checkpatch.pl 60.4 KB
Newer Older
1 2
#!/usr/bin/perl -w
# (c) 2001, Dave Jones. <davej@codemonkey.org.uk> (the file handling bit)
3
# (c) 2005, Joel Schopp <jschopp@austin.ibm.com> (the ugly bit)
4 5 6 7 8 9
# (c) 2007, Andy Whitcroft <apw@uk.ibm.com> (new conditions, test suite, etc)
# Licensed under the terms of the GNU GPL License version 2

use strict;

my $P = $0;
10
$P =~ s@.*/@@g;
11

Andy Whitcroft's avatar
Andy Whitcroft committed
12
my $V = '0.20';
13 14 15 16 17 18 19

use Getopt::Long qw(:config no_auto_abbrev);

my $quiet = 0;
my $tree = 1;
my $chk_signoff = 1;
my $chk_patch = 1;
20
my $tst_only;
21
my $emacs = 0;
22
my $terse = 0;
23 24
my $file = 0;
my $check = 0;
25 26
my $summary = 1;
my $mailback = 0;
27
my $summary_file = 0;
28
my $root;
29
my %debug;
30
GetOptions(
31
	'q|quiet+'	=> \$quiet,
32 33 34
	'tree!'		=> \$tree,
	'signoff!'	=> \$chk_signoff,
	'patch!'	=> \$chk_patch,
35
	'emacs!'	=> \$emacs,
36
	'terse!'	=> \$terse,
37 38 39 40
	'file!'		=> \$file,
	'subjective!'	=> \$check,
	'strict!'	=> \$check,
	'root=s'	=> \$root,
41 42
	'summary!'	=> \$summary,
	'mailback!'	=> \$mailback,
43 44
	'summary-file!'	=> \$summary_file,

45
	'debug=s'	=> \%debug,
46
	'test-only=s'	=> \$tst_only,
47 48 49 50 51
) or exit;

my $exit = 0;

if ($#ARGV < 0) {
52
	print "usage: $P [options] patchfile\n";
53
	print "version: $V\n";
54 55 56 57 58 59 60 61 62
	print "options: -q               => quiet\n";
	print "         --no-tree        => run without a kernel tree\n";
	print "         --terse          => one line per report\n";
	print "         --emacs          => emacs compile window format\n";
	print "         --file           => check a source file\n";
	print "         --strict         => enable more subjective tests\n";
	print "         --root           => path to the kernel tree root\n";
	print "         --no-summary     => suppress the per-file summary\n";
	print "         --summary-file   => include the filename in summary\n";
63 64 65
	exit(1);
}

66 67
my $dbg_values = 0;
my $dbg_possible = 0;
68
my $dbg_type = 0;
69 70 71 72
for my $key (keys %debug) {
	eval "\${dbg_$key} = '$debug{$key}';"
}

73 74 75 76 77
if ($terse) {
	$emacs = 1;
	$quiet++;
}

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
if ($tree) {
	if (defined $root) {
		if (!top_of_kernel_tree($root)) {
			die "$P: $root: --root does not point at a valid tree\n";
		}
	} else {
		if (top_of_kernel_tree('.')) {
			$root = '.';
		} elsif ($0 =~ m@(.*)/scripts/[^/]*$@ &&
						top_of_kernel_tree($1)) {
			$root = $1;
		}
	}

	if (!defined $root) {
		print "Must be run from the top-level dir. of a kernel tree\n";
		exit(2);
	}
96 97
}

98 99 100 101 102 103 104 105 106 107 108
my $emitted_corrupt = 0;

our $Ident       = qr{[A-Za-z_][A-Za-z\d_]*};
our $Storage	= qr{extern|static|asmlinkage};
our $Sparse	= qr{
			__user|
			__kernel|
			__force|
			__iomem|
			__must_check|
			__init_refok|
109
			__kprobes
110 111 112 113 114 115 116
		}x;
our $Attribute	= qr{
			const|
			__read_mostly|
			__kprobes|
			__(?:mem|cpu|dev|)(?:initdata|init)
		  }x;
117
our $Modifier;
118 119 120 121 122 123 124 125 126
our $Inline	= qr{inline|__always_inline|noinline};
our $Member	= qr{->$Ident|\.$Ident|\[[^]]*\]};
our $Lval	= qr{$Ident(?:$Member)*};

our $Constant	= qr{(?:[0-9]+|0x[0-9a-fA-F]+)[UL]*};
our $Assignment	= qr{(?:\*\=|/=|%=|\+=|-=|<<=|>>=|&=|\^=|\|=|=)};
our $Operators	= qr{
			<=|>=|==|!=|
			=>|->|<<|>>|<|>|!|~|
127
			&&|\|\||,|\^|\+\+|--|&|\||\+|-|\*|\/|%
128 129
		  }x;

130 131 132 133
our $NonptrType;
our $Type;
our $Declare;

134 135 136 137 138 139 140 141 142 143 144
our $UTF8	= qr {
	[\x09\x0A\x0D\x20-\x7E]              # ASCII
	| [\xC2-\xDF][\x80-\xBF]             # non-overlong 2-byte
	|  \xE0[\xA0-\xBF][\x80-\xBF]        # excluding overlongs
	| [\xE1-\xEC\xEE\xEF][\x80-\xBF]{2}  # straight 3-byte
	|  \xED[\x80-\x9F][\x80-\xBF]        # excluding surrogates
	|  \xF0[\x90-\xBF][\x80-\xBF]{2}     # planes 1-3
	| [\xF1-\xF3][\x80-\xBF]{3}          # planes 4-15
	|  \xF4[\x80-\x8F][\x80-\xBF]{2}     # plane 16
}x;

145 146
our @typeList = (
	qr{void},
147 148 149 150 151 152 153
	qr{(?:unsigned\s+)?char},
	qr{(?:unsigned\s+)?short},
	qr{(?:unsigned\s+)?int},
	qr{(?:unsigned\s+)?long},
	qr{(?:unsigned\s+)?long\s+int},
	qr{(?:unsigned\s+)?long\s+long},
	qr{(?:unsigned\s+)?long\s+long\s+int},
154 155 156 157 158 159 160 161 162 163 164 165
	qr{unsigned},
	qr{float},
	qr{double},
	qr{bool},
	qr{(?:__)?(?:u|s|be|le)(?:8|16|32|64)},
	qr{struct\s+$Ident},
	qr{union\s+$Ident},
	qr{enum\s+$Ident},
	qr{${Ident}_t},
	qr{${Ident}_handler},
	qr{${Ident}_handler_fn},
);
166 167 168
our @modifierList = (
	qr{fastcall},
);
169 170

sub build_types {
171 172
	my $mods = "(?x:  \n" . join("|\n  ", @modifierList) . "\n)";
	my $all = "(?x:  \n" . join("|\n  ", @typeList) . "\n)";
173
	$Modifier	= qr{(?:$Attribute|$Sparse|$mods)};
174
	$NonptrType	= qr{
175
			(?:$Modifier\s+|const\s+)*
176
			(?:
177 178
				(?:typeof|__typeof__)\s*\(\s*\**\s*$Ident\s*\)|
				(?:${all}\b)
179
			)
180
			(?:\s+$Modifier|\s+const)*
181 182
		  }x;
	$Type	= qr{
183
			$NonptrType
184
			(?:\s*\*+\s*const|\s*\*+|(?:\s*\[\s*\])+)?
185
			(?:\s+$Inline|\s+$Modifier)*
186 187 188 189
		  }x;
	$Declare	= qr{(?:$Storage\s+)?$Type};
}
build_types();
190 191 192

$chk_signoff = 0 if ($file);

193 194
my @dep_includes = ();
my @dep_functions = ();
195 196 197 198
my $removal = "Documentation/feature-removal-schedule.txt";
if ($tree && -f "$root/$removal") {
	open(REMOVE, "<$root/$removal") ||
				die "$P: $removal: open failed - $!\n";
199
	while (<REMOVE>) {
200 201 202
		if (/^Check:\s+(.*\S)/) {
			for my $entry (split(/[, ]+/, $1)) {
				if ($entry =~ m@include/(.*)@) {
203 204
					push(@dep_includes, $1);

205 206 207
				} elsif ($entry !~ m@/@) {
					push(@dep_functions, $entry);
				}
208
			}
209 210 211 212
		}
	}
}

213
my @rawlines = ();
214 215
my @lines = ();
my $vname;
216 217 218 219 220 221 222
for my $filename (@ARGV) {
	if ($file) {
		open(FILE, "diff -u /dev/null $filename|") ||
			die "$P: $filename: diff failed - $!\n";
	} else {
		open(FILE, "<$filename") ||
			die "$P: $filename: open failed - $!\n";
223
	}
224 225 226 227 228
	if ($filename eq '-') {
		$vname = 'Your patch';
	} else {
		$vname = $filename;
	}
229 230 231 232 233
	while (<FILE>) {
		chomp;
		push(@rawlines, $_);
	}
	close(FILE);
234
	if (!process($filename)) {
235 236 237
		$exit = 1;
	}
	@rawlines = ();
238
	@lines = ();
239 240 241 242 243
}

exit($exit);

sub top_of_kernel_tree {
244 245 246 247 248 249 250 251 252 253 254 255
	my ($root) = @_;

	my @tree_check = (
		"COPYING", "CREDITS", "Kbuild", "MAINTAINERS", "Makefile",
		"README", "Documentation", "arch", "include", "drivers",
		"fs", "init", "ipc", "kernel", "lib", "scripts",
	);

	foreach my $check (@tree_check) {
		if (! -e $root . '/' . $check) {
			return 0;
		}
256
	}
257
	return 1;
258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
}

sub expand_tabs {
	my ($str) = @_;

	my $res = '';
	my $n = 0;
	for my $c (split(//, $str)) {
		if ($c eq "\t") {
			$res .= ' ';
			$n++;
			for (; ($n % 8) != 0; $n++) {
				$res .= ' ';
			}
			next;
		}
		$res .= $c;
		$n++;
	}

	return $res;
}
280
sub copy_spacing {
281
	(my $res = shift) =~ tr/\t/ /c;
282 283
	return $res;
}
284

285 286 287 288 289 290 291 292 293 294 295 296 297
sub line_stats {
	my ($line) = @_;

	# Drop the diff line leader and expand tabs
	$line =~ s/^.//;
	$line = expand_tabs($line);

	# Pick the indent from the front of the line.
	my ($white) = ($line =~ /^(\s*)/);

	return (length($line), length($white));
}

298 299 300 301 302 303 304 305 306 307 308
my $sanitise_quote = '';

sub sanitise_line_reset {
	my ($in_comment) = @_;

	if ($in_comment) {
		$sanitise_quote = '*/';
	} else {
		$sanitise_quote = '';
	}
}
309 310 311 312 313 314
sub sanitise_line {
	my ($line) = @_;

	my $res = '';
	my $l = '';

315
	my $qlen = 0;
316 317
	my $off = 0;
	my $c;
318

319 320 321 322 323 324 325 326 327 328 329 330 331 332
	# Always copy over the diff marker.
	$res = substr($line, 0, 1);

	for ($off = 1; $off < length($line); $off++) {
		$c = substr($line, $off, 1);

		# Comments we are wacking completly including the begin
		# and end, all to $;.
		if ($sanitise_quote eq '' && substr($line, $off, 2) eq '/*') {
			$sanitise_quote = '*/';

			substr($res, $off, 2, "$;$;");
			$off++;
			next;
333
		}
334
		if (substr($line, $off, 2) eq '*/') {
335 336 337 338
			$sanitise_quote = '';
			substr($res, $off, 2, "$;$;");
			$off++;
			next;
339
		}
340 341 342 343 344 345 346

		# A \ in a string means ignore the next character.
		if (($sanitise_quote eq "'" || $sanitise_quote eq '"') &&
		    $c eq "\\") {
			substr($res, $off, 2, 'XX');
			$off++;
			next;
347
		}
348 349 350 351
		# Regular quotes.
		if ($c eq "'" || $c eq '"') {
			if ($sanitise_quote eq '') {
				$sanitise_quote = $c;
352

353 354 355 356 357 358
				substr($res, $off, 1, $c);
				next;
			} elsif ($sanitise_quote eq $c) {
				$sanitise_quote = '';
			}
		}
359

360 361 362 363 364 365 366 367
		#print "SQ:$sanitise_quote\n";
		if ($off != 0 && $sanitise_quote eq '*/' && $c ne "\t") {
			substr($res, $off, 1, $;);
		} elsif ($off != 0 && $sanitise_quote && $c ne "\t") {
			substr($res, $off, 1, 'X');
		} else {
			substr($res, $off, 1, $c);
		}
368 369 370
	}

	# The pathname on a #include may be surrounded by '<' and '>'.
371
	if ($res =~ /^.\s*\#\s*include\s+\<(.*)\>/) {
372 373 374 375
		my $clean = 'X' x length($1);
		$res =~ s@\<.*\>@<$clean>@;

	# The whole of a #error is a string.
376
	} elsif ($res =~ /^.\s*\#\s*(?:error|warning)\s+(.*)\b/) {
377
		my $clean = 'X' x length($1);
378
		$res =~ s@(\#\s*(?:error|warning)\s+).*@$1$clean@;
379 380
	}

381 382 383
	return $res;
}

384 385 386 387 388 389
sub ctx_statement_block {
	my ($linenr, $remain, $off) = @_;
	my $line = $linenr - 1;
	my $blk = '';
	my $soff = $off;
	my $coff = $off - 1;
390
	my $coff_set = 0;
391

392 393
	my $loff = 0;

394 395
	my $type = '';
	my $level = 0;
396
	my $p;
397 398
	my $c;
	my $len = 0;
399 400

	my $remainder;
401
	while (1) {
402
		#warn "CSB: blk<$blk> remain<$remain>\n";
403 404 405 406
		# If we are about to drop off the end, pull in more
		# context.
		if ($off >= $len) {
			for (; $remain > 0; $line++) {
407
				next if ($lines[$line] =~ /^-/);
408
				$remain--;
409
				$loff = $len;
410
				$blk .= $lines[$line] . "\n";
411 412 413 414 415 416
				$len = length($blk);
				$line++;
				last;
			}
			# Bail if there is no further context.
			#warn "CSB: blk<$blk> off<$off> len<$len>\n";
417
			if ($off >= $len) {
418 419 420
				last;
			}
		}
421
		$p = $c;
422
		$c = substr($blk, $off, 1);
423
		$remainder = substr($blk, $off);
424

425
		#warn "CSB: c<$c> type<$type> level<$level> remainder<$remainder> coff_set<$coff_set>\n";
426 427 428 429 430 431
		# Statement ends at the ';' or a close '}' at the
		# outermost level.
		if ($level == 0 && $c eq ';') {
			last;
		}

432
		# An else is really a conditional as long as its not else if
433 434 435 436 437 438 439 440
		if ($level == 0 && $coff_set == 0 &&
				(!defined($p) || $p =~ /(?:\s|\}|\+)/) &&
				$remainder =~ /^(else)(?:\s|{)/ &&
				$remainder !~ /^else\s+if\b/) {
			$coff = $off + length($1) - 1;
			$coff_set = 1;
			#warn "CSB: mark coff<$coff> soff<$soff> 1<$1>\n";
			#warn "[" . substr($blk, $soff, $coff - $soff + 1) . "]\n";
441 442
		}

443 444 445 446 447 448 449 450 451 452
		if (($type eq '' || $type eq '(') && $c eq '(') {
			$level++;
			$type = '(';
		}
		if ($type eq '(' && $c eq ')') {
			$level--;
			$type = ($level != 0)? '(' : '';

			if ($level == 0 && $coff < $soff) {
				$coff = $off;
453 454
				$coff_set = 1;
				#warn "CSB: mark coff<$coff>\n";
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470
			}
		}
		if (($type eq '' || $type eq '{') && $c eq '{') {
			$level++;
			$type = '{';
		}
		if ($type eq '{' && $c eq '}') {
			$level--;
			$type = ($level != 0)? '{' : '';

			if ($level == 0) {
				last;
			}
		}
		$off++;
	}
471
	# We are truly at the end, so shuffle to the next line.
472
	if ($off == $len) {
473
		$loff = $len + 1;
474 475 476
		$line++;
		$remain--;
	}
477 478 479 480 481 482 483

	my $statement = substr($blk, $soff, $off - $soff + 1);
	my $condition = substr($blk, $soff, $coff - $soff + 1);

	#warn "STATEMENT<$statement>\n";
	#warn "CONDITION<$condition>\n";

484
	#print "coff<$coff> soff<$off> loff<$loff>\n";
485 486 487 488 489

	return ($statement, $condition,
			$line, $remain + 1, $off - $loff + 1, $level);
}

490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
sub statement_lines {
	my ($stmt) = @_;

	# Strip the diff line prefixes and rip blank lines at start and end.
	$stmt =~ s/(^|\n)./$1/g;
	$stmt =~ s/^\s*//;
	$stmt =~ s/\s*$//;

	my @stmt_lines = ($stmt =~ /\n/g);

	return $#stmt_lines + 2;
}

sub statement_rawlines {
	my ($stmt) = @_;

	my @stmt_lines = ($stmt =~ /\n/g);

	return $#stmt_lines + 2;
}

sub statement_block_size {
	my ($stmt) = @_;

	$stmt =~ s/(^|\n)./$1/g;
	$stmt =~ s/^\s*{//;
	$stmt =~ s/}\s*$//;
	$stmt =~ s/^\s*//;
	$stmt =~ s/\s*$//;

	my @stmt_lines = ($stmt =~ /\n/g);
	my @stmt_statements = ($stmt =~ /;/g);

	my $stmt_lines = $#stmt_lines + 2;
	my $stmt_statements = $#stmt_statements + 1;

	if ($stmt_lines > $stmt_statements) {
		return $stmt_lines;
	} else {
		return $stmt_statements;
	}
}

533 534 535 536 537 538
sub ctx_statement_full {
	my ($linenr, $remain, $off) = @_;
	my ($statement, $condition, $level);

	my (@chunks);

539
	# Grab the first conditional/block pair.
540 541
	($statement, $condition, $linenr, $remain, $off, $level) =
				ctx_statement_block($linenr, $remain, $off);
542
	#print "F: c<$condition> s<$statement> remain<$remain>\n";
543 544 545 546 547 548 549
	push(@chunks, [ $condition, $statement ]);
	if (!($remain > 0 && $condition =~ /^\s*(?:\n[+-])?\s*(?:if|else|do)\b/s)) {
		return ($level, $linenr, @chunks);
	}

	# Pull in the following conditional/block pairs and see if they
	# could continue the statement.
550 551 552
	for (;;) {
		($statement, $condition, $linenr, $remain, $off, $level) =
				ctx_statement_block($linenr, $remain, $off);
553
		#print "C: c<$condition> s<$statement> remain<$remain>\n";
554
		last if (!($remain > 0 && $condition =~ /^(?:\s*\n[+-])*\s*(?:else|do)\b/s));
555 556
		#print "C: push\n";
		push(@chunks, [ $condition, $statement ]);
557 558 559
	}

	return ($level, $linenr, @chunks);
560 561
}

562
sub ctx_block_get {
563
	my ($linenr, $remain, $outer, $open, $close, $off) = @_;
564 565 566 567 568 569 570
	my $line;
	my $start = $linenr - 1;
	my $blk = '';
	my @o;
	my @c;
	my @res = ();

571
	my $level = 0;
572 573 574 575 576
	for ($line = $start; $remain > 0; $line++) {
		next if ($rawlines[$line] =~ /^-/);
		$remain--;

		$blk .= $rawlines[$line];
577 578 579 580 581 582
		foreach my $c (split(//, $rawlines[$line])) {
			##print "C<$c>L<$level><$open$close>O<$off>\n";
			if ($off > 0) {
				$off--;
				next;
			}
583

584 585 586 587 588 589 590
			if ($c eq $close && $level > 0) {
				$level--;
				last if ($level == 0);
			} elsif ($c eq $open) {
				$level++;
			}
		}
591

592
		if (!$outer || $level <= 1) {
593
			push(@res, $rawlines[$line]);
594 595
		}

596
		last if ($level == 0);
597 598
	}

599
	return ($level, @res);
600 601 602 603
}
sub ctx_block_outer {
	my ($linenr, $remain) = @_;

604 605
	my ($level, @r) = ctx_block_get($linenr, $remain, 1, '{', '}', 0);
	return @r;
606 607 608 609
}
sub ctx_block {
	my ($linenr, $remain) = @_;

610 611
	my ($level, @r) = ctx_block_get($linenr, $remain, 0, '{', '}', 0);
	return @r;
612 613
}
sub ctx_statement {
614 615 616 617 618 619
	my ($linenr, $remain, $off) = @_;

	my ($level, @r) = ctx_block_get($linenr, $remain, 0, '(', ')', $off);
	return @r;
}
sub ctx_block_level {
620 621
	my ($linenr, $remain) = @_;

622
	return ctx_block_get($linenr, $remain, 0, '{', '}', 0);
623
}
624 625 626 627 628
sub ctx_statement_level {
	my ($linenr, $remain, $off) = @_;

	return ctx_block_get($linenr, $remain, 0, '(', ')', $off);
}
629 630 631 632 633

sub ctx_locate_comment {
	my ($first_line, $end_line) = @_;

	# Catch a comment on the end of the line itself.
634
	my ($current_comment) = ($rawlines[$end_line - 1] =~ m@.*(/\*.*\*/)\s*(?:\\\s*)?$@);
635 636 637 638 639 640 641
	return $current_comment if (defined $current_comment);

	# Look through the context and try and figure out if there is a
	# comment.
	my $in_comment = 0;
	$current_comment = '';
	for (my $linenr = $first_line; $linenr < $end_line; $linenr++) {
642 643
		my $line = $rawlines[$linenr - 1];
		#warn "           $line\n";
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665
		if ($linenr == $first_line and $line =~ m@^.\s*\*@) {
			$in_comment = 1;
		}
		if ($line =~ m@/\*@) {
			$in_comment = 1;
		}
		if (!$in_comment && $current_comment ne '') {
			$current_comment = '';
		}
		$current_comment .= $line . "\n" if ($in_comment);
		if ($line =~ m@\*/@) {
			$in_comment = 0;
		}
	}

	chomp($current_comment);
	return($current_comment);
}
sub ctx_has_comment {
	my ($first_line, $end_line) = @_;
	my $cmt = ctx_locate_comment($first_line, $end_line);

666
	##print "LINE: $rawlines[$end_line - 1 ]\n";
667 668 669 670 671
	##print "CMMT: $cmt\n";

	return ($cmt ne '');
}

672 673 674
sub cat_vet {
	my ($vet) = @_;
	my ($res, $coded);
675

676 677 678 679 680 681
	$res = '';
	while ($vet =~ /([^[:cntrl:]]*)([[:cntrl:]]|$)/g) {
		$res .= $1;
		if ($2 ne '') {
			$coded = sprintf("^%c", unpack('C', $2) + 64);
			$res .= $coded;
682 683
		}
	}
684
	$res =~ s/$/\$/;
685

686
	return $res;
687 688
}

689
my $av_preprocessor = 0;
690
my $av_pending;
691 692 693 694
my @av_paren_type;

sub annotate_reset {
	$av_preprocessor = 0;
695 696
	$av_pending = '_';
	@av_paren_type = ('E');
697 698
}

699 700
sub annotate_values {
	my ($stream, $type) = @_;
701

702 703 704
	my $res;
	my $cur = $stream;

705
	print "$stream\n" if ($dbg_values > 1);
706 707

	while (length($cur)) {
708
		@av_paren_type = ('E') if ($#av_paren_type < 0);
709
		print " <" . join('', @av_paren_type) .
710
				"> <$type> <$av_pending>" if ($dbg_values > 1);
711
		if ($cur =~ /^(\s+)/o) {
712 713
			print "WS($1)\n" if ($dbg_values > 1);
			if ($1 =~ /\n/ && $av_preprocessor) {
714
				$type = pop(@av_paren_type);
715
				$av_preprocessor = 0;
716 717
			}

718
		} elsif ($cur =~ /^($Type)\s*(?:$Ident|,|\))/) {
719
			print "DECLARE($1)\n" if ($dbg_values > 1);
720 721
			$type = 'T';

722 723 724 725
		} elsif ($cur =~ /^($Modifier)\s*/) {
			print "MODIFIER($1)\n" if ($dbg_values > 1);
			$type = 'T';

726
		} elsif ($cur =~ /^(\#\s*define\s*$Ident)(\(?)/o) {
727
			print "DEFINE($1,$2)\n" if ($dbg_values > 1);
728
			$av_preprocessor = 1;
729 730 731 732 733 734
			push(@av_paren_type, $type);
			if ($2 ne '') {
				$av_pending = 'N';
			}
			$type = 'E';

735
		} elsif ($cur =~ /^(\#\s*(?:undef\s*$Ident|include\b))/o) {
736 737 738
			print "UNDEF($1)\n" if ($dbg_values > 1);
			$av_preprocessor = 1;
			push(@av_paren_type, $type);
739

740
		} elsif ($cur =~ /^(\#\s*(?:ifdef|ifndef|if))/o) {
741
			print "PRE_START($1)\n" if ($dbg_values > 1);
742
			$av_preprocessor = 1;
743 744 745

			push(@av_paren_type, $type);
			push(@av_paren_type, $type);
746
			$type = 'E';
747

748
		} elsif ($cur =~ /^(\#\s*(?:else|elif))/o) {
749 750 751 752 753
			print "PRE_RESTART($1)\n" if ($dbg_values > 1);
			$av_preprocessor = 1;

			push(@av_paren_type, $av_paren_type[$#av_paren_type]);

754
			$type = 'E';
755

756
		} elsif ($cur =~ /^(\#\s*(?:endif))/o) {
757 758 759 760 761 762 763 764
			print "PRE_END($1)\n" if ($dbg_values > 1);

			$av_preprocessor = 1;

			# Assume all arms of the conditional end as this
			# one does, and continue as if the #endif was not here.
			pop(@av_paren_type);
			push(@av_paren_type, $type);
765
			$type = 'E';
766 767

		} elsif ($cur =~ /^(\\\n)/o) {
768
			print "PRECONT($1)\n" if ($dbg_values > 1);
769

770 771 772 773 774
		} elsif ($cur =~ /^(__attribute__)\s*\(?/o) {
			print "ATTR($1)\n" if ($dbg_values > 1);
			$av_pending = $type;
			$type = 'N';

775
		} elsif ($cur =~ /^(sizeof)\s*(\()?/o) {
776
			print "SIZEOF($1)\n" if ($dbg_values > 1);
777
			if (defined $2) {
778
				$av_pending = 'V';
779 780 781
			}
			$type = 'N';

782
		} elsif ($cur =~ /^(if|while|typeof|__typeof__|for)\b/o) {
783
			print "COND($1)\n" if ($dbg_values > 1);
784
			$av_pending = 'N';
785 786
			$type = 'N';

787
		} elsif ($cur =~/^(return|case|else|goto)/o) {
788
			print "KEYWORD($1)\n" if ($dbg_values > 1);
789 790 791
			$type = 'N';

		} elsif ($cur =~ /^(\()/o) {
792
			print "PAREN('$1')\n" if ($dbg_values > 1);
793 794
			push(@av_paren_type, $av_pending);
			$av_pending = '_';
795 796 797
			$type = 'N';

		} elsif ($cur =~ /^(\))/o) {
798 799 800
			my $new_type = pop(@av_paren_type);
			if ($new_type ne '_') {
				$type = $new_type;
801 802
				print "PAREN('$1') -> $type\n"
							if ($dbg_values > 1);
803
			} else {
804
				print "PAREN('$1')\n" if ($dbg_values > 1);
805 806
			}

807
		} elsif ($cur =~ /^($Ident)\s*\(/o) {
808
			print "FUNC($1)\n" if ($dbg_values > 1);
809
			$type = 'V';
810
			$av_pending = 'V';
811 812

		} elsif ($cur =~ /^($Ident|$Constant)/o) {
813
			print "IDENT($1)\n" if ($dbg_values > 1);
814 815 816
			$type = 'V';

		} elsif ($cur =~ /^($Assignment)/o) {
817
			print "ASSIGN($1)\n" if ($dbg_values > 1);
818 819
			$type = 'N';

820
		} elsif ($cur =~/^(;|{|})/) {
821
			print "END($1)\n" if ($dbg_values > 1);
822 823
			$type = 'E';

824
		} elsif ($cur =~ /^(;|\?|:|\[)/o) {
825
			print "CLOSE($1)\n" if ($dbg_values > 1);
826 827 828
			$type = 'N';

		} elsif ($cur =~ /^($Operators)/o) {
829
			print "OP($1)\n" if ($dbg_values > 1);
830 831 832 833 834
			if ($1 ne '++' && $1 ne '--') {
				$type = 'N';
			}

		} elsif ($cur =~ /(^.)/o) {
835
			print "C($1)\n" if ($dbg_values > 1);
836 837 838 839 840
		}
		if (defined $1) {
			$cur = substr($cur, length($1));
			$res .= $type x length($1);
		}
841
	}
842

843
	return $res;
844 845
}

846
sub possible {
847
	my ($possible, $line) = @_;
848

849
	print "CHECK<$possible> ($line)\n" if ($dbg_possible > 1);
850 851 852
	if ($possible !~ /^(?:$Storage|$Type|DEFINE_\S+)$/ &&
	    $possible ne 'goto' && $possible ne 'return' &&
	    $possible ne 'case' && $possible ne 'else' &&
853
	    $possible ne 'asm' && $possible ne '__asm__' &&
854 855 856 857 858 859 860 861 862 863 864 865 866 867 868
	    $possible !~ /^(typedef|struct|enum)\b/) {
		# Check for modifiers.
		$possible =~ s/\s*$Storage\s*//g;
		$possible =~ s/\s*$Sparse\s*//g;
		if ($possible =~ /^\s*$/) {

		} elsif ($possible =~ /\s/) {
			$possible =~ s/\s*$Type\s*//g;
			warn "MODIFIER: $possible ($line)\n" if ($dbg_possible);
			push(@modifierList, $possible);

		} else {
			warn "POSSIBLE: $possible ($line)\n" if ($dbg_possible);
			push(@typeList, $possible);
		}
869 870 871 872
		build_types();
	}
}

873 874
my $prefix = '';

875
sub report {
876 877 878
	if (defined $tst_only && $_[0] !~ /\Q$tst_only\E/) {
		return 0;
	}
879 880 881 882
	my $line = $prefix . $_[0];

	$line = (split('\n', $line))[0] . "\n" if ($terse);

883
	push(our @report, $line);
884 885

	return 1;
886 887
}
sub report_dump {
888
	our @report;
889
}
890
sub ERROR {
891 892 893 894
	if (report("ERROR: $_[0]\n")) {
		our $clean = 0;
		our $cnt_error++;
	}
895 896
}
sub WARN {
897 898 899 900
	if (report("WARNING: $_[0]\n")) {
		our $clean = 0;
		our $cnt_warn++;
	}
901 902
}
sub CHK {
903
	if ($check && report("CHECK: $_[0]\n")) {
904 905 906
		our $clean = 0;
		our $cnt_chk++;
	}
907 908
}

909 910 911 912 913
sub process {
	my $filename = shift;

	my $linenr=0;
	my $prevline="";
914
	my $prevrawline="";
915
	my $stashline="";
916
	my $stashrawline="";
917

918
	my $length;
919 920 921 922
	my $indent;
	my $previndent=0;
	my $stashindent=0;

923
	our $clean = 1;
924 925 926
	my $signoff = 0;
	my $is_patch = 0;

927
	our @report = ();
928 929 930 931 932
	our $cnt_lines = 0;
	our $cnt_error = 0;
	our $cnt_warn = 0;
	our $cnt_chk = 0;

933 934 935 936 937 938
	# Trace the real file/line as we go.
	my $realfile = '';
	my $realline = 0;
	my $realcnt = 0;
	my $here = '';
	my $in_comment = 0;
939
	my $comment_edge = 0;
940 941
	my $first_line = 0;

942 943 944
	my $prev_values = 'E';

	# suppression flags
945
	my %suppress_ifbraces;
946

947
	# Pre-scan the patch sanitizing the lines.
948
	# Pre-scan the patch looking for any __setup documentation.
949
	#
950 951
	my @setup_docs = ();
	my $setup_docs = 0;
952 953

	sanitise_line_reset();
954 955
	my $line;
	foreach my $rawline (@rawlines) {
956 957
		$linenr++;
		$line = $rawline;
958

959
		if ($rawline=~/^\+\+\+\s+(\S+)/) {
960 961 962 963
			$setup_docs = 0;
			if ($1 =~ m@Documentation/kernel-parameters.txt$@) {
				$setup_docs = 1;
			}
964 965 966 967 968 969 970 971 972
			#next;
		}
		if ($rawline=~/^\@\@ -\d+(?:,\d+)? \+(\d+)(,(\d+))? \@\@/) {
			$realline=$1-1;
			if (defined $2) {
				$realcnt=$3+1;
			} else {
				$realcnt=1+1;
			}
973
			$in_comment = 0;
974 975 976 977 978 979

			# Guestimate if this is a continuing comment.  Run
			# the context looking for a comment "edge".  If this
			# edge is a close comment then we must be in a comment
			# at context start.
			my $edge;
980
			for (my $ln = $linenr + 1; $ln < ($linenr + $realcnt); $ln++) {
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
				next if ($line =~ /^-/);
				($edge) = ($rawlines[$ln - 1] =~ m@(/\*|\*/)@);
				last if (defined $edge);
			}
			if (defined $edge && $edge eq '*/') {
				$in_comment = 1;
			}

			# Guestimate if this is a continuing comment.  If this
			# is the start of a diff block and this line starts
			# ' *' then it is very likely a comment.
			if (!defined $edge &&
			    $rawlines[$linenr] =~ m@^.\s* \*(?:\s|$)@)
			{
				$in_comment = 1;
			}

			##print "COMMENT:$in_comment edge<$edge> $rawline\n";
			sanitise_line_reset($in_comment);

1001
		} elsif ($realcnt && $rawline =~ /^(?:\+| |$)/) {
1002
			# Standardise the strings and chars within the input to
1003
			# simplify matching -- only bother with positive lines.
1004
			$line = sanitise_line($rawline);
1005
		}
1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		push(@lines, $line);

		if ($realcnt > 1) {
			$realcnt-- if ($line =~ /^(?:\+| |$)/);
		} else {
			$realcnt = 0;
		}

		#print "==>$rawline\n";
		#print "-->$line\n";
1016 1017 1018 1019 1020 1021

		if ($setup_docs && $line =~ /^\+/) {
			push(@setup_docs, $line);
		}
	}

1022 1023
	$prefix = '';

1024 1025
	$realcnt = 0;
	$linenr = 0;
1026 1027 1028
	foreach my $line (@lines) {
		$linenr++;

1029
		my $rawline = $rawlines[$linenr - 1];
1030

1031
#extract the line range in the file after the patch is applied
1032
		if ($line=~/^\@\@ -\d+(?:,\d+)? \+(\d+)(,(\d+))? \@\@/) {
1033
			$is_patch = 1;
1034
			$first_line = $linenr + 1;
1035 1036 1037 1038 1039 1040
			$realline=$1-1;
			if (defined $2) {
				$realcnt=$3+1;
			} else {
				$realcnt=1+1;
			}
1041
			annotate_reset();
1042 1043
			$prev_values = 'E';

1044
			%suppress_ifbraces = ();
1045 1046
			next;

1047 1048 1049
# track the line number as we move through the hunk, note that
# new versions of GNU diff omit the leading space on completely
# blank context lines so we need to count that too.
1050
		} elsif ($line =~ /^( |\+|$)/) {
1051
			$realline++;
1052
			$realcnt-- if ($realcnt != 0);
1053

1054
			# Measure the line length and indent.
1055
			($length, $indent) = line_stats($rawline);
1056 1057 1058 1059

			# Track the previous line.
			($prevline, $stashline) = ($stashline, $line);
			($previndent, $stashindent) = ($stashindent, $indent);
1060 1061
			($prevrawline, $stashrawline) = ($stashrawline, $rawline);

1062
			#warn "line<$line>\n";
1063

1064 1065
		} elsif ($realcnt == 1) {
			$realcnt--;
1066 1067 1068
		}

#make up the handle for any error we report on this line
1069 1070 1071
		$prefix = "$filename:$realline: " if ($emacs && $file);
		$prefix = "$filename:$linenr: " if ($emacs && !$file);

1072 1073
		$here = "#$linenr: " if (!$file);
		$here = "#$realline: " if ($file);
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085

		# extract the filename as it passes
		if ($line=~/^\+\+\+\s+(\S+)/) {
			$realfile = $1;
			$realfile =~ s@^[^/]*/@@;

			if ($realfile =~ m@include/asm/@) {
				ERROR("do not modify files in include/asm, change architecture specific files in include/asm-<architecture>\n" . "$here$rawline\n");
			}
			next;
		}

1086
		$here .= "FILE: $realfile:$realline:" if ($realcnt != 0);
1087

1088 1089 1090
		my $hereline = "$here\n$rawline\n";
		my $herecurr = "$here\n$rawline\n";
		my $hereprev = "$here\n$prevrawline\n$rawline\n";
1091

1092 1093
		$cnt_lines++ if ($realcnt != 0);

1094
#check the patch for a signoff:
1095
		if ($line =~ /^\s*signed-off-by:/i) {
1096 1097
			# This is a signoff, if ugly, so do not double report.
			$signoff++;
1098
			if (!($line =~ /^\s*Signed-off-by:/)) {
1099 1100
				WARN("Signed-off-by: is the preferred form\n" .
					$herecurr);
1101 1102
			}
			if ($line =~ /^\s*signed-off-by:\S/i) {
1103
				WARN("space required after Signed-off-by:\n" .
1104
					$herecurr);
1105 1106 1107
			}
		}

1108
# Check for wrappage within a valid hunk of the file
1109
		if ($realcnt != 0 && $line !~ m{^(?:\+|-| |\\ No newline|$)}) {
1110
			ERROR("patch seems to be corrupt (line wrapped?)\n" .
1111
				$herecurr) if (!$emitted_corrupt++);
1112 1113 1114 1115
		}

# UTF-8 regex found at http://www.w3.org/International/questions/qa-forms-utf-8.en.php
		if (($realfile =~ /^$/ || $line =~ /^\+/) &&
1116 1117 1118 1119 1120 1121 1122 1123
		    $rawline !~ m/^$UTF8*$/) {
			my ($utf8_prefix) = ($rawline =~ /^($UTF8*)/);

			my $blank = copy_spacing($rawline);
			my $ptr = substr($blank, 0, length($utf8_prefix)) . "^";
			my $hereptr = "$hereline$ptr\n";

			ERROR("Invalid UTF-8, patch and commit message should be encoded in UTF-8\n" . $hereptr);
1124 1125 1126 1127
		}

#ignore lines being removed
		if ($line=~/^-/) {next;}
1128

1129 1130
# check we are in a valid source file if not then ignore this hunk
		next if ($realfile !~ /\.(h|c|s|S|pl|sh)$/);
1131 1132

#trailing whitespace
1133
		if ($line =~ /^\+.*\015/) {
1134
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
1135 1136
			ERROR("DOS line endings\n" . $herevet);

1137 1138
		} elsif ($rawline =~ /^\+.*\S\s+$/ || $rawline =~ /^\+\s+$/) {
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
1139
			ERROR("trailing whitespace\n" . $herevet);
1140 1141
		}
#80 column limit
1142
		if ($line =~ /^\+/ && $prevrawline !~ /\/\*\*/ &&
1143 1144 1145
		    $rawline !~ /^.\s*\*\s*\@$Ident\s/ &&
		    $line !~ /^\+\s*printk\s*\(\s*(?:KERN_\S+\s*)?"[X\t]*"\s*(?:,|\)\s*;)\s*$/ &&
		    $length > 80)
1146
		{
1147
			WARN("line over 80 characters\n" . $herecurr);
1148 1149
		}

1150 1151 1152 1153 1154
# check for adding lines without a newline.
		if ($line =~ /^\+/ && defined $lines[$linenr] && $lines[$linenr] =~ /^\\ No newline at end of file/) {
			WARN("adding a line without newline at end of file\n" . $herecurr);
		}

1155 1156 1157 1158 1159
# check we are in a valid source file *.[hc] if not then ignore this hunk
		next if ($realfile !~ /\.[hc]$/);

# at the beginning of a line any tabs must come first and anything
# more than 8 must use tabs.
1160 1161 1162
		if ($rawline =~ /^\+\s* \t\s*\S/ ||
		    $rawline =~ /^\+\s*        \s*/) {
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
1163
			ERROR("code indent should use tabs where possible\n" . $herevet);
1164 1165
		}

1166
# check for RCS/CVS revision markers
1167
		if ($rawline =~ /^\+.*\$(Revision|Log|Id)(?:\$|)/) {
1168 1169
			WARN("CVS style keyword markers, these will _not_ be updated\n". $herecurr);
		}
1170

1171
# Check for potential 'bare' types
1172
		my ($stat, $cond, $line_nr_next, $remain_next);
1173
		if ($realcnt && $line =~ /.\s*\S/) {
1174 1175
			($stat, $cond, $line_nr_next, $remain_next) =
				ctx_statement_block($linenr, $realcnt, 0);
1176 1177 1178 1179 1180
			$stat =~ s/\n./\n /g;
			$cond =~ s/\n./\n /g;

			my $s = $stat;
			$s =~ s/{.*$//s;
1181

1182
			# Ignore goto labels.
1183
			if ($s =~ /$Ident:\*$/s) {
1184 1185

			# Ignore functions being called