contrib/make_sunver.pl

   1 #!/usr/bin/perl -w
   2
   3 # make_sunver.pl
   4 #
   5 # This script takes at least two arguments, a GNU style version script and
   6 # a list of object and archive files, and generates a corresponding Sun
   7 # style version script as follows:
   8 #
   9 # Each glob pattern, C++ mangled pattern or literal in the input script is
  10 # matched against all global symbols in the input objects, emitting those
  11 # that matched (or nothing if no match was found).
  12 # A comment with the original pattern and its type is left in the output
  13 # file to make it easy to understand the matches.
  14 #
  15 # It uses elfdump when present (native), GNU readelf otherwise.
  16 # It depends on the GNU version of c++filt, since it must understand the
  17 # GNU mangling style.
  18
  19 use FileHandle;
  20 use IPC::Open2;
  21
  22 # Input version script, GNU style.
  23 my $symvers = shift;
  24
  25 ##########
  26 # Get all the symbols from the library, match them, and add them to a hash.
  27
  28 my %sym_hash = ();
  29
  30 # List of objects and archives to process.
  31 my @OBJECTS = ();
  32
  33 # List of shared objects to omit from processing.
  34 my @SHAREDOBJS = ();
  35
  36 # Filter out those input archives that have corresponding shared objects to
  37 # avoid adding all symbols matched in the archive to the output map.
  38 foreach $file (@ARGV) {
  39     if (($so = $file) =~ s/\.a$/.so/ && -e $so) {
  40         printf STDERR "omitted $file -> $so\n";
  41         push (@SHAREDOBJS, $so);
  42     } else {
  43         push (@OBJECTS, $file);
  44     }
  45 }
  46
  47 # We need to detect and ignore hidden symbols.  Solaris nm can only detect
  48 # this in the harder to parse default output format, and GNU nm not at all,
  49 # so use elfdump -s in the native case and GNU readelf -s otherwise.
  50 # GNU objdump -t cannot be used since it produces a variable number of
  51 # columns.
  52
  53 # The path to elfdump.
  54 my $elfdump = "/usr/ccs/bin/elfdump";
  55
  56 if (-f $elfdump) {
  57     open ELFDUMP,$elfdump.' -s '.(join ' ',@OBJECTS).'|' or die $!;
  58     my $skip_arsym = 0;
  59
  60     while (<ELFDUMP>) {
  61         chomp;
  62
  63         # Ignore empty lines.
  64         if (/^$/) {
  65             # End of archive symbol table, stop skipping.
  66             $skip_arsym = 0 if $skip_arsym;
  67             next;
  68         }
  69
  70         # Keep skipping until end of archive symbol table.
  71         next if ($skip_arsym);
  72
  73         # Ignore object name header for individual objects and archives.
  74         next if (/:$/);
  75
  76         # Ignore table header lines.
  77         next if (/^Symbol Table Section:/);
  78         next if (/index.*value.*size/);
  79
  80         # Start of archive symbol table: start skipping.
  81         if (/^Symbol Table: \(archive/) {
  82             $skip_arsym = 1;
  83             next;
  84         }
  85
  86         # Split table.
  87         (undef, undef, undef, undef, $bind, $oth, undef, $shndx, $name) = split;
  88
  89         # Error out for unknown input.
  90         die "unknown input line:\n$_" unless defined($bind);
  91
  92         # Ignore local symbols.
  93         next if ($bind eq "LOCL");
  94         # Ignore hidden symbols.
  95         next if ($oth eq "H");
  96         # Ignore undefined symbols.
  97         next if ($shndx eq "UNDEF");
  98         # Error out for unhandled cases.
  99         if ($bind !~ /^(GLOB|WEAK)/ or $oth ne "D") {
 100             die "unhandled symbol:\n$_";
 101         }
 102
 103         # Remember symbol.
 104         $sym_hash{$name}++;
 105     }
 106     close ELFDUMP or die "$elfdump error";
 107 } else {
 108     open READELF, 'readelf -s -W '.(join ' ',@OBJECTS).'|' or die $!;
 109     # Process each symbol.
 110     while (<READELF>) {
 111         chomp;
 112
 113         # Ignore empty lines.
 114         next if (/^$/);
 115
 116         # Ignore object name header.
 117         next if (/^File: .*$/);
 118
 119         # Ignore table header lines.
 120         next if (/^Symbol table.*contains.*:/);
 121         next if (/Num:.*Value.*Size/);
 122
 123         # Split table.
 124         (undef, undef, undef, undef, $bind, $vis, $ndx, $name) = split;
 125
 126         # Error out for unknown input.
 127         die "unknown input line:\n$_" unless defined($bind);
 128
 129         # Ignore local symbols.
 130         next if ($bind eq "LOCAL");
 131         # Ignore hidden symbols.
 132         next if ($vis eq "HIDDEN");
 133         # Ignore undefined symbols.
 134         next if ($ndx eq "UND");
 135         # Error out for unhandled cases.
 136         if ($bind !~ /^(GLOBAL|WEAK)/ or $vis ne "DEFAULT") {
 137             die "unhandled symbol:\n$_";
 138         }
 139
 140         # Remember symbol.
 141         $sym_hash{$name}++;
 142     }
 143     close READELF or die "readelf error";
 144 }
 145
 146 ##########
 147 # The various types of glob patterns.
 148 #
 149 # A glob pattern that is to be applied to the demangled name: 'cxx'.
 150 # A glob patterns that applies directly to the name in the .o files: 'glob'.
 151 # This pattern is ignored; used for local variables (usually just '*'): 'ign'.
 152
 153 # The type of the current pattern.
 154 my $glob = 'glob';
 155
 156 # We're currently inside `extern "C++"', which Sun ld doesn't understand.
 157 my $in_extern = 0;
 158
 159 # The c++filt command to use.  This *must* be GNU c++filt; the Sun Studio
 160 # c++filt doesn't handle the GNU mangling style.
 161 my $cxxfilt = $ENV{'CXXFILT'} || "c++filt";
 162
 163 # The current version name.
 164 my $current_version = "";
 165
 166 # Was there any attempt to match a symbol to this version?
 167 my $matches_attempted;
 168
 169 # The number of versions which matched this symbol.
 170 my $matched_symbols;
 171
 172 open F,$symvers or die $!;
 173
 174 # Print information about generating this file
 175 print "# This file was generated by make_sunver.pl.  DO NOT EDIT!\n";
 176 print "# It was generated by:\n";
 177 printf "# %s %s %s\n", $0, $symvers, (join ' ',@ARGV);
 178 printf "# Omitted archives with corresponding shared libraries: %s\n",
 179     (join ' ', @SHAREDOBJS) if $#SHAREDOBJS >= 0;
 180 print "#\n\n";
 181
 182 while (<F>) {
 183     # Lines of the form '};'
 184     if (/^([ \t]*)(\}[ \t]*;[ \t]*)$/) {
 185         $glob = 'glob';
 186         if ($in_extern) {
 187             $in_extern--;
 188             print "$1##$2\n";
 189         } else {
 190             print;
 191         }
 192         next;
 193     }
 194
 195     # Lines of the form '} SOME_VERSION_NAME_1.0;'
 196     if (/^[ \t]*\}[ \tA-Z0-9_.a-z]+;[ \t]*$/) {
 197         $glob = 'glob';
 198         # We tried to match symbols agains this version, but none matched.
 199         # Emit dummy hidden symbol to avoid marking this version WEAK.
 200         if ($matches_attempted && $matched_symbols == 0) {
 201             print "  hidden:\n";
 202             print "    .force_WEAK_off_$current_version = DATA S0x0 V0x0;\n";
 203         }
 204         print; next;
 205     }
 206
 207     # Comment and blank lines
 208     if (/^[ \t]*\#/) { print; next; }
 209     if (/^[ \t]*$/) { print; next; }
 210
 211     # Lines of the form '{'
 212     if (/^([ \t]*){$/) {
 213         if ($in_extern) {
 214             print "$1##{\n";
 215         } else {
 216             print;
 217         }
 218         next;
 219     }
 220
 221     # Lines of the form 'SOME_VERSION_NAME_1.1 {'
 222     if (/^([A-Z0-9_.]+)[ \t]+{$/) {
 223         # Record version name.
 224         $current_version = $1;
 225         # Reset match attempts, #matched symbols for this version.
 226         $matches_attempted = 0;
 227         $matched_symbols = 0;
 228         print;
 229         next;
 230     }
 231
 232     # Ignore 'global:'
 233     if (/^[ \t]*global:$/) { print; next; }
 234
 235     # After 'local:', globs should be ignored, they won't be exported.
 236     if (/^[ \t]*local:$/) {
 237         $glob = 'ign';
 238         print;
 239         next;
 240     }
 241
 242     # After 'extern "C++"', globs are C++ patterns
 243     if (/^([ \t]*)(extern \"C\+\+\"[ \t]*)$/) {
 244         $in_extern++;
 245         $glob = 'cxx';
 246         # Need to comment, Sun ld cannot handle this.
 247         print "$1##$2\n"; next;
 248     }
 249
 250     # Chomp newline now we're done with passing through the input file.
 251     chomp;
 252
 253     # Catch globs.  Note that '{}' is not allowed in globs by this script,
 254     # so only '*' and '[]' are available.
 255     if (/^([ \t]*)([^ \t;{}#]+);?[ \t]*$/) {
 256         my $ws = $1;
 257         my $ptn = $2;
 258         # Turn the glob into a regex by replacing '*' with '.*', '?' with '.'.
 259         # Keep $ptn so we can still print the original form.
 260         ($pattern = $ptn) =~ s/\*/\.\*/g;
 261         $pattern =~ s/\?/\./g;
 262
 263         if ($glob eq 'ign') {
 264             # We're in a local: * section; just continue.
 265             print "$_\n";
 266             next;
 267         }
 268
 269         # Print the glob commented for human readers.
 270         print "$ws##$ptn ($glob)\n";
 271         # We tried to match a symbol to this version.
 272         $matches_attempted++;
 273
 274         if ($glob eq 'glob') {
 275             my %ptn_syms = ();
 276
 277             # Match ptn against symbols in %sym_hash.
 278             foreach my $sym (keys %sym_hash) {
 279                 # Maybe it matches one of the patterns based on the symbol in
 280                 # the .o file.
 281                 $ptn_syms{$sym}++ if ($sym =~ /^$pattern$/);
 282             }
 283
 284             foreach my $sym (sort keys(%ptn_syms)) {
 285                 $matched_symbols++;
 286                 print "$ws$sym;\n";
 287             }
 288         } elsif ($glob eq 'cxx') {
 289             my %dem_syms = ();
 290
 291             # Verify that we're actually using GNU c++filt.  Other versions
 292             # most likely cannot handle GNU style symbol mangling.
 293             my $cxxout = `$cxxfilt --version 2>&1`;
 294             $cxxout =~ m/GNU/ or die "$0 requires GNU c++filt to function";
 295
 296             # Talk to c++filt through a pair of file descriptors.
 297             # Need to start a fresh instance per pattern, otherwise the
 298             # process grows to 500+ MB.
 299             my $pid = open2(*FILTIN, *FILTOUT, $cxxfilt) or die $!;
 300
 301             # Match ptn against symbols in %sym_hash.
 302             foreach my $sym (keys %sym_hash) {
 303                 # No?  Well, maybe its demangled form matches one of those
 304                 # patterns.
 305                 printf FILTOUT "%s\n",$sym;
 306                 my $dem = <FILTIN>;
 307                 chomp $dem;
 308                 $dem_syms{$sym}++ if ($dem =~ /^$pattern$/);
 309             }
 310
 311             close FILTOUT or die "c++filt error";
 312             close FILTIN or die "c++filt error";
 313             # Need to wait for the c++filt process to avoid lots of zombies.
 314             waitpid $pid, 0;
 315
 316             foreach my $sym (sort keys(%dem_syms)) {
 317                 $matched_symbols++;
 318                 print "$ws$sym;\n";
 319             }
 320         } else {
 321             # No?  Well, then ignore it.
 322         }
 323         next;
 324     }
 325     # Important sanity check.  This script can't handle lots of formats
 326     # that GNU ld can, so be sure to error out if one is seen!
 327     die "strange line `$_'";
 328 }
 329 close F;