summaryrefslogtreecommitdiff
path: root/Master/tlpkg/tlperl/lib/B/Assembler.pm
blob: f312273ce36034427c64b8ada4708ea339b2eb8e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
#      Assembler.pm
#
#      Copyright (c) 1996 Malcolm Beattie
#
#      You may distribute under the terms of either the GNU General Public
#      License or the Artistic License, as specified in the README file.

package B::Assembler;
use Exporter;
use B qw(ppname);
use B::Asmdata qw(%insn_data @insn_name);
use Config qw(%Config);
require ByteLoader;		# we just need its $VERSION

no warnings;			# XXX

@ISA = qw(Exporter);
@EXPORT_OK = qw(assemble_fh newasm endasm assemble asm);
$VERSION = 0.07;

use strict;
my %opnumber;
my ($i, $opname);
for ($i = 0; defined($opname = ppname($i)); $i++) {
    $opnumber{$opname} = $i;
}

my($linenum, $errors, $out); #	global state, set up by newasm

sub error {
    my $str = shift;
    warn "$linenum: $str\n";
    $errors++;
}

my $debug = 0;
sub debug { $debug = shift }

sub limcheck($$$$){
    my( $val, $lo, $hi, $loc ) = @_;
    if( $val < $lo || $hi < $val ){
        error "argument for $loc outside [$lo, $hi]: $val";
        $val = $hi;
    }
    return $val;
}

#
# First define all the data conversion subs to which Asmdata will refer
#

sub B::Asmdata::PUT_U8 {
    my $arg = shift;
    my $c = uncstring($arg);
    if (defined($c)) {
	if (length($c) != 1) {
	    error "argument for U8 is too long: $c";
	    $c = substr($c, 0, 1);
	}
    } else {
        $arg = limcheck( $arg, 0, 0xff, 'U8' );
	$c = chr($arg);
    }
    return $c;
}

sub B::Asmdata::PUT_U16 {
    my $arg = limcheck( $_[0], 0, 0xffff, 'U16' );
    pack("S", $arg);
}
sub B::Asmdata::PUT_U32 {
    my $arg = limcheck( $_[0], 0, 0xffffffff, 'U32' );
    pack("L", $arg);
}
sub B::Asmdata::PUT_I32 {
    my $arg = limcheck( $_[0], -0x80000000, 0x7fffffff, 'I32' );
    pack("l", $arg);
}
sub B::Asmdata::PUT_NV  { sprintf("%s\0", $_[0]) } # "%lf" looses precision and pack('d',...)
						   # may not even be portable between compilers
sub B::Asmdata::PUT_objindex { # could allow names here
    my $arg = limcheck( $_[0], 0, 0xffffffff, '*index' );
    pack("L", $arg);
} 
sub B::Asmdata::PUT_svindex { &B::Asmdata::PUT_objindex }
sub B::Asmdata::PUT_opindex { &B::Asmdata::PUT_objindex }
sub B::Asmdata::PUT_pvindex { &B::Asmdata::PUT_objindex }

sub B::Asmdata::PUT_strconst {
    my $arg = shift;
    my $str = uncstring($arg);
    if (!defined($str)) {
	error "bad string constant: $arg";
	$str = '';
    }
    if ($str =~ s/\0//g) {
	error "string constant argument contains NUL: $arg";
        $str = '';
    }
    return $str . "\0";
}

sub B::Asmdata::PUT_pvcontents {
    my $arg = shift;
    error "extraneous argument: $arg" if defined $arg;
    return "";
}
sub B::Asmdata::PUT_PV {
    my $arg = shift;
    my $str = uncstring($arg);
    if( ! defined($str) ){
        error "bad string argument: $arg";
        $str = '';
    }
    return pack("L", length($str)) . $str;
}
sub B::Asmdata::PUT_comment_t {
    my $arg = shift;
    $arg = uncstring($arg);
    error "bad string argument: $arg" unless defined($arg);
    if ($arg =~ s/\n//g) {
	error "comment argument contains linefeed: $arg";
    }
    return $arg . "\n";
}
sub B::Asmdata::PUT_double { sprintf("%s\0", $_[0]) } # see PUT_NV above
sub B::Asmdata::PUT_none {
    my $arg = shift;
    error "extraneous argument: $arg" if defined $arg;
    return "";
}
sub B::Asmdata::PUT_op_tr_array {
    my @ary = split /\s*,\s*/, shift;
    return pack "S*", @ary;
}

sub B::Asmdata::PUT_IV64 {
    return pack "Q", shift;
}

sub B::Asmdata::PUT_IV {
    $Config{ivsize} == 4 ? &B::Asmdata::PUT_I32 : &B::Asmdata::PUT_IV64;
}

sub B::Asmdata::PUT_PADOFFSET {
    $Config{ptrsize} == 8 ? &B::Asmdata::PUT_IV64 : &B::Asmdata::PUT_U32;
}

sub B::Asmdata::PUT_long {
    $Config{longsize} == 8 ? &B::Asmdata::PUT_IV64 : &B::Asmdata::PUT_U32;
}

my %unesc = (n => "\n", r => "\r", t => "\t", a => "\a",
	     b => "\b", f => "\f", v => "\013");

sub uncstring {
    my $s = shift;
    $s =~ s/^"// and $s =~ s/"$// or return undef;
    $s =~ s/\\(\d\d\d|.)/length($1) == 3 ? chr(oct($1)) : ($unesc{$1}||$1)/eg;
    return $s;
}

sub strip_comments {
    my $stmt = shift;
    # Comments only allowed in instructions which don't take string arguments
    # Treat string as a single line so .* eats \n characters.
    $stmt =~ s{
	^\s*	# Ignore leading whitespace
	(
	  [^"]*	# A double quote '"' indicates a string argument. If we
		# find a double quote, the match fails and we strip nothing.
	)
	\s*\#	# Any amount of whitespace plus the comment marker...
	.*$	# ...which carries on to end-of-string.
    }{$1}sx;	# Keep only the instruction and optional argument.
    return $stmt;
}

# create the ByteCode header: magic, archname, ByteLoader $VERSION, ivsize,
# 	ptrsize, byteorder
# nvtype is irrelevant (floats are stored as strings)
# byteorder is strconst not U32 because of varying size issues

sub gen_header {
    my $header = "";

    $header .= B::Asmdata::PUT_U32(0x43424c50);	# 'PLBC'
    $header .= B::Asmdata::PUT_strconst('"' . $Config{archname}. '"');
    $header .= B::Asmdata::PUT_strconst(qq["$ByteLoader::VERSION"]);
    $header .= B::Asmdata::PUT_U32($Config{ivsize});
    $header .= B::Asmdata::PUT_U32($Config{ptrsize});
    $header;
}

sub parse_statement {
    my $stmt = shift;
    my ($insn, $arg) = $stmt =~ m{
	^\s*	# allow (but ignore) leading whitespace
	(.*?)	# Instruction continues up until...
	(?:	# ...an optional whitespace+argument group
	    \s+		# first whitespace.
	    (.*)	# The argument is all the rest (newlines included).
	)?$	# anchor at end-of-line
    }sx;
    if (defined($arg)) {
	if ($arg =~ s/^0x(?=[0-9a-fA-F]+$)//) {
	    $arg = hex($arg);
	} elsif ($arg =~ s/^0(?=[0-7]+$)//) {
	    $arg = oct($arg);
	} elsif ($arg =~ /^pp_/) {
	    $arg =~ s/\s*$//; # strip trailing whitespace
	    my $opnum = $opnumber{$arg};
	    if (defined($opnum)) {
		$arg = $opnum;
	    } else {
		error qq(No such op type "$arg");
		$arg = 0;
	    }
	}
    }
    return ($insn, $arg);
}

sub assemble_insn {
    my ($insn, $arg) = @_;
    my $data = $insn_data{$insn};
    if (defined($data)) {
	my ($bytecode, $putsub) = @{$data}[0, 1];
	my $argcode = &$putsub($arg);
	return chr($bytecode).$argcode;
    } else {
	error qq(no such instruction "$insn");
	return "";
    }
}

sub assemble_fh {
    my ($fh, $out) = @_;
    my $line;
    my $asm = newasm($out);
    while ($line = <$fh>) {
	assemble($line);
    }
    endasm();
}

sub newasm {
    my($outsub) = @_;

    die "Invalid printing routine for B::Assembler\n" unless ref $outsub eq 'CODE';
    die <<EOD if ref $out;
Can't have multiple byteassembly sessions at once!
	(perhaps you forgot an endasm()?)
EOD

    $linenum = $errors = 0;
    $out = $outsub;

    $out->(gen_header());
}

sub endasm {
    if ($errors) {
	die "There were $errors assembly errors\n";
    }
    $linenum = $errors = $out = 0;
}

sub assemble {
    my($line) = @_;
    my ($insn, $arg);
    $linenum++;
    chomp $line;
    if ($debug) {
	my $quotedline = $line;
	$quotedline =~ s/\\/\\\\/g;
	$quotedline =~ s/"/\\"/g;
	$out->(assemble_insn("comment", qq("$quotedline")));
    }
    if( $line = strip_comments($line) ){
        ($insn, $arg) = parse_statement($line);
        $out->(assemble_insn($insn, $arg));
        if ($debug) {
	    $out->(assemble_insn("nop", undef));
        }
    }
}

### temporary workaround

sub asm {
    return if $_[0] =~ /\s*\W/;
    if (defined $_[1]) {
	return if $_[1] eq "0" and
	    $_[0] !~ /^(?:newsvx?|av_pushx?|av_extend|xav_flags)$/;
	return if $_[1] eq "1" and $_[0] =~ /^(?:sv_refcnt)$/;
    }
    assemble "@_";
}

1;

__END__

=head1 NAME

B::Assembler - Assemble Perl bytecode

=head1 SYNOPSIS

	use B::Assembler qw(newasm endasm assemble);
	newasm(\&printsub);	# sets up for assembly
	assemble($buf); 	# assembles one line
	endasm();		# closes down

	use B::Assembler qw(assemble_fh);
	assemble_fh($fh, \&printsub);	# assemble everything in $fh

=head1 DESCRIPTION

See F<ext/B/B/Assembler.pm>.

=head1 AUTHORS

Malcolm Beattie, C<mbeattie@sable.ox.ac.uk>
Per-statement interface by Benjamin Stuhl, C<sho_pi@hotmail.com>

=cut