summaryrefslogtreecommitdiff
path: root/insns.pl
blob: 011012102a8bd8f2607feb74733fc3785f090b47 (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
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
#!/usr/bin/perl
#
# insns.pl   produce insnsa.c, insnsd.c, insnsi.h, insnsn.c from insns.dat
#
# The Netwide Assembler is copyright (C) 1996 Simon Tatham and
# Julian Hall. All rights reserved. The software is
# redistributable under the license given in the file "LICENSE"
# distributed in the NASM archive.

# Opcode prefixes which need their own opcode tables
# LONGER PREFIXES FIRST!
@disasm_prefixes = qw(0F24 0F25 0F38 0F3A 0F7A 0FA6 0FA7 0F);

# This should match MAX_OPERANDS from nasm.h
$MAX_OPERANDS = 5;

print STDERR "Reading insns.dat...\n";

@args   = ();
undef $output;
foreach $arg ( @ARGV ) {
    if ( $arg =~ /^\-/ ) {
	if  ( $arg =~ /^\-([abdin])$/ ) {
	    $output = $1;
	} else {
	    die "$0: Unknown option: ${arg}\n";
	}
    } else {
	push (@args, $arg);
    }
}

$fname = "insns.dat" unless $fname = $args[0];
open (F, $fname) || die "unable to open $fname";

%dinstables = ();
@bytecode_list = ();

$line = 0;
$insns = 0;
while (<F>) {
  $line++;
  next if /^\s*;/;   # comments
  chomp;
  split;
  next if $#_ == -1; # blank lines
  (warn "line $line does not contain four fields\n"), next if $#_ != 3;
  ($formatted, $nd) = &format(@_);
  if ($formatted) {
    $insns++;
    $aname = "aa_$_[0]";
    push @$aname, $formatted;
  }
  if ( $_[0] =~ /cc$/ ) {
      # Conditional instruction
      $k_opcodes_cc{$_[0]}++;
  } else {
      # Unconditional instruction
      $k_opcodes{$_[0]}++;
  }
  if ($formatted && !$nd) {
    push @big, $formatted;
    my @sseq = startseq($_[2]);
    foreach $i (@sseq) {
	if (!defined($dinstables{$i})) {
	    $dinstables{$i} = [];
	}
	push(@{$dinstables{$i}}, $#big);
    }
  }
}

close F;

#
# Generate the bytecode array.  At this point, @bytecode_list contains
# the full set of bytecodes.
#

# Sort by descending length
@bytecode_list = sort { scalar(@$b) <=> scalar(@$a) } @bytecode_list;
@bytecode_array = ();
%bytecode_pos = ();
$bytecode_next = 0;
foreach $bl (@bytecode_list) {
    my $h = hexstr(@$bl);
    next if (defined($bytecode_pos{$h}));

    push(@bytecode_array, $bl);
    while ($h ne '') {
	$bytecode_pos{$h} = $bytecode_next;
	$h = substr($h, 2);
	$bytecode_next++;
    }
}
undef @bytecode_list;

@opcodes    = sort keys(%k_opcodes);
@opcodes_cc = sort keys(%k_opcodes_cc);

if ( !defined($output) || $output eq 'b') {
    print STDERR "Writing insnsb.c...\n";

    open B, ">insnsb.c";
    
    print B "/* This file auto-generated from insns.dat by insns.pl" .
        " - don't edit it */\n\n";

    print B "#include \"nasm.h\"\n";
    print B "#include \"insns.h\"\n\n";

    print B "static const uint8_t nasm_bytecodes[$bytecode_next] = {\n";

    $p = 0;
    foreach $bl (@bytecode_array) {
	printf B "    /* %5d */ ", $p;
	foreach $d (@$bl) {
	    printf B "%#o,", $d;
	    $p++;
	}
	printf B "\n";
    }
    print B "};\n";

    close B;
}
    
if ( !defined($output) || $output eq 'a' ) {
    print STDERR "Writing insnsa.c...\n";

    open A, ">insnsa.c";

    print A "/* This file auto-generated from insns.dat by insns.pl" .
        " - don't edit it */\n\n";

    print A "#include \"insnsb.c\"\n\n";

    foreach $i (@opcodes, @opcodes_cc) {
	print A "static const struct itemplate instrux_${i}[] = {\n";
	$aname = "aa_$i";
	foreach $j (@$aname) {
	    print A "    ", codesubst($j), "\n";
	}
	print A "    ITEMPLATE_END\n};\n\n";
    }
    print A "const struct itemplate * const nasm_instructions[] = {\n";
    foreach $i (@opcodes, @opcodes_cc) {
	print A "    instrux_${i},\n";
    }
    print A "};\n";

    close A;
}

if ( !defined($output) || $output eq 'd' ) {
    print STDERR "Writing insnsd.c...\n";

    open D, ">insnsd.c";

    print D "/* This file auto-generated from insns.dat by insns.pl" .
        " - don't edit it */\n\n";

    print D "#include \"insnsb.c\"\n\n";

    print D "static const struct itemplate instrux[] = {\n";
    $n = 0;
    foreach $j (@big) {
	printf D "    /* %4d */ %s\n", $n++, codesubst($j);
    }
    print D "};\n";

    foreach $h (sort(keys(%dinstables))) {
	print D "\nstatic const struct itemplate * const itable_${h}[] = {\n";
	foreach $j (@{$dinstables{$h}}) {
	    print D "    instrux + $j,\n";
	}
	print D "};\n";
    }

    foreach $h (@disasm_prefixes, '') {
	$is_prefix{$h} = 1;
	print D "\n";
	print D "static " unless ($h eq '');
	print D "const struct disasm_index ";
	print D ($h eq '') ? 'itable' : "itable_$h";
	print D "[256] = {\n";
	for ($c = 0; $c < 256; $c++) {
	    $nn = sprintf("%s%02X", $h, $c);
	    if ($is_prefix{$nn}) {
		die "$0: ambiguous decoding of $nn\n"
		    if (defined($dinstables{$nn}));
		printf D "    { itable_%s, -1 },\n", $nn;
	    } elsif (defined($dinstables{$nn})) {
		printf D "    { itable_%s, %u },\n",
		$nn, scalar(@{$dinstables{$nn}});
	    } else {
		printf D "    { NULL, 0 },\n";
	    }
	}
    print D "};\n";
    }

    close D;
}

if ( !defined($output) || $output eq 'i' ) {
    print STDERR "Writing insnsi.h...\n";

    open I, ">insnsi.h";

    print I "/* This file is auto-generated from insns.dat by insns.pl" .
        " - don't edit it */\n\n";
    print I "/* This file in included by nasm.h */\n\n";

    print I "/* Instruction names */\n\n";
    print I "#ifndef NASM_INSNSI_H\n";
    print I "#define NASM_INSNSI_H 1\n\n";
    print I "enum opcode {\n";
    $maxlen = 0;
    foreach $i (@opcodes, @opcodes_cc) {
	print I "\tI_${i},\n";
	$len = length($i);
	$len++ if ( $i =~ /cc$/ );	# Condition codes can be 3 characters long
	$maxlen = $len if ( $len > $maxlen );
    }
    print I "\tI_none = -1\n";
    print I "\n};\n\n";
    print I "#define MAX_INSLEN ", $maxlen, "\n\n";
    print I "#endif /* NASM_INSNSI_H */\n";

    close I;
}

if ( !defined($output) || $output eq 'n' ) {
    print STDERR "Writing insnsn.c...\n";

    open N, ">insnsn.c";

    print N "/* This file is auto-generated from insns.dat by insns.pl" .
        " - don't edit it */\n\n";
    print N "/* This file in included by names.c */\n\n";

    print N "static const char * const insn_names[] = {";
    $first = 1;
    foreach $i (@opcodes) {
	print N "," if ( !$first );
	$first = 0;
	$ilower = $i;
	$ilower =~ tr/A-Z/a-z/;	# Change to lower case (Perl 4 compatible)
	print N "\n\t\"${ilower}\"";
    }
    print N "\n};\n\n";
    print N "/* Conditional instructions */\n";
    print N "static const char *icn[] = {";
    $first = 1;
    foreach $i (@opcodes_cc) {
	print N "," if ( !$first );
	$first = 0;
	$ilower = $i;
	$ilower =~ s/cc$//;		# Skip cc suffix
	$ilower =~ tr/A-Z/a-z/;	# Change to lower case (Perl 4 compatible)
	print N "\n\t\"${ilower}\"";
    }

    print N "\n};\n\n";
    print N "/* and the corresponding opcodes */\n";
    print N "static const enum opcode ico[] = {";
    $first = 1;
    foreach $i (@opcodes_cc) {
	print N "," if ( !$first );
	$first = 0;
	print N "\n\tI_$i";
    }

    print N "\n};\n";

    close N;
}

printf STDERR "Done: %d instructions\n", $insns;

sub format {
    my ($opcode, $operands, $codes, $flags) = @_;
    my $num, $nd = 0;
    my @bytecode;

    return (undef, undef) if $operands eq "ignore";

    # format the operands
    $operands =~ s/:/|colon,/g;
    $operands =~ s/mem(\d+)/mem|bits$1/g;
    $operands =~ s/mem/memory/g;
    $operands =~ s/memory_offs/mem_offs/g;
    $operands =~ s/imm(\d+)/imm|bits$1/g;
    $operands =~ s/imm/immediate/g;
    $operands =~ s/rm(\d+)/rm_gpr|bits$1/g;
    $operands =~ s/(mmx|xmm|ymm)rm/rm_$1/g;
    $operands =~ s/\=([0-9]+)/same_as|$1/g;
    if ($operands eq 'void') {
	@ops = ();
    } else {
	@ops = split(/\,/, $operands);
    }
    $num = scalar(@ops);
    while (scalar(@ops) < $MAX_OPERANDS) {
	push(@ops, '0');
    }
    $operands = join(',', @ops);
    $operands =~ tr/a-z/A-Z/;

    # format the flags
    $flags =~ s/,/|IF_/g;
    $flags =~ s/(\|IF_ND|IF_ND\|)//, $nd = 1 if $flags =~ /IF_ND/;
    $flags = "IF_" . $flags;

    @bytecode = (decodify($codes), 0);
    push(@bytecode_list, [@bytecode]);
    $codes = hexstr(@bytecode);

    ("{I_$opcode, $num, {$operands}, \@\@CODES-$codes\@\@, $flags},", $nd);
}

#
# Look for @@CODES-xxx@@ sequences and replace them with the appropriate
# offset into nasm_bytecodes
#
sub codesubst($) {
    my($s) = @_;
    my $n;

    while ($s =~ /\@\@CODES-([0-9A-F]+)\@\@/) {
	my $pos = $bytecode_pos{$1};
	if (!defined($pos)) {
	    die "$0: no position assigned to byte code $1\n";
	}
	$s = $` . "nasm_bytecodes+${pos}" . "$'";
    }
    return $s;
}

sub addprefix ($@) {
    my ($prefix, @list) = @_;
    my $x;
    my @l = ();

    foreach $x (@list) {
	push(@l, sprintf("%s%02X", $prefix, $x));
    }

    return @l;
}

#
# Turn a code string into a sequence of bytes
#
sub decodify($) {
  # Although these are C-syntax strings, by convention they should have
  # only octal escapes (for directives) and hexadecimal escapes
  # (for verbatim bytes)
    my($codestr) = @_;
    my $c = $codestr;
    my @codes = ();

    while ($c ne '') {
	if ($c =~ /^\\x([0-9a-f]+)(.*)$/i) {
	    push(@codes, hex $1);
	    $c = $2;
	    next;
	} elsif ($c =~ /^\\([0-7]{1,3})(.*)$/) {
	    push(@codes, oct $1);
	    $c = $2;
	    next;
	} else {
	    die "$0: unknown code format in \"$codestr\"\n";
	}
    }

    return @codes;
}

# Turn a numeric list into a hex string
sub hexstr(@) {
    my $s = '';
    my $c;

    foreach $c (@_) {
	$s .= sprintf("%02X", $c);
    }
    return $s;
}

# Here we determine the range of possible starting bytes for a given
# instruction. We need only consider the codes:
# \1 \2 \3     mean literal bytes, of course
# \4 \5 \6 \7  mean PUSH/POP of segment registers: special case
# \1[0123]     mean byte plus register value
# \330         means byte plus condition code
# \0 or \340   mean give up and return empty set
sub startseq($) {
  my ($codestr) = @_;
  my $word, @range;
  my @codes = ();
  my $c = $codestr;
  my $c0, $c1, $i;
  my $prefix = '';

  @codes = decodify($codestr);

  while ($c0 = shift(@codes)) {
      $c1 = $codes[0];
      if ($c0 == 01 || $c0 == 02 || $c0 == 03) {
	  # Fixed byte string
	  my $fbs = $prefix;
	  while (1) {
	      if ($c0 == 01 || $c0 == 02 || $c0 == 03) {
		  while ($c0--) {
		      $fbs .= sprintf("%02X", shift(@codes));
		  }
	      } else {
		  last;
	      }
	      $c0 = shift(@codes);
	  }

	  foreach $pfx (@disasm_prefixes) {
	      if (substr($fbs, 0, length($pfx)) eq $pfx) {
		  $prefix = $pfx;
		  $fbs = substr($fbs, length($pfx));
		  last;
	      }
	  }

	  if ($fbs ne '') {
	      return ($prefix.substr($fbs,0,2));
	  }

	  unshift(@codes, $c0);
      } elsif ($c0 == 04) {
	  return addprefix($prefix, 0x07, 0x17, 0x1F);
      } elsif ($c0 == 05) {
	  return addprefix($prefix, 0xA1, 0xA9);
      } elsif ($c0 == 06) {
	  return addprefix($prefix, 0x06, 0x0E, 0x16, 0x1E);
      } elsif ($c0 == 07) {
	  return addprefix($prefix, 0xA0, 0xA8);
      } elsif ($c0 >= 010 && $c0 <= 013) {
	  return addprefix($prefix, $c1..($c1+7));
      } elsif (($c0 & ~013) == 0144) {
	  return addprefix($prefix, $c1, $c1|2);
      } elsif ($c0 == 0330) {
	  return addprefix($prefix, $c1..($c1+15));
      } elsif ($c0 == 0 || $c0 == 0340) {
	  return $prefix;
      } elsif (($c0 & ~3) == 0260 || $c0 == 270) {
	  shift(@codes);
	  shift(@codes);
      } elsif ($c0 == 0172) {
	  shift(@codes);
      } else {
	  # We really need to be able to distinguish "forbidden"
	  # and "ignorable" codes here
      }
  }
  return $prefix;
}