checkpatch.pl

#!/usr/bin/perl -w
# (c) 2001, Dave Jones. <davej@codemonkey.org.uk> (the file handling bit)
# (c) 2005, Joel Schopp <jschopp@austin.ibm.com> (the ugly bit)
# (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;
$P =~ s@.*/@@g;

my $V = '0.21';

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

my $quiet = 0;
my $tree = 1;
my $chk_signoff = 1;
my $chk_patch = 1;
my $tst_only;
my $emacs = 0;
my $terse = 0;
my $file = 0;
my $check = 0;
my $summary = 1;
my $mailback = 0;
my $summary_file = 0;
my $root;
my %debug;
GetOptions(
	'q|quiet+'	=> \$quiet,
	'tree!'		=> \$tree,
	'signoff!'	=> \$chk_signoff,
	'patch!'	=> \$chk_patch,
	'emacs!'	=> \$emacs,
	'terse!'	=> \$terse,
	'file!'		=> \$file,
	'subjective!'	=> \$check,
	'strict!'	=> \$check,
	'root=s'	=> \$root,
	'summary!'	=> \$summary,
	'mailback!'	=> \$mailback,
	'summary-file!'	=> \$summary_file,

	'debug=s'	=> \%debug,
	'test-only=s'	=> \$tst_only,
) or exit;

my $exit = 0;

if ($#ARGV < 0) {
	print "usage: $P [options] patchfile\n";
	print "version: $V\n";
	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";
	exit(1);
}

my $dbg_values = 0;
my $dbg_possible = 0;
my $dbg_type = 0;
my $dbg_attr = 0;
for my $key (keys %debug) {
	eval "\${dbg_$key} = '$debug{$key}';"
}

if ($terse) {
	$emacs = 1;
	$quiet++;
}

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);
	}
}

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
		}x;
our $Attribute	= qr{
			const|
			__read_mostly|
			__kprobes|
			__(?:mem|cpu|dev|)(?:initdata|init)|
			____cacheline_aligned|
			____cacheline_aligned_in_smp|
			____cacheline_internodealigned_in_smp
		  }x;
our $Modifier;
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{
			<=|>=|==|!=|
			=>|->|<<|>>|<|>|!|~|
			&&|\|\||,|\^|\+\+|--|&|\||\+|-|\*|\/|%
		  }x;

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

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;

our @typeList = (
	qr{void},
	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},
	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},
);
our @modifierList = (
	qr{fastcall},
);

sub build_types {
	my $mods = "(?x:  \n" . join("|\n  ", @modifierList) . "\n)";
	my $all = "(?x:  \n" . join("|\n  ", @typeList) . "\n)";
	$Modifier	= qr{(?:$Attribute|$Sparse|$mods)};
	$NonptrType	= qr{
			(?:$Modifier\s+|const\s+)*
			(?:
				(?:typeof|__typeof__)\s*\(\s*\**\s*$Ident\s*\)|
				(?:${all}\b)
			)
			(?:\s+$Modifier|\s+const)*
		  }x;
	$Type	= qr{
			$NonptrType
			(?:\s*\*+\s*const|\s*\*+|(?:\s*\[\s*\])+)?
			(?:\s+$Inline|\s+$Modifier)*
		  }x;
	$Declare	= qr{(?:$Storage\s+)?$Type};
}
build_types();

$chk_signoff = 0 if ($file);

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

				} elsif ($entry !~ m@/@) {
					push(@dep_functions, $entry);
				}
			}
		}
	}
}

my @rawlines = ();
my @lines = ();
my $vname;
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 - $!\n";
	}
	if ($filename eq '-') {
		$vname = 'Your patch';
	} else {
		$vname = $filename;
	}
	while (<FILE>) {
		chomp;
		push(@rawlines, $_);
	}
	close(FILE);
	if (!process($filename)) {
		$exit = 1;
	}
	@rawlines = ();
	@lines = ();
}

exit($exit);

sub top_of_kernel_tree {
	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;
		}
	}
	return 1;
}

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;
}
sub copy_spacing {
	(my $res = shift) =~ tr/\t/ /c;
	return $res;
}

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));
}

my $sanitise_quote = '';

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

	if ($in_comment) {
		$sanitise_quote = '*/';
	} else {
		$sanitise_quote = '';
	}
}
sub sanitise_line {
	my ($line) = @_;

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

	my $qlen = 0;
	my $off = 0;
	my $c;

	# 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, $off, 2, "$;$;");
			$off++;
			next;
		}
		if (substr($line, $off, 2) eq '*/') {
			$sanitise_quote = '';
			substr($res, $off, 2, "$;$;");
			$off++;
			next;
		}

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

				substr($res, $off, 1, $c);
				next;
			} elsif ($sanitise_quote eq $c) {
				$sanitise_quote = '';
			}
		}

		#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);
		}
	}

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

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

	return $res;
}

sub ctx_statement_block {
	my ($linenr, $remain, $off) = @_;
	my $line = $linenr - 1;
	my $blk = '';
	my $soff = $off;
	my $coff = $off - 1;
	my $coff_set = 0;

	my $loff = 0;

	my $type = '';
	my $level = 0;
	my $p;
	my $c;
	my $len = 0;

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

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

		# An else is really a conditional as long as its not else if
		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";
		}

		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;
				$coff_set = 1;
				#warn "CSB: mark coff<$coff>\n";
			}
		}
		if (($type eq '' || $type eq '{') && $c eq '{') {
			$level++;
			$type = '{';
		}
		if ($type eq '{' && $c eq '}') {
			$level--;
			$type = ($level != 0)? '{' : '';

			if ($level == 0) {
				last;
			}
		}
		$off++;
	}
	# We are truly at the end, so shuffle to the next line.
	if ($off == $len) {
		$loff = $len + 1;
		$line++;
		$remain--;
	}

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

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

	#print "coff<$coff> soff<$off> loff<$loff>\n";

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

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;
	}
}

sub ctx_statement_full {
	my ($linenr, $remain, $off) = @_;
	my ($statement, $condition, $level);

	my (@chunks);

	# Grab the first conditional/block pair.
	($statement, $condition, $linenr, $remain, $off, $level) =
				ctx_statement_block($linenr, $remain, $off);
	#print "F: c<$condition> s<$statement> remain<$remain>\n";
	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.
	for (;;) {
		($statement, $condition, $linenr, $remain, $off, $level) =
				ctx_statement_block($linenr, $remain, $off);
		#print "C: c<$condition> s<$statement> remain<$remain>\n";
		last if (!($remain > 0 && $condition =~ /^(?:\s*\n[+-])*\s*(?:else|do)\b/s));
		#print "C: push\n";
		push(@chunks, [ $condition, $statement ]);
	}

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

sub ctx_block_get {
	my ($linenr, $remain, $outer, $open, $close, $off) = @_;
	my $line;
	my $start = $linenr - 1;
	my $blk = '';
	my @o;
	my @c;
	my @res = ();

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

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

			if ($c eq $close && $level > 0) {
				$level--;
				last if ($level == 0);
			} elsif ($c eq $open) {
				$level++;
			}
		}

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

		last if ($level == 0);
	}

	return ($level, @res);
}
sub ctx_block_outer {
	my ($linenr, $remain) = @_;

	my ($level, @r) = ctx_block_get($linenr, $remain, 1, '{', '}', 0);
	return @r;
}
sub ctx_block {
	my ($linenr, $remain) = @_;

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

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

	return ctx_block_get($linenr, $remain, 0, '{', '}', 0);
}
sub ctx_statement_level {
	my ($linenr, $remain, $off) = @_;

	return ctx_block_get($linenr, $remain, 0, '(', ')', $off);
}

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

	# Catch a comment on the end of the line itself.
	my ($current_comment) = ($rawlines[$end_line - 1] =~ m@.*(/\*.*\*/)\s*(?:\\\s*)?$@);
	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++) {
		my $line = $rawlines[$linenr - 1];
		#warn "           $line\n";
		if ($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);

	##print "LINE: $rawlines[$end_line - 1 ]\n";
	##print "CMMT: $cmt\n";

	return ($cmt ne '');
}

sub cat_vet {
	my ($vet) = @_;
	my ($res, $coded);

	$res = '';
	while ($vet =~ /([^[:cntrl:]]*)([[:cntrl:]]|$)/g) {
		$res .= $1;
		if ($2 ne '') {
			$coded = sprintf("^%c", unpack('C', $2) + 64);
			$res .= $coded;
		}
	}
	$res =~ s/$/\$/;

	return $res;
}

my $av_preprocessor = 0;
my $av_pending;
my @av_paren_type;
my $av_pend_colon;

sub annotate_reset {
	$av_preprocessor = 0;
	$av_pending = '_';
	@av_paren_type = ('E');
	$av_pend_colon = 'O';
}

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

	my $res;
	my $var = '_' x length($stream);
	my $cur = $stream;

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

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

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

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

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

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

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

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

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

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

			$type = 'E';

		} elsif ($cur =~ /^(\#\s*(?:endif))/o) {
			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);
			$type = 'E';

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

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

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

		} elsif ($cur =~ /^(if|while|for)\b/o) {
			print "COND($1)\n" if ($dbg_values > 1);
			$av_pending = 'E';
			$type = 'N';

		} elsif ($cur =~/^(case)/o) {
			print "CASE($1)\n" if ($dbg_values > 1);
			$av_pend_colon = 'C';
			$type = 'N';

		} elsif ($cur =~/^(return|else|goto|typeof|__typeof__)\b/o) {
			print "KEYWORD($1)\n" if ($dbg_values > 1);
			$type = 'N';

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

		} elsif ($cur =~ /^(\))/o) {
			my $new_type = pop(@av_paren_type);
			if ($new_type ne '_') {
				$type = $new_type;
				print "PAREN('$1') -> $type\n"
							if ($dbg_values > 1);
			} else {
				print "PAREN('$1')\n" if ($dbg_values > 1);
			}

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

		} elsif ($cur =~ /^($Ident\s*):/) {
			if ($type eq 'E') {
				$av_pend_colon = 'L';
			} elsif ($type eq 'T') {
				$av_pend_colon = 'B';
			}
			print "IDENT_COLON($1,$type>$av_pend_colon)\n" if ($dbg_values > 1);
			$type = 'V';

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

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

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

		} elsif ($cur =~ /^(\?)/o) {
			print "QUESTION($1)\n" if ($dbg_values > 1);
			$type = 'N';

		} elsif ($cur =~ /^(:)/o) {
			print "COLON($1,$av_pend_colon)\n" if ($dbg_values > 1);

			substr($var, length($res), 1, $av_pend_colon);
			if ($av_pend_colon eq 'C' || $av_pend_colon eq 'L') {
				$type = 'E';
			} else {
				$type = 'N';
			}
			$av_pend_colon = 'O';

		} elsif ($cur =~ /^(;|\[)/o) {
			print "CLOSE($1)\n" if ($dbg_values > 1);
			$type = 'N';

		} elsif ($cur =~ /^(-(?![->])|\+(?!\+)|\*|\&\&|\&)/o) {
			my $variant;

			print "OPV($1)\n" if ($dbg_values > 1);
			if ($type eq 'V') {
				$variant = 'B';
			} else {
				$variant = 'U';
			}

			substr($var, length($res), 1, $variant);
			$type = 'N';

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

		} elsif ($cur =~ /(^.)/o) {
			print "C($1)\n" if ($dbg_values > 1);
		}
		if (defined $1) {
			$cur = substr($cur, length($1));
			$res .= $type x length($1);
		}
	}

	return ($res, $var);
}

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

	print "CHECK<$possible> ($line)\n" if ($dbg_possible > 1);
	if ($possible !~ /^(?:$Modifier|$Storage|$Type|DEFINE_\S+)$/ &&
	    $possible ne 'goto' && $possible ne 'return' &&
	    $possible ne 'case' && $possible ne 'else' &&
	    $possible ne 'asm' && $possible ne '__asm__' &&
	    $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;
			for my $modifier (split(' ', $possible)) {
				warn "MODIFIER: $modifier ($possible) ($line)\n" if ($dbg_possible);
				push(@modifierList, $modifier);
			}

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

my $prefix = '';

sub report {
	if (defined $tst_only && $_[0] !~ /\Q$tst_only\E/) {
		return 0;
	}
	my $line = $prefix . $_[0];

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

	push(our @report, $line);

	return 1;
}
sub report_dump {
	our @report;
}
sub ERROR {
	if (report("ERROR: $_[0]\n")) {
		our $clean = 0;
		our $cnt_error++;
	}
}
sub WARN {
	if (report("WARNING: $_[0]\n")) {
		our $clean = 0;
		our $cnt_warn++;
	}
}
sub CHK {
	if ($check && report("CHECK: $_[0]\n")) {
		our $clean = 0;
		our $cnt_chk++;
	}
}

sub check_absolute_file {
	my ($absolute, $herecurr) = @_;
	my $file = $absolute;

	##print "absolute<$absolute>\n";

	# See if any suffix of this path is a path within the tree.
	while ($file =~ s@^[^/]*/@@) {
		if (-f "$root/$file") {
			##print "file<$file>\n";
			last;
		}
	}
	if (! -f _)  {
		return 0;
	}

	# It is, so see if the prefix is acceptable.
	my $prefix = $absolute;
	substr($prefix, -length($file)) = '';

	##print "prefix<$prefix>\n";
	if ($prefix ne ".../") {
		WARN("use relative pathname instead of absolute in changelog text\n" . $herecurr);
	}
}

sub process {
	my $filename = shift;

	my $linenr=0;
	my $prevline="";
	my $prevrawline="";
	my $stashline="";
	my $stashrawline="";

	my $length;
	my $indent;
	my $previndent=0;
	my $stashindent=0;

	our $clean = 1;
	my $signoff = 0;
	my $is_patch = 0;

	our @report = ();
	our $cnt_lines = 0;
	our $cnt_error = 0;
	our $cnt_warn = 0;
	our $cnt_chk = 0;

	# Trace the real file/line as we go.
	my $realfile = '';
	my $realline = 0;
	my $realcnt = 0;
	my $here = '';
	my $in_comment = 0;
	my $comment_edge = 0;
	my $first_line = 0;

	my $prev_values = 'E';

	# suppression flags
	my %suppress_ifbraces;

	# Pre-scan the patch sanitizing the lines.
	# Pre-scan the patch looking for any __setup documentation.
	#
	my @setup_docs = ();
	my $setup_docs = 0;

	sanitise_line_reset();
	my $line;
	foreach my $rawline (@rawlines) {
		$linenr++;
		$line = $rawline;

		if ($rawline=~/^\+\+\+\s+(\S+)/) {
			$setup_docs = 0;
			if ($1 =~ m@Documentation/kernel-parameters.txt$@) {
				$setup_docs = 1;
			}
			#next;
		}
		if ($rawline=~/^\@\@ -\d+(?:,\d+)? \+(\d+)(,(\d+))? \@\@/) {
			$realline=$1-1;
			if (defined $2) {
				$realcnt=$3+1;
			} else {
				$realcnt=1+1;
			}
			$in_comment = 0;

			# 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;
			my $cnt = $realcnt;
			for (my $ln = $linenr + 1; $cnt > 0; $ln++) {
				next if (defined $rawlines[$ln - 1] &&
					 $rawlines[$ln - 1] =~ /^-/);
				$cnt--;
				#print "RAW<$rawlines[$ln - 1]>\n";
				($edge) = (defined $rawlines[$ln - 1] &&
					$rawlines[$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);

		} elsif ($realcnt && $rawline =~ /^(?:\+| |$)/) {
			# Standardise the strings and chars within the input to
			# simplify matching -- only bother with positive lines.
			$line = sanitise_line($rawline);
		}
		push(@lines, $line);

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

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

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

	$prefix = '';

	$realcnt = 0;
	$linenr = 0;
	foreach my $line (@lines) {
		$linenr++;

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

#extract the line range in the file after the patch is applied
		if ($line=~/^\@\@ -\d+(?:,\d+)? \+(\d+)(,(\d+))? \@\@/) {
			$is_patch = 1;
			$first_line = $linenr + 1;
			$realline=$1-1;
			if (defined $2) {
				$realcnt=$3+1;
			} else {
				$realcnt=1+1;
			}
			annotate_reset();
			$prev_values = 'E';

			%suppress_ifbraces = ();
			next;

# 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.
		} elsif ($line =~ /^( |\+|$)/) {
			$realline++;
			$realcnt-- if ($realcnt != 0);

			# Measure the line length and indent.
			($length, $indent) = line_stats($rawline);

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

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

		} elsif ($realcnt == 1) {
			$realcnt--;
		}

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

		$here = "#$linenr: " if (!$file);
		$here = "#$realline: " if ($file);

		# 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;
		}

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

		my $hereline = "$here\n$rawline\n";
		my $herecurr = "$here\n$rawline\n";
		my $hereprev = "$here\n$prevrawline\n$rawline\n";

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

#check the patch for a signoff:
		if ($line =~ /^\s*signed-off-by:/i) {