args.clang_args
266      clang_args = exec_args
267      clang_args[0:1] = SUBST[clang_args[0]]
268      for s in subs:
269        clang_args = [i.replace(s, subs[s]) if s in i else i for i in clang_args]
270      clang_args += ti.args.clang_args
271
272      # Extract -check-prefix in FileCheck args
273      filecheck_cmd = commands[-1]
274      common.verify_filecheck_prefixes(filecheck_cmd)
275      if not filecheck_cmd.startswith('FileCheck '):
276        # Execute non-filechecked clang runline.
277        exe = [ti.args.clang] + clang_args
278        run_list.append((None, exe, None, None))
279        continue
280
281      check_prefixes = [item for m in common.CHECK_PREFIX_RE.finditer(filecheck_cmd)
282                               for item in m.group(1).split(',')]
283      if not check_prefixes:
284        check_prefixes = ['CHECK']
285      run_list.append((check_prefixes, clang_args, commands[1:-1], triple_in_cmd))
286
287    # Execute clang, generate LLVM IR, and extract functions.
288
289    # Store only filechecked runlines.
290    filecheck_run_list = [i for i in run_list if i[0]]
291    builder = common.FunctionTestBuilder(
292      run_list=filecheck_run_list,
293      flags=ti.args,
294      scrubber_args=[],
295      path=ti.path)
296
297    for prefixes, args, extra_commands, triple_in_cmd in run_list:
298      # Execute non-filechecked runline.
299      if not prefixes:
300        print('NOTE: Executing non-FileChecked RUN line: ' + ' '.join(args), file=sys.stderr)
301        exec_run_line(args)
302        continue
303
304      clang_args = args
305      common.debug('Extracted clang cmd: clang {}'.format(clang_args))
306      common.debug('Extracted FileCheck prefixes: {}'.format(prefixes))
307
308      get_function_body(builder, ti.args, ti.path, clang_args, extra_commands,
309                        prefixes)
310
311      # Invoke clang -Xclang -ast-dump=json to get mapping from start lines to
312      # mangled names. Forward all clang args for now.
313      for k, v in get_line2func_list(ti.args, clang_args).items():
314        line2func_list[k].extend(v)
315
316    func_dict = builder.finish_and_get_func_dict()
317    global_vars_seen_dict = {}
318    prefix_set = set([prefix for p in filecheck_run_list for prefix in p[0]])
319    output_lines = []
320    has_checked_pre_function_globals = False
321
322    include_generated_funcs = common.find_arg_in_test(ti,
323                                                      lambda args: ti.args.include_generated_funcs,
324                                                      '--include-generated-funcs',
325                                                      True)
326
327    if include_generated_funcs:
328      # Generate the appropriate checks for each function.  We need to emit
329      # these in the order according to the generated output so that CHECK-LABEL
330      # works properly.  func_order provides that.
331
332      # It turns out that when clang generates functions (for example, with
333      # -fopenmp), it can sometimes cause functions to be re-ordered in the
334      # output, even functions that exist in the source file.  Therefore we
335      # can't insert check lines before each source function and instead have to
336      # put them at the end.  So the first thing to do is dump out the source
337      # lines.
338      common.dump_input_lines(output_lines, ti, prefix_set, '//')
339
340      # Now generate all the checks.
341      def check_generator(my_output_lines, prefixes, func):
342        if '-emit-llvm' in clang_args:
343          return common.add_ir_checks(my_output_lines, '//',
344                                      prefixes,
345                                      func_dict, func, False,
346                                      ti.args.function_signature,
347                                      global_vars_seen_dict,
348                                      is_filtered=builder.is_filtered())
349        else:
350          return asm.add_checks(my_output_lines, '//',
351                                prefixes,
352                                func_dict, func, global_vars_seen_dict,
353                                is_filtered=builder.is_filtered())
354
355      if ti.args.check_globals:
356        common.add_global_checks(builder.global_var_dict(), '//', run_list,
357                                 output_lines, global_vars_seen_dict, True,
358                                 True)
359      common.add_checks_at_end(output_lines, filecheck_run_list, builder.func_order(),
360                               '//', lambda my_output_lines, prefixes, func:
361                               check_generator(my_output_lines,
362                                               prefixes, func))
363    else:
364      # Normal mode.  Put checks before each source function.
365      for line_info in ti.iterlines(output_lines):
366        idx = line_info.line_number
367        line = line_info.line
368        args = line_info.args
369        include_line = True
370        m = common.CHECK_RE.match(line)
371        if m and m.group(1) in prefix_set:
372          continue  # Don't append the existing CHECK lines
373        # Skip special separator comments added by commmon.add_global_checks.
374        if line.strip() == '//' + common.SEPARATOR:
375          continue
376        if idx in line2func_list:
377          added = set()
378          for spell, mangled, search in line2func_list[idx]:
379            # One line may contain multiple function declarations.
380            # Skip if the mangled name has been added before.
381            # The line number may come from an included file, we simply require
382            # the search string (normally the function's spelling name, but is
383            # the class's spelling name for class specializations) to appear on
384            # the line to exclude functions from other files.
385            if mangled in added or search not in line:
386              continue
387            if args.functions is None or any(re.search(regex, spell) for regex in args.functions):
388              last_line = output_lines[-1].strip()
389              while last_line == '//':
390                # Remove the comment line since we will generate a new  comment
391                # line as part of common.add_ir_checks()
392                output_lines.pop()
393                last_line = output_lines[-1].strip()
394              if ti.args.check_globals and not has_checked_pre_function_globals:
395                common.add_global_checks(builder.global_var_dict(), '//',
396                                         run_list, output_lines,
397                                         global_vars_seen_dict, True, True)
398                has_checked_pre_function_globals = True
399              if added:
400                output_lines.append('//')
401              added.add(mangled)
402              common.add_ir_checks(output_lines, '//', filecheck_run_list, func_dict, mangled,
403                                   False, args.function_signature, global_vars_seen_dict,
404                                   is_filtered=builder.is_filtered())
405              if line.rstrip('\n') == '//':
406                include_line = False
407
408        if include_line:
409          output_lines.append(line.rstrip('\n'))
410
411    if ti.args.check_globals:
412      common.add_global_checks(builder.global_var_dict(), '//', run_list,
413                               output_lines, global_vars_seen_dict, True, False)
414    common.debug('Writing %d lines to %s...' % (len(output_lines), ti.path))
415    with open(ti.path, 'wb') as f:
416      f.writelines(['{}\n'.format(l).encode('utf-8') for l in output_lines])
417
418  return 0
419
420
421if __name__ == '__main__':
422  sys.exit(main())
423