view release on metacpan or  search on metacpan

lib/B/Disassembler.pm  view on Meta::CPAN

    printf "#archflag    %d\n", $archflag  if defined $archflag;
    printf "#perlversion %s\n", $perlversion if $perlversion;
    print "\n";
  }
  while ( defined( $c = $fh->getc ) ) {
    $c    = ord($c);
    $insn = $insn_name[$c];
    if ( !defined($insn) || $insn eq "unused" ) {
      my $pos = $fh->tell - 1;
      warn "Illegal instruction code $c at stream offset $pos.\n";
    }
    $getmeth = $insn_data{$insn}->[2];
    #warn "EOF at $insn $getmeth" if $fh->eof();
    $arg     = $fh->$getmeth();
    if ( defined($arg) ) {
      &$out( $insn, $arg, $verbose );
    }
    else {
      &$out( $insn, undef, $verbose );
    }
  }
}

1;

__END__

=head1 NAME

B::Disassembler - Disassemble Perl bytecode

=head1 SYNOPSIS

	use Disassembler qw(print_insn);
        my $fh = new FileHandle "<$ARGV[0]";
	disassemble_fh($fh, \&print_insn);

=head1 DESCRIPTION

disassemble_fh takes an filehandle with bytecode and a printer function.
The printer function gets three arguments: insn, arg (optional) and the comment.

See F<lib/B/Disassembler.pm> and F<scripts/disassemble>.

=head1 disassemble_fh (filehandle, printer_coderef, [ verbose ])

disassemble_fh takes an filehandle with bytecode and a printer coderef.

Two default printer functions are provided:

  print_insn print_insn_bare

=head1 print_insn

Callback function for disassemble_fh, which gets three arguments from
the disassembler.  insn (a string), arg (a string or number or undef)
and the comment (an optional string).

This supports the new behaviour in F<scripts/disassemble>.  It prints
each insn and optional argument with some additional comments, which
looks similar to B::Assembler with option -S (commented source).

=head1 print_insn_bare

This is the same as the old behaviour of scripts/disassemble.  It
prints each insn and optional argument without any comments. Line per
line.

=head1 get_header

Returns the .plc header as array of

  ( magic, archname, blversion, ivsize, ptrsize,
    byteorder, longsize, archflag, perlversion )

in ARRAY context, or in SCALAR context the array from above as named hash.

B<magic> is always "PLBC". "PLJC" is reserved for JIT'ted code also
loaded via ByteLoader.

B<archname> is the archname string and is in the ByteLoader up to 0.06
checked strictly. Starting with ByteLoader 0.06_05 platform
compatibility is implemented by checking the $archflag, and doing
byteorder swapping for same length longsize, and adjusting the ivsize
and ptrsize.

B<blversion> is the ByteLoader version from the header as string.
Up to ByteLoader 0.06 this version must have matched exactly, since 0.07
earlier ByteLoader versions are also accepted in the ByteLoader.

B<ivsize> matches $Config{ivsize} of the assembling perl.
A number, 4 or 8 only supported.

B<ptrsize> matches $Config{ptrsize} of the assembling perl.
A number, 4 or 8 only supported.

B<longsize> is $Config{longsize} of the assembling perl.
A number, 4 or 8.
Only since blversion 0.06_03.

B<byteorder> is a string of "0x12345678" on big-endian or "0x78563412" (?)
on little-endian machines. The beginning "0x" is stripped for compatibility
with intermediate ByteLoader versions, i.e. 5.6.1 to 5.8.0,
Added with blversion 0.06_03, and also with blversion 0.04.
See L<BcVersions>

B<archflag> is a bitmask of opcode platform-dependencies.
Currently used:
  bit 1 for USE_ITHREADS
  bit 2 for MULTIPLICITY
Added with  blversion 0.06_05.

B<perlversion> $] of the perl which produced this bytecode as string.
Added with blversion 0.06_06.

=head1 AUTHORS

Malcolm Beattie C<MICB at cpan.org> I<(retired)>,
Reini Urban C<perl-compiler@googlegroups.com> since 2008.

=cut