scripts/glibcextract.py

   1 #!/usr/bin/python3
   2 # Extract information from C headers.
   3 # Copyright (C) 2018-2023 Free Software Foundation, Inc.
   4 # This file is part of the GNU C Library.
   5 #
   6 # The GNU C Library is free software; you can redistribute it and/or
   7 # modify it under the terms of the GNU Lesser General Public
   8 # License as published by the Free Software Foundation; either
   9 # version 2.1 of the License, or (at your option) any later version.
  10 #
  11 # The GNU C Library is distributed in the hope that it will be useful,
  12 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14 # Lesser General Public License for more details.
  15 #
  16 # You should have received a copy of the GNU Lesser General Public
  17 # License along with the GNU C Library; if not, see
  18 # <https://www.gnu.org/licenses/>.
  19
  20 import collections
  21 import os.path
  22 import re
  23 import subprocess
  24 import tempfile
  25
  26
  27 def compute_c_consts(sym_data, cc):
  28     """Compute the values of some C constants.
  29
  30     The first argument is a list whose elements are either strings
  31     (preprocessor directives, or the special string 'START' to
  32     indicate this function should insert its initial boilerplate text
  33     in the output there) or pairs of strings (a name and a C
  34     expression for the corresponding value).  Preprocessor directives
  35     in the middle of the list may be used to select which constants
  36     end up being evaluated using which expressions.
  37
  38     """
  39     out_lines = []
  40     for arg in sym_data:
  41         if isinstance(arg, str):
  42             if arg == 'START':
  43                 out_lines.append('void\ndummy (void)\n{')
  44             else:
  45                 out_lines.append(arg)
  46             continue
  47         name = arg[0]
  48         value = arg[1]
  49         out_lines.append('asm ("/* @@@name@@@%s@@@value@@@%%0@@@end@@@ */" '
  50                          ': : \"i\" ((long int) (%s)));'
  51                          % (name, value))
  52     out_lines.append('}')
  53     out_lines.append('')
  54     out_text = '\n'.join(out_lines)
  55     with tempfile.TemporaryDirectory() as temp_dir:
  56         c_file_name = os.path.join(temp_dir, 'test.c')
  57         s_file_name = os.path.join(temp_dir, 'test.s')
  58         with open(c_file_name, 'w') as c_file:
  59             c_file.write(out_text)
  60         # Compilation has to be from stdin to avoid the temporary file
  61         # name being written into the generated dependencies.
  62         cmd = ('%s -S -o %s -x c - < %s' % (cc, s_file_name, c_file_name))
  63         subprocess.check_call(cmd, shell=True)
  64         consts = {}
  65         with open(s_file_name, 'r') as s_file:
  66             for line in s_file:
  67                 match = re.search('@@@name@@@([^@]*)'
  68                                   '@@@value@@@[^0-9Xxa-fA-F-]*'
  69                                   '([0-9Xxa-fA-F-]+).*@@@end@@@', line)
  70                 if match:
  71                     if (match.group(1) in consts
  72                         and match.group(2) != consts[match.group(1)]):
  73                         raise ValueError('duplicate constant %s'
  74                                          % match.group(1))
  75                     consts[match.group(1)] = match.group(2)
  76         return consts
  77
  78
  79 def list_macros(source_text, cc):
  80     """List the preprocessor macros defined by the given source code.
  81
  82     The return value is a pair of dicts, the first one mapping macro
  83     names to their expansions and the second one mapping macro names
  84     to lists of their arguments, or to None for object-like macros.
  85
  86     """
  87     with tempfile.TemporaryDirectory() as temp_dir:
  88         c_file_name = os.path.join(temp_dir, 'test.c')
  89         i_file_name = os.path.join(temp_dir, 'test.i')
  90         with open(c_file_name, 'w') as c_file:
  91             c_file.write(source_text)
  92         cmd = ('%s -E -dM -o %s %s' % (cc, i_file_name, c_file_name))
  93         subprocess.check_call(cmd, shell=True)
  94         macros_exp = {}
  95         macros_args = {}
  96         with open(i_file_name, 'r') as i_file:
  97             for line in i_file:
  98                 match = re.fullmatch('#define ([0-9A-Za-z_]+)(.*)\n', line)
  99                 if not match:
 100                     raise ValueError('bad -dM output line: %s' % line)
 101                 name = match.group(1)
 102                 value = match.group(2)
 103                 if value.startswith(' '):
 104                     value = value[1:]
 105                     args = None
 106                 elif value.startswith('('):
 107                     match = re.fullmatch(r'\((.*?)\) (.*)', value)
 108                     if not match:
 109                         raise ValueError('bad -dM output line: %s' % line)
 110                     args = match.group(1).split(',')
 111                     value = match.group(2)
 112                 else:
 113                     raise ValueError('bad -dM output line: %s' % line)
 114                 if name in macros_exp:
 115                     raise ValueError('duplicate macro: %s' % line)
 116                 macros_exp[name] = value
 117                 macros_args[name] = args
 118     return macros_exp, macros_args
 119
 120
 121 def compute_macro_consts(source_text, cc, macro_re, exclude_re=None):
 122     """Compute the integer constant values of macros defined by source_text.
 123
 124     Macros must match the regular expression macro_re, and if
 125     exclude_re is defined they must not match exclude_re.  Values are
 126     computed with compute_c_consts.
 127
 128     """
 129     macros_exp, macros_args = list_macros(source_text, cc)
 130     macros_set = {m for m in macros_exp
 131                   if (macros_args[m] is None
 132                       and re.fullmatch(macro_re, m)
 133                       and (exclude_re is None
 134                            or not re.fullmatch(exclude_re, m)))}
 135     sym_data = [source_text, 'START']
 136     sym_data.extend(sorted((m, m) for m in macros_set))
 137     return compute_c_consts(sym_data, cc)
 138
 139
 140 def compare_macro_consts(source_1, source_2, cc, macro_re, exclude_re=None,
 141                          allow_extra_1=False, allow_extra_2=False):
 142     """Compare the values of macros defined by two different sources.
 143
 144     The sources would typically be includes of a glibc header and a
 145     kernel header.  If allow_extra_1, the first source may define
 146     extra macros (typically if the kernel headers are older than the
 147     version glibc has taken definitions from); if allow_extra_2, the
 148     second source may define extra macros (typically if the kernel
 149     headers are newer than the version glibc has taken definitions
 150     from).  Return 1 if there were any differences other than those
 151     allowed, 0 if the macro values were the same apart from any
 152     allowed differences.
 153
 154     """
 155     macros_1 = compute_macro_consts(source_1, cc, macro_re, exclude_re)
 156     macros_2 = compute_macro_consts(source_2, cc, macro_re, exclude_re)
 157     if macros_1 == macros_2:
 158         return 0
 159     print('First source:\n%s\n' % source_1)
 160     print('Second source:\n%s\n' % source_2)
 161     ret = 0
 162     for name, value in sorted(macros_1.items()):
 163         if name not in macros_2:
 164             print('Only in first source: %s' % name)
 165             if not allow_extra_1:
 166                 ret = 1
 167         elif macros_1[name] != macros_2[name]:
 168             print('Different values for %s: %s != %s'
 169                   % (name, macros_1[name], macros_2[name]))
 170             ret = 1
 171     for name in sorted(macros_2.keys()):
 172         if name not in macros_1:
 173             print('Only in second source: %s' % name)
 174             if not allow_extra_2:
 175                 ret = 1
 176     return ret
 177
 178 CompileResult = collections.namedtuple("CompileResult", "returncode output")
 179
 180 def compile_c_snippet(snippet, cc, extra_cc_args=''):
 181     """Compile and return whether the SNIPPET can be build with CC along
 182        EXTRA_CC_ARGS compiler flags.  Return a CompileResult with RETURNCODE
 183        being 0 for success, or the failure value and the compiler output.
 184     """
 185     with tempfile.TemporaryDirectory() as temp_dir:
 186         c_file_name = os.path.join(temp_dir, 'test.c')
 187         obj_file_name = os.path.join(temp_dir, 'test.o')
 188         with open(c_file_name, 'w') as c_file:
 189             c_file.write(snippet + '\n')
 190         cmd = cc.split() + extra_cc_args.split() + ['-c', '-o', obj_file_name,
 191                 c_file_name]
 192         r = subprocess.run(cmd, check=False, stdout=subprocess.PIPE,
 193                 stderr=subprocess.STDOUT)
 194         return CompileResult(r.returncode, r.stdout)