checkpatch.pl 44.6 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

12
my $V = '0.13';
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_type = 0;
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 $root;
28
my %debug;
29
GetOptions(
30
	'q|quiet+'	=> \$quiet,
31
32
33
	'tree!'		=> \$tree,
	'signoff!'	=> \$chk_signoff,
	'patch!'	=> \$chk_patch,
34
	'test-type!'	=> \$tst_type,
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
	'debug=s'	=> \%debug,
44
45
46
47
48
) or exit;

my $exit = 0;

if ($#ARGV < 0) {
49
	print "usage: $P [options] patchfile\n";
50
51
52
	print "version: $V\n";
	print "options: -q           => quiet\n";
	print "         --no-tree    => run without a kernel tree\n";
53
	print "         --terse      => one line per report\n";
54
55
56
57
	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";
58
59
60
	exit(1);
}

61
62
63
64
65
66
my $dbg_values = 0;
my $dbg_possible = 0;
for my $key (keys %debug) {
	eval "\${dbg_$key} = '$debug{$key}';"
}

67
68
69
70
71
if ($terse) {
	$emacs = 1;
	$quiet++;
}

72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
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);
	}
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
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|
			__kprobes|
			fastcall
		}x;
our $Attribute	= qr{
			const|
			__read_mostly|
			__kprobes|
			__(?:mem|cpu|dev|)(?:initdata|init)
		  }x;
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{
			<=|>=|==|!=|
			=>|->|<<|>>|<|>|!|~|
121
			&&|\|\||,|\^|\+\+|--|&|\||\+|-|\*|\/|%
122
123
		  }x;

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
our $NonptrType;
our $Type;
our $Declare;

our @typeList = (
	qr{void},
	qr{char},
	qr{short},
	qr{int},
	qr{long},
	qr{unsigned},
	qr{float},
	qr{double},
	qr{bool},
	qr{long\s+int},
	qr{long\s+long},
	qr{long\s+long\s+int},
	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},
);

sub build_types {
	my $all = "(?:  \n" . join("|\n  ", @typeList) . "\n)";
	$NonptrType	= qr{
			\b
			(?:const\s+)?
			(?:unsigned\s+)?
			$all
			(?:\s+$Sparse|\s+const)*
			\b
		  }x;
	$Type	= qr{
			\b$NonptrType\b
			(?:\s*\*+\s*const|\s*\*+|(?:\s*\[\s*\])+)?
163
			(?:\s+$Inline|\s+$Sparse|\s+$Attribute)*
164
165
166
167
		  }x;
	$Declare	= qr{(?:$Storage\s+)?$Type};
}
build_types();
168
169
170

$chk_signoff = 0 if ($file);

171
172
my @dep_includes = ();
my @dep_functions = ();
173
174
175
176
my $removal = "Documentation/feature-removal-schedule.txt";
if ($tree && -f "$root/$removal") {
	open(REMOVE, "<$root/$removal") ||
				die "$P: $removal: open failed - $!\n";
177
	while (<REMOVE>) {
178
179
180
		if (/^Check:\s+(.*\S)/) {
			for my $entry (split(/[, ]+/, $1)) {
				if ($entry =~ m@include/(.*)@) {
181
182
					push(@dep_includes, $1);

183
184
185
				} elsif ($entry !~ m@/@) {
					push(@dep_functions, $entry);
				}
186
			}
187
188
189
190
		}
	}
}

191
my @rawlines = ();
192
193
my @lines = ();
my $vname;
194
195
196
197
198
199
200
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 - $!\n";
201
	}
202
203
204
205
206
	if ($filename eq '-') {
		$vname = 'Your patch';
	} else {
		$vname = $filename;
	}
207
208
209
210
211
	while (<FILE>) {
		chomp;
		push(@rawlines, $_);
	}
	close(FILE);
212
	if (!process($filename)) {
213
214
215
		$exit = 1;
	}
	@rawlines = ();
216
217
218
219
220
}

exit($exit);

sub top_of_kernel_tree {
221
222
223
224
225
226
227
228
229
230
231
232
	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;
		}
233
	}
234
	return 1;
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
}

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 $res;
}
257
258
259
260
261
262
263
264
265
266
267
268
269
270
sub copy_spacing {
	my ($str) = @_;

	my $res = '';
	for my $c (split(//, $str)) {
		if ($c eq "\t") {
			$res .= $c;
		} else {
			$res .= ' ';
		}
	}

	return $res;
}
271

272
273
274
275
276
277
278
279
280
281
282
283
284
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));
}

285
286
287
288
289
290
291
sub sanitise_line {
	my ($line) = @_;

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

	my $quote = '';
292
	my $qlen = 0;
293
294

	foreach my $c (split(//, $line)) {
295
296
297
298
		# The second backslash of a pair is not a "quote".
		if ($l eq "\\" && $c eq "\\") {
			$c = 'X';
		}
299
300
301
302
303
		if ($l ne "\\" && ($c eq "'" || $c eq '"')) {
			if ($quote eq '') {
				$quote = $c;
				$res .= $c;
				$l = $c;
304
				$qlen = 0;
305
306
307
308
309
				next;
			} elsif ($quote eq $c) {
				$quote = '';
			}
		}
310
311
312
		if ($quote eq "'" && $qlen > 1) {
			$quote = '';
		}
313
314
		if ($quote && $c ne "\t") {
			$res .= "X";
315
			$qlen++;
316
317
318
319
320
321
322
		} else {
			$res .= $c;
		}

		$l = $c;
	}

323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
	# Clear out the comments.
	while ($res =~ m@(/\*.*?\*/)@) {
		substr($res, $-[1], $+[1] - $-[1]) = ' ' x ($+[1] - $-[1]);
	}
	if ($res =~ m@(/\*.*)@) {
		substr($res, $-[1], $+[1] - $-[1]) = ' ' x ($+[1] - $-[1]);
	}
	if ($res =~ m@^.(.*\*/)@) {
		substr($res, $-[1], $+[1] - $-[1]) = ' ' x ($+[1] - $-[1]);
	}

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

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

345
346
347
	return $res;
}

348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
sub ctx_statement_block {
	my ($linenr, $remain, $off) = @_;
	my $line = $linenr - 1;
	my $blk = '';
	my $soff = $off;
	my $coff = $off - 1;

	my $type = '';
	my $level = 0;
	my $c;
	my $len = 0;
	while (1) {
		#warn "CSB: blk<$blk>\n";
		# If we are about to drop off the end, pull in more
		# context.
		if ($off >= $len) {
			for (; $remain > 0; $line++) {
365
				next if ($lines[$line] =~ /^-/);
366
				$remain--;
367
				$blk .= $lines[$line] . "\n";
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
				$len = length($blk);
				$line++;
				last;
			}
			# Bail if there is no further context.
			#warn "CSB: blk<$blk> off<$off> len<$len>\n";
			if ($off == $len) {
				last;
			}
		}
		$c = substr($blk, $off, 1);

		#warn "CSB: c<$c> type<$type> level<$level>\n";
		# Statement ends at the ';' or a close '}' at the
		# outermost level.
		if ($level == 0 && $c eq ';') {
			last;
		}

		if (($type eq '' || $type eq '(') && $c eq '(') {
			$level++;
			$type = '(';
		}
		if ($type eq '(' && $c eq ')') {
			$level--;
			$type = ($level != 0)? '(' : '';

			if ($level == 0 && $coff < $soff) {
				$coff = $off;
			}
		}
		if (($type eq '' || $type eq '{') && $c eq '{') {
			$level++;
			$type = '{';
		}
		if ($type eq '{' && $c eq '}') {
			$level--;
			$type = ($level != 0)? '{' : '';

			if ($level == 0) {
				last;
			}
		}
		$off++;
	}

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

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

	return ($statement, $condition);
}

423
sub ctx_block_get {
424
	my ($linenr, $remain, $outer, $open, $close, $off) = @_;
425
426
427
428
429
430
431
	my $line;
	my $start = $linenr - 1;
	my $blk = '';
	my @o;
	my @c;
	my @res = ();

432
	my $level = 0;
433
434
435
436
437
	for ($line = $start; $remain > 0; $line++) {
		next if ($rawlines[$line] =~ /^-/);
		$remain--;

		$blk .= $rawlines[$line];
438
439
440
441
442
443
		foreach my $c (split(//, $rawlines[$line])) {
			##print "C<$c>L<$level><$open$close>O<$off>\n";
			if ($off > 0) {
				$off--;
				next;
			}
444

445
446
447
448
449
450
451
			if ($c eq $close && $level > 0) {
				$level--;
				last if ($level == 0);
			} elsif ($c eq $open) {
				$level++;
			}
		}
452

453
		if (!$outer || $level <= 1) {
454
			push(@res, $rawlines[$line]);
455
456
		}

457
		last if ($level == 0);
458
459
	}

460
	return ($level, @res);
461
462
463
464
}
sub ctx_block_outer {
	my ($linenr, $remain) = @_;

465
466
	my ($level, @r) = ctx_block_get($linenr, $remain, 1, '{', '}', 0);
	return @r;
467
468
469
470
}
sub ctx_block {
	my ($linenr, $remain) = @_;

471
472
	my ($level, @r) = ctx_block_get($linenr, $remain, 0, '{', '}', 0);
	return @r;
473
474
}
sub ctx_statement {
475
476
477
478
479
480
	my ($linenr, $remain, $off) = @_;

	my ($level, @r) = ctx_block_get($linenr, $remain, 0, '(', ')', $off);
	return @r;
}
sub ctx_block_level {
481
482
	my ($linenr, $remain) = @_;

483
	return ctx_block_get($linenr, $remain, 0, '{', '}', 0);
484
}
485
486
487
488
489
sub ctx_statement_level {
	my ($linenr, $remain, $off) = @_;

	return ctx_block_get($linenr, $remain, 0, '(', ')', $off);
}
490
491
492
493
494

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

	# Catch a comment on the end of the line itself.
495
	my ($current_comment) = ($rawlines[$end_line - 1] =~ m@.*(/\*.*\*/)\s*$@);
496
497
498
499
500
501
502
	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++) {
503
504
		my $line = $rawlines[$linenr - 1];
		#warn "           $line\n";
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
		if ($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);

527
	##print "LINE: $rawlines[$end_line - 1 ]\n";
528
529
530
531
532
	##print "CMMT: $cmt\n";

	return ($cmt ne '');
}

533
534
535
sub cat_vet {
	my ($vet) = @_;
	my ($res, $coded);
536

537
538
539
540
541
542
	$res = '';
	while ($vet =~ /([^[:cntrl:]]*)([[:cntrl:]]|$)/g) {
		$res .= $1;
		if ($2 ne '') {
			$coded = sprintf("^%c", unpack('C', $2) + 64);
			$res .= $coded;
543
544
		}
	}
545
	$res =~ s/$/\$/;
546

547
	return $res;
548
549
}

550
551
552
553
554
555
556
557
558
559
my $av_preprocessor = 0;
my $av_paren = 0;
my @av_paren_type;

sub annotate_reset {
	$av_preprocessor = 0;
	$av_paren = 0;
	@av_paren_type = ();
}

560
561
sub annotate_values {
	my ($stream, $type) = @_;
562

563
564
565
	my $res;
	my $cur = $stream;

566
	print "$stream\n" if ($dbg_values > 1);
567
568

	while (length($cur)) {
569
		print " <$type> " if ($dbg_values > 1);
570
		if ($cur =~ /^(\s+)/o) {
571
572
573
			print "WS($1)\n" if ($dbg_values > 1);
			if ($1 =~ /\n/ && $av_preprocessor) {
				$av_preprocessor = 0;
574
575
576
				$type = 'N';
			}

577
		} elsif ($cur =~ /^($Type)/) {
578
			print "DECLARE($1)\n" if ($dbg_values > 1);
579
580
581
			$type = 'T';

		} elsif ($cur =~ /^(#\s*define\s*$Ident)(\(?)/o) {
582
583
584
			print "DEFINE($1)\n" if ($dbg_values > 1);
			$av_preprocessor = 1;
			$av_paren_type[$av_paren] = 'N';
585

586
587
588
		} elsif ($cur =~ /^(#\s*(?:ifdef|ifndef|if|else|elif|endif))/o) {
			print "PRE($1)\n" if ($dbg_values > 1);
			$av_preprocessor = 1;
589
590
591
			$type = 'N';

		} elsif ($cur =~ /^(\\\n)/o) {
592
			print "PRECONT($1)\n" if ($dbg_values > 1);
593
594

		} elsif ($cur =~ /^(sizeof)\s*(\()?/o) {
595
			print "SIZEOF($1)\n" if ($dbg_values > 1);
596
			if (defined $2) {
597
				$av_paren_type[$av_paren] = 'V';
598
599
600
			}
			$type = 'N';

601
		} elsif ($cur =~ /^(if|while|typeof|for)\b/o) {
602
603
			print "COND($1)\n" if ($dbg_values > 1);
			$av_paren_type[$av_paren] = 'N';
604
605
606
			$type = 'N';

		} elsif ($cur =~/^(return|case|else)/o) {
607
			print "KEYWORD($1)\n" if ($dbg_values > 1);
608
609
610
			$type = 'N';

		} elsif ($cur =~ /^(\()/o) {
611
612
			print "PAREN('$1')\n" if ($dbg_values > 1);
			$av_paren++;
613
614
615
			$type = 'N';

		} elsif ($cur =~ /^(\))/o) {
616
617
618
619
620
621
			$av_paren-- if ($av_paren > 0);
			if (defined $av_paren_type[$av_paren]) {
				$type = $av_paren_type[$av_paren];
				undef $av_paren_type[$av_paren];
				print "PAREN('$1') -> $type\n"
							if ($dbg_values > 1);
622
			} else {
623
				print "PAREN('$1')\n" if ($dbg_values > 1);
624
625
626
			}

		} elsif ($cur =~ /^($Ident)\(/o) {
627
628
			print "FUNC($1)\n" if ($dbg_values > 1);
			$av_paren_type[$av_paren] = 'V';
629
630

		} elsif ($cur =~ /^($Ident|$Constant)/o) {
631
			print "IDENT($1)\n" if ($dbg_values > 1);
632
633
634
			$type = 'V';

		} elsif ($cur =~ /^($Assignment)/o) {
635
			print "ASSIGN($1)\n" if ($dbg_values > 1);
636
637
638
			$type = 'N';

		} elsif ($cur =~ /^(;|{|}|\?|:|\[)/o) {
639
			print "END($1)\n" if ($dbg_values > 1);
640
641
642
			$type = 'N';

		} elsif ($cur =~ /^($Operators)/o) {
643
			print "OP($1)\n" if ($dbg_values > 1);
644
645
646
647
648
			if ($1 ne '++' && $1 ne '--') {
				$type = 'N';
			}

		} elsif ($cur =~ /(^.)/o) {
649
			print "C($1)\n" if ($dbg_values > 1);
650
651
652
653
654
		}
		if (defined $1) {
			$cur = substr($cur, length($1));
			$res .= $type x length($1);
		}
655
	}
656

657
	return $res;
658
659
}

660
661
662
663
664
665
666
667
668
sub possible {
	my ($possible) = @_;

	#print "CHECK<$possible>\n";
	if ($possible !~ /^(?:$Storage|$Type|DEFINE_\S+)$/ &&
	    $possible ne 'goto' && $possible ne 'return' &&
	    $possible ne 'struct' && $possible ne 'enum' &&
	    $possible ne 'case' && $possible ne 'else' &&
	    $possible ne 'typedef') {
669
		warn "POSSIBLE: $possible\n" if ($dbg_possible);
670
671
672
673
674
		push(@typeList, $possible);
		build_types();
	}
}

675
676
my $prefix = '';

677
678
my @report = ();
sub report {
679
680
681
682
683
	my $line = $prefix . $_[0];

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

	push(@report, $line);
684
685
686
687
}
sub report_dump {
	@report;
}
688
sub ERROR {
689
	report("ERROR: $_[0]\n");
690
	our $clean = 0;
691
	our $cnt_error++;
692
693
}
sub WARN {
694
	report("WARNING: $_[0]\n");
695
	our $clean = 0;
696
	our $cnt_warn++;
697
698
}
sub CHK {
699
700
701
702
703
	if ($check) {
		report("CHECK: $_[0]\n");
		our $clean = 0;
		our $cnt_chk++;
	}
704
705
}

706
707
708
709
710
sub process {
	my $filename = shift;

	my $linenr=0;
	my $prevline="";
711
	my $prevrawline="";
712
	my $stashline="";
713
	my $stashrawline="";
714

715
	my $length;
716
717
718
719
	my $indent;
	my $previndent=0;
	my $stashindent=0;

720
	our $clean = 1;
721
722
723
	my $signoff = 0;
	my $is_patch = 0;

724
725
726
727
728
	our $cnt_lines = 0;
	our $cnt_error = 0;
	our $cnt_warn = 0;
	our $cnt_chk = 0;

729
730
731
732
733
734
	# Trace the real file/line as we go.
	my $realfile = '';
	my $realline = 0;
	my $realcnt = 0;
	my $here = '';
	my $in_comment = 0;
735
	my $comment_edge = 0;
736
737
	my $first_line = 0;

738
	my $prev_values = 'N';
739

740
	# Pre-scan the patch sanitizing the lines.
741
	# Pre-scan the patch looking for any __setup documentation.
742
	#
743
744
	my @setup_docs = ();
	my $setup_docs = 0;
745
746
747
748
749
750
751
752
753
754
	my $line;
	foreach my $rawline (@rawlines) {
		# Standardise the strings and chars within the input to
		# simplify matching.
		$line = sanitise_line($rawline);
		push(@lines, $line);

		##print "==>$rawline\n";
		##print "-->$line\n";

755
756
757
758
759
760
761
762
763
764
765
766
767
		if ($line=~/^\+\+\+\s+(\S+)/) {
			$setup_docs = 0;
			if ($1 =~ m@Documentation/kernel-parameters.txt$@) {
				$setup_docs = 1;
			}
			next;
		}

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

768
769
	$prefix = '';

770
771
772
	foreach my $line (@lines) {
		$linenr++;

773
		my $rawline = $rawlines[$linenr - 1];
774

775
776
777
#extract the filename as it passes
		if ($line=~/^\+\+\+\s+(\S+)/) {
			$realfile=$1;
778
			$realfile =~ s@^[^/]*/@@;
779
780
781
782
			$in_comment = 0;
			next;
		}
#extract the line range in the file after the patch is applied
783
		if ($line=~/^\@\@ -\d+(?:,\d+)? \+(\d+)(,(\d+))? \@\@/) {
784
			$is_patch = 1;
785
			$first_line = $linenr + 1;
786
787
788
789
790
791
792
			$in_comment = 0;
			$realline=$1-1;
			if (defined $2) {
				$realcnt=$3+1;
			} else {
				$realcnt=1+1;
			}
793
			annotate_reset();
794
			$prev_values = 'N';
795
796
797
			next;
		}

798
799
800
801
# 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.
		if ($line =~ /^( |\+|$)/) {
802
			$realline++;
803
			$realcnt-- if ($realcnt != 0);
804

805
806
807
808
809
810
811
			# 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.
			if ($linenr == $first_line) {
				my $edge;
				for (my $ln = $first_line; $ln < ($linenr + $realcnt); $ln++) {
812
					($edge) = ($rawlines[$ln - 1] =~ m@(/\*|\*/)@);
813
814
815
816
817
818
819
					last if (defined $edge);
				}
				if (defined $edge && $edge eq '*/') {
					$in_comment = 1;
				}
			}

820
821
822
			# 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.
823
			if ($linenr == $first_line and $rawline =~ m@^.\s* \*(?:\s|$)@) {
824
825
				$in_comment = 1;
			}
826
827

			# Find the last comment edge on _this_ line.
828
829
			$comment_edge = 0;
			while (($rawline =~ m@(/\*|\*/)@g)) {
830
831
832
833
834
				if ($1 eq '/*') {
					$in_comment = 1;
				} else {
					$in_comment = 0;
				}
835
				$comment_edge = 1;
836
837
			}

838
			# Measure the line length and indent.
839
			($length, $indent) = line_stats($rawline);
840
841
842
843

			# Track the previous line.
			($prevline, $stashline) = ($stashline, $line);
			($previndent, $stashindent) = ($stashindent, $indent);
844
845
846
			($prevrawline, $stashrawline) = ($stashrawline, $rawline);

			#warn "ic<$in_comment> ce<$comment_edge> line<$line>\n";
847

848
849
		} elsif ($realcnt == 1) {
			$realcnt--;
850
851
852
		}

#make up the handle for any error we report on this line
853
854
		$here = "#$linenr: " if (!$file);
		$here = "#$realline: " if ($file);
855
		$here .= "FILE: $realfile:$realline:" if ($realcnt != 0);
856

857
858
859
		my $hereline = "$here\n$rawline\n";
		my $herecurr = "$here\n$rawline\n";
		my $hereprev = "$here\n$prevrawline\n$rawline\n";
860

861
862
863
864
		$prefix = "$filename:$realline: " if ($emacs && $file);
		$prefix = "$filename:$linenr: " if ($emacs && !$file);
		$cnt_lines++ if ($realcnt != 0);

865
#check the patch for a signoff:
866
		if ($line =~ /^\s*signed-off-by:/i) {
867
868
			# This is a signoff, if ugly, so do not double report.
			$signoff++;
869
			if (!($line =~ /^\s*Signed-off-by:/)) {
870
871
				WARN("Signed-off-by: is the preferred form\n" .
					$herecurr);
872
873
			}
			if ($line =~ /^\s*signed-off-by:\S/i) {
874
875
				WARN("need space after Signed-off-by:\n" .
					$herecurr);
876
877
878
			}
		}

879
# Check for wrappage within a valid hunk of the file
880
		if ($realcnt != 0 && $line !~ m{^(?:\+|-| |\\ No newline|$)}) {
881
			ERROR("patch seems to be corrupt (line wrapped?)\n" .
882
				$herecurr) if (!$emitted_corrupt++);
883
884
885
886
		}

# UTF-8 regex found at http://www.w3.org/International/questions/qa-forms-utf-8.en.php
		if (($realfile =~ /^$/ || $line =~ /^\+/) &&
887
		     !($rawline =~ m/^(
888
889
890
891
892
893
894
895
896
				[\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 )) {
897
			ERROR("Invalid UTF-8, patch and commit message should be encoded in UTF-8\n" . $herecurr);
898
899
900
901
		}

#ignore lines being removed
		if ($line=~/^-/) {next;}
902

903
904
# check we are in a valid source file if not then ignore this hunk
		next if ($realfile !~ /\.(h|c|s|S|pl|sh)$/);
905
906

#trailing whitespace
907
		if ($line =~ /^\+.*\015/) {
908
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
909
910
			ERROR("DOS line endings\n" . $herevet);

911
912
		} elsif ($rawline =~ /^\+.*\S\s+$/ || $rawline =~ /^\+\s+$/) {
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
913
			ERROR("trailing whitespace\n" . $herevet);
914
915
		}
#80 column limit
916
		if ($line =~ /^\+/ && !($prevrawline=~/\/\*\*/) && $length > 80) {
917
			WARN("line over 80 characters\n" . $herecurr);
918
919
		}

920
921
922
923
924
# 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);
		}

925
926
927
928
929
# 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.
930
931
932
		if ($rawline =~ /^\+\s* \t\s*\S/ ||
		    $rawline =~ /^\+\s*        \s*/) {
			my $herevet = "$here\n" . cat_vet($rawline) . "\n";
933
			ERROR("use tabs not spaces\n" . $herevet);
934
935
		}

936
937
938
939
# check for RCS/CVS revision markers
		if ($rawline =~ /\$(Revision|Log|Id)(?:\$|)/) {
			WARN("CVS style keyword markers, these will _not_ be updated\n". $herecurr);
		}
940
941
942
943
944

# The rest of our checks refer specifically to C style
# only apply those _outside_ comments.  Only skip
# lines in the middle of comments.
		next if (!$comment_edge && $in_comment);
945

946
# Check for potential 'bare' types
947
948
949
950
951
952
953
		if ($realcnt) {
			# Ignore goto labels.
			if ($line =~ /$Ident:\*$/) {

			# Ignore functions being called
			} elsif ($line =~ /^.\s*$Ident\s*\(/) {

954
			# definitions in global scope can only start with types
955
			} elsif ($line =~ /^.(?:$Storage\s+)?(?:$Inline\s+)?(?:const\s+)?($Ident)\b/) {
956
957
958
				possible($1);

			# declarations always start with types
959
			} elsif ($prev_values eq 'N' && $line =~ /^.\s*(?:$Storage\s+)?(?:const\s+)?($Ident)\b(:?\s+$Sparse)?\s*\**\s*$Ident\s*(?:;|=)/) {
960
				possible($1);
961
			}
962
963

			# any (foo ... *) is a pointer cast, and foo is a type
964
			while ($line =~ /\(($Ident)(?:\s+$Sparse)*\s*\*+\s*\)/g) {
965
966
967
968
969
970
				possible($1);
			}

			# Check for any sort of function declaration.
			# int foo(something bar, other baz);
			# void (*store_gdt)(x86_descr_ptr *);
971
			if ($prev_values eq 'N' && $line =~ /^(.(?:typedef\s*)?(?:(?:$Storage|$Inline)\s*)*\s*$Type\s*(?:\b$Ident|\(\*\s*$Ident\))\s*)\(/) {
972
973
974
975
976
977
978
979
980
981
982
983
984
				my ($name_len) = length($1);
				my ($level, @ctx) = ctx_statement_level($linenr, $realcnt, $name_len);
				my $ctx = join("\n", @ctx);

				$ctx =~ s/\n.//;
				substr($ctx, 0, $name_len + 1) = '';
				$ctx =~ s/\)[^\)]*$//;
				for my $arg (split(/\s*,\s*/, $ctx)) {
					if ($arg =~ /^(?:const\s+)?($Ident)(?:\s+$Sparse)*\s*\**\s*(:?\b$Ident)?$/ || $arg =~ /^($Ident)$/) {

						possible($1);
					}
				}
985
			}
986

987
988
		}

989
990
991
#
# Checks which may be anchored in the context.
#
992

993
994
# Check for switch () and associated case and default
# statements should be at the same indent.
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
		if ($line=~/\bswitch\s*\(.*\)/) {
			my $err = '';
			my $sep = '';
			my @ctx = ctx_block_outer($linenr, $realcnt);
			shift(@ctx);
			for my $ctx (@ctx) {
				my ($clen, $cindent) = line_stats($ctx);
				if ($ctx =~ /^\+\s*(case\s+|default:)/ &&
							$indent != $cindent) {
					$err .= "$sep$ctx\n";
					$sep = '';
				} else {
					$sep = "[...]\n";
				}
			}
			if ($err ne '') {
1011
				ERROR("switch and case should be at the same indent\n$hereline$err");
1012
1013
1014
1015
1016
1017
			}
		}

# if/while/etc brace do not go on next line, unless defining a do while loop,
# or if that brace on the next line is for something else
		if ($line =~ /\b(?:(if|while|for|switch)\s*\(|do\b|else\b)/ && $line !~ /^.#/) {
1018
			my ($level, @ctx) = ctx_statement_level($linenr, $realcnt, 0);
1019
1020
1021
1022
			my $ctx_ln = $linenr + $#ctx + 1;
			my $ctx_cnt = $realcnt - $#ctx - 1;
			my $ctx = join("\n", @ctx);

1023
			# Skip over any removed lines in the context following statement.
1024
1025
1026
1027
1028
1029
1030
			while ($ctx_cnt > 0 && $lines[$ctx_ln - 1] =~ /^-/) {
				$ctx_ln++;
				$ctx_cnt--;
			}
			##warn "line<$line>\nctx<$ctx>\nnext<$lines[$ctx_ln - 1]>";

			if ($ctx !~ /{\s*/ && $ctx_cnt > 0 && $lines[$ctx_ln - 1] =~ /^\+\s*{/) {
1031
				ERROR("That open brace { should be on the previous line\n" .
1032
					"$here\n$ctx\n$lines[$ctx_ln - 1]");
1033
			}
1034
1035
1036
1037
1038
1039
1040
			if ($level == 0 && $ctx =~ /\)\s*\;\s*$/ && defined $lines[$ctx_ln - 1]) {
				my ($nlength, $nindent) = line_stats($lines[$ctx_ln - 1]);
				if ($nindent > $indent) {
					WARN("Trailing semicolon indicates no statements, indent implies otherwise\n" .
						"$here\n$ctx\n$lines[$ctx_ln - 1]");
				}
			}
1041
1042
		}

1043
1044
1045
1046
		# Track the 'values' across context and added lines.
		my $opline = $line; $opline =~ s/^./ /;
		my $curr_values = annotate_values($opline . "\n", $prev_values);
		$curr_values = $prev_values . $curr_values;
1047
1048
1049
1050
1051
		if ($dbg_values) {
			my $outline = $opline; $outline =~ s/\t/ /g;
			warn "--> .$outline\n";
			warn "--> $curr_values\n";
		}
1052
1053
		$prev_values = substr($curr_values, -1);

1054
1055
1056
#ignore lines not being added
		if ($line=~/^[^\+]/) {next;}

1057
1058
# TEST: allow direct testing of the type matcher.
		if ($tst_type && $line =~ /^.$Declare$/) {
1059
			ERROR("TEST: is type $Declare\n" . $herecurr);
1060
1061
1062
			next;
		}

1063
1064
1065
1066
1067
1068
# check for initialisation to aggregates open brace on the next line
		if ($prevline =~ /$Declare\s*$Ident\s*=\s*$/ &&
		    $line =~ /^.\s*{/) {
			ERROR("That open brace { should be on the previous line\n" . $hereprev);
		}

1069
1070
1071
1072
1073
1074
1075
1076
#
# Checks which are anchored on the added line.
#

# check for malformed paths in #include statements (uses RAW line)
		if ($rawline =~ m{^.#\s*include\s+[<"](.*)[">]}) {
			my $path = $1;
			if ($path =~ m{//}) {
1077
1078
				ERROR("malformed #include filename\n" .
					$herecurr);
1079
1080
			}
		}
1081

1082
# no C99 // comments
1083
		if ($line =~ m{//}) {
1084
			ERROR("do not use C99 // comments\n" . $herecurr);
1085
		}
1086
		# Remove C99 comments.
1087
		$line =~ s@//.*@@;
1088
		$opline =~ s@//.*@@;
1089
1090

#EXPORT_SYMBOL should immediately follow its function closing }.
1091
1092
1093
		if (($line =~ /EXPORT_SYMBOL.*\((.*)\)/) ||
		    ($line =~ /EXPORT_UNUSED_SYMBOL.*\((.*)\)/)) {
			my $name = $1;
1094
1095
			if (($prevline !~ /^}/) &&
			   ($prevline !~ /^\+}/) &&
1096
			   ($prevline !~ /^ }/) &&
1097
			   ($prevline !~ /\b\Q$name\E(?:\s+$Attribute)?\s*(?:;|=)/)) {
1098
				WARN("EXPORT_SYMBOL(foo); should immediately follow its function/variable\n" . $herecurr);
1099
1100
1101
			}
		}

1102
1103
1104
1105
1106
# check for external initialisers.
		if ($line =~ /^.$Type\s*$Ident\s*=\s*(0|NULL);/) {
			ERROR("do not initialise externals to 0 or NULL\n" .
				$herecurr);
		}
1107
# check for static initialisers.
1108
		if ($line =~ /\s*static\s.*=\s*(0|NULL);/) {
1109
1110
			ERROR("do not initialise statics to 0 or NULL\n" .
				$herecurr);
1111
1112
		}

1113
1114
1115
# check for new typedefs, only function parameters and sparse annotations
# make sense.
		if ($line =~ /\btypedef\s/ &&
1116
		    $line !~ /\btypedef\s+$Type\s+\(\s*\*?$Ident\s*\)\s*\(/ &&
1117
		    $line !~ /\b__bitwise(?:__|)\b/) {
1118
			WARN("do not add new typedefs\n" . $herecurr);
1119
1120
1121
		}

# * goes on variable not on type
1122
		if ($line =~ m{\($NonptrType(\*+)(?:\s+const)?\)}) {
1123
1124
			ERROR("\"(foo$1)\" should be \"(foo $1)\"\n" .
				$herecurr);
1125
1126

		} elsif ($line =~ m{\($NonptrType\s+(\*+)(?!\s+const)\s+\)}) {
1127
1128
			ERROR("\"(foo $1 )\" should be \"(foo $1)\"\n" .
				$herecurr);
1129

1130
		} elsif ($line =~ m{$NonptrType(\*+)(?:\s+(?:$Attribute|$Sparse))?\s+[A-Za-z\d_]+}) {
1131
1132
			ERROR("\"foo$1 bar\" should be \"foo $1bar\"\n" .
				$herecurr);
1133

1134
		} elsif ($line =~ m{$NonptrType\s+(\*+)(?!\s+(?:$Attribute|$Sparse))\s+[A-Za-z\d_]+}) {
1135
1136
			ERROR("\"foo $1 bar\" should be \"foo $1bar\"\n" .
				$herecurr);
1137
1138
1139
1140
1141
1142
1143
1144
1145
		}

# # no BUG() or BUG_ON()
# 		if ($line =~ /\b(BUG|BUG_ON)\b/) {
# 			print "Try to use WARN_ON & Recovery code rather than BUG() or BUG_ON()\n";
# 			print "$herecurr";
# 			$clean = 0;
# 		}

1146
		if ($line =~ /\bLINUX_VERSION_CODE\b/) {
1147
			WARN("LINUX_VERSION_CODE should be avoided, code should be for the version to which it is merged\n" . $herecurr);
1148
1149
		}

1150
1151
1152
1153
1154
# printk should use KERN_* levels.  Note that follow on printk's on the
# same line do not need a level, so we use the current block context
# to try and find and validate the current printk.  In summary the current
# printk includes all preceeding printk's which have no newline on the end.
# we assume the first bad printk is the one to report.
1155
		if ($line =~ /\bprintk\((?!KERN_)\s*"/) {
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
			my $ok = 0;
			for (my $ln = $linenr - 1; $ln >= $first_line; $ln--) {
				#print "CHECK<$lines[$ln - 1]\n";
				# we have a preceeding printk if it ends
				# with "\n" ignore it, else it is to blame
				if ($lines[$ln - 1] =~ m{\bprintk\(}) {
					if ($rawlines[$ln - 1] !~ m{\\n"}) {
						$ok = 1;
					}
					last;
				}
			}
			if ($ok == 0) {
1169
				WARN("printk() should include KERN_ facility level\n" . $herecurr);
1170
			}
1171
1172
		}

1173
1174
# function brace can't be on same line, except for #defines of do while,
# or if closed on same line
1175
		if (($line=~/$Type\s*[A-Za-z\d_]+\(.*\).*\s{/) and
1176
		    !($line=~/\#define.*do\s{/) and !($line=~/}/)) {
1177
			ERROR("open brace '{' following function declarations go on the next line\n" . $herecurr);
1178
		}
1179

1180
1181
1182
1183
1184
1185
# open braces for enum, union and struct go on the same line.
		if ($line =~ /^.\s*{/ &&
		    $prevline =~ /^.\s*(?:typedef\s+)?(enum|union|struct)(?:\s+$Ident)?\s*$/) {
			ERROR("open brace '{' following $1 go on the same line\n" . $hereprev);
		}

1186
# check for spaces between functions and their parentheses.
1187
		while ($line =~ /($Ident)\s+\(/g) {
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
			my $name = $1;
			my $ctx = substr($line, 0, $-[1]);

			# Ignore those directives where spaces _are_ permitted.
			if ($name =~ /^(?:if|for|while|switch|return|volatile|__volatile__|__attribute__|format|__extension__|Copyright|case)$/) {

			# cpp #define statements have non-optional spaces, ie
			# if there is a space between the name and the open
			# parenthesis it is simply not a parameter group.
			} elsif ($ctx =~ /^.\#\s*define\s*$/) {

			# If this whole things ends with a type its most
			# likely a typedef for a function.
			} elsif ("$ctx$name" =~ /$Type$/) {

			} else {
1204
1205
				WARN("no space between function name and open parenthesis '('\n" . $herecurr);
			}
1206
		}
1207
# Check operator spacing.
1208
		if (!($line=~/\#\s*include/)) {
1209
1210
1211
1212
			my $ops = qr{
				<<=|>>=|<=|>=|==|!=|
				\+=|-=|\*=|\/=|%=|\^=|\|=|&=|
				=>|->|<<|>>|<|>|=|!|~|
1213
				&&|\|\||,|\^|\+\+|--|&|\||\+|-|\*|\/|%
1214
1215
			}x;
			my @elements = split(/($ops|;)/, $opline);
1216
			my $off = 0;
1217
1218
1219

			my $blank = copy_spacing($opline);

1220
			for (my $n = 0; $n < $#elements; $n += 2) {
1221
1222
1223
1224
1225
1226
1227
1228
1229
				$off += length($elements[$n]);

				my $a = '';
				$a = 'V' if ($elements[$n] ne '');
				$a = 'W' if ($elements[$n] =~ /\s$/);
				$a = 'B' if ($elements[$n] =~ /(\[|\()$/);
				$a = 'O' if ($elements[$n] eq '');
				$a = 'E' if ($elements[$n] eq '' && $n == 0);

1230
				my $op = $elements[$n + 1];
1231
1232

				my $c = '';
1233
				if (defined $elements[$n + 2]) {
1234
1235
1236
1237
					$c = 'V' if ($elements[$n + 2] ne '');
					$c = 'W' if ($elements[$n + 2] =~ /^\s/);
					$c = 'B' if ($elements[$n + 2] =~ /^(\)|\]|;)/);
					$c = 'O' if ($elements[$n + 2] eq '');
1238
					$c = 'E' if ($elements[$n + 2] =~ /\s*\\$/);