Skip to contentSkip to conversation

checkpatch: update warnings

PATCH In review Tobin C. Harding 2 participants 7 messages Dec 7, 2017 – Dec 9, 2017 Checkpatch

lore
Share & tools
lore
#0 lore raw

cover checkpatch: update warnings

This set is a combination of previous submissions

[PATCH v2 0/3] checkpatch: warn for use of %px [PATCH] checkpatch: add check for tag Co-Developed-by

Although seemingly unrelated, they both patch checkpatch.pl and they were both contributed to by Joe and myself. They are also tied together by (previously incorrect) usage of the new Co-Developed-by tag.

Retains Joe's Ack'ed by tag on the %px patch from v2.

Retains Greg's Reviewed-by tag for the Co-Developed-by patch. The code is totally different but the sentiment is the same, in this case, I believe this is what the review was implying.

The Joe Perches patch is from LKML with minor adjustments after testing. Needs review please.

As discussed, if this is all ok it will go through Tobin's tree.

thanks,
Tobin.

Joe Perches (1):
checkpatch: add check for tag Co-Developed-by

Tobin C. Harding (3):
checkpatch: add sub routine get_stat_real()
checkpatch: add sub routine get_stat_here()
checkpatch: warn for use of %px

 scripts/checkpatch.pl | 172 +++++++++++++++++++++++++++-----------------------
 1 file changed, 93 insertions(+), 79 deletions(-)
Signature · 1 line
2.7.4
#1 lore raw

PATCH 1/4 checkpatch: add sub routine get_stat_real()

checkpatch currently contains duplicate code. We can define a sub routine and call that instead. This reduces code duplication and line count.

Add subroutine get_stat_real()

---
 scripts/checkpatch.pl | 31 +++++++++++++++----------------
 1 file changed, 15 insertions(+), 16 deletions(-)
scripts/checkpatch.pl +15 -16
index 040aa79e1d9d..5a63aa36300a 100755
@@ -1612,6 +1612,17 @@ sub raw_line {
 	return $line;
 }
 
+sub get_stat_real {
+	my ($linenr, $lc) = @_;
+
+	my $stat_real = raw_line($linenr, 0);
+	for (my $count = $linenr + 1; $count <= $lc; $count++) {
+		$stat_real = $stat_real . "\n" . raw_line($count, 0);
+	}
+
+	return $stat_real;
+}
+
 sub cat_vet {
 	my ($vet) = @_;
 	my ($res, $coded);
@@ -5759,10 +5770,7 @@ sub process {
 				}
 			}
 			if ($bad_extension ne "") {
-				my $stat_real = raw_line($linenr, 0);
-				for (my $count = $linenr + 1; $count <= $lc; $count++) {
-					$stat_real = $stat_real . "\n" . raw_line($count, 0);
-				}
+				my $stat_real = get_stat_real($linenr, $lc);
 				WARN("VSPRINTF_POINTER_EXTENSION",
 				     "Invalid vsprintf pointer extension '$bad_extension'\n" . "$here\n$stat_real\n");
 			}
@@ -5877,10 +5885,7 @@ sub process {
 		     $stat !~ /(?:$Compare)\s*\bsscanf\s*$balanced_parens/)) {
 			my $lc = $stat =~ tr@\n@@;
 			$lc = $lc + $linenr;
-			my $stat_real = raw_line($linenr, 0);
-		        for (my $count = $linenr + 1; $count <= $lc; $count++) {
-				$stat_real = $stat_real . "\n" . raw_line($count, 0);
-			}
+			my $stat_real = get_stat_real($linenr, $lc);
 			WARN("NAKED_SSCANF",
 			     "unchecked sscanf return value\n" . "$here\n$stat_real\n");
 		}
@@ -5891,10 +5896,7 @@ sub process {
 		    $line =~ /\bsscanf\b/) {
 			my $lc = $stat =~ tr@\n@@;
 			$lc = $lc + $linenr;
-			my $stat_real = raw_line($linenr, 0);
-		        for (my $count = $linenr + 1; $count <= $lc; $count++) {
-				$stat_real = $stat_real . "\n" . raw_line($count, 0);
-			}
+			my $stat_real = get_stat_real($linenr, $lc);
 			if ($stat_real =~ /\bsscanf\b\s*\(\s*$FuncArg\s*,\s*("[^"]+")/) {
 				my $format = $6;
 				my $count = $format =~ tr@%@%@;
@@ -6289,10 +6291,7 @@ sub process {
 
 				my $lc = $stat =~ tr@\n@@;
 				$lc = $lc + $linenr;
-				my $stat_real = raw_line($linenr, 0);
-				for (my $count = $linenr + 1; $count <= $lc; $count++) {
-					$stat_real = $stat_real . "\n" . raw_line($count, 0);
-				}
+				my $stat_real = get_stat_real($linenr, $lc);
 
 				my $skip_args = "";
 				if ($arg_pos > 1) {
Signature · 1 line
2.7.4
#2 lore raw

PATCH 2/4 checkpatch: add sub routine get_stat_here()

In reply to Tobin C. Harding

checkpatch currently contains duplicate code. We can define a sub routine and call that instead. This reduces code duplication and line count.

Add subroutine get_stat_here()

---
 scripts/checkpatch.pl | 54 +++++++++++++++++++--------------------------------
 1 file changed, 20 insertions(+), 34 deletions(-)
scripts/checkpatch.pl +20 -34
index 5a63aa36300a..4534c9a9aefa 100755
@@ -1623,6 +1623,17 @@ sub get_stat_real {
 	return $stat_real;
 }
 
+sub get_stat_here {
+	my ($linenr, $cnt, $here) = @_;
+
+	my $herectx = $here . "\n";
+	for (my $n = 0; $n < $cnt; $n++) {
+		$herectx .= raw_line($linenr, $n) . "\n";
+	}
+
+	return $herectx;
+}
+
 sub cat_vet {
 	my ($vet) = @_;
 	my ($res, $coded);
@@ -4911,12 +4922,8 @@ sub process {
 			#print "REST<$rest> dstat<$dstat> ctx<$ctx>\n";
 
 			$ctx =~ s/\n*$//;
-			my $herectx = $here . "\n";
 			my $stmt_cnt = statement_rawlines($ctx);
-
-			for (my $n = 0; $n < $stmt_cnt; $n++) {
-				$herectx .= raw_line($linenr, $n) . "\n";
-			}
+			my $herectx = get_stat_here($linenr, $stmt_cnt, $here);
 
 			if ($dstat ne '' &&
 			    $dstat !~ /^(?:$Ident|-?$Constant),$/ &&			# 10, // foo(),
@@ -4988,12 +4995,9 @@ sub process {
 # check for macros with flow control, but without ## concatenation
 # ## concatenation is commonly a macro that defines a function so ignore those
 			if ($has_flow_statement && !$has_arg_concat) {
-				my $herectx = $here . "\n";
 				my $cnt = statement_rawlines($ctx);
+				my $herectx = get_stat_here($linenr, $cnt, $here);
 
-				for (my $n = 0; $n < $cnt; $n++) {
-					$herectx .= raw_line($linenr, $n) . "\n";
-				}
 				WARN("MACRO_WITH_FLOW_CONTROL",
 				     "Macros with flow control statements should be avoided\n" . "$herectx");
 			}
@@ -5033,11 +5037,7 @@ sub process {
 
 				$ctx =~ s/\n*$//;
 				my $cnt = statement_rawlines($ctx);
-				my $herectx = $here . "\n";
-
-				for (my $n = 0; $n < $cnt; $n++) {
-					$herectx .= raw_line($linenr, $n) . "\n";
-				}
+				my $herectx = get_stat_here($linenr, $cnt, $here);
 
 				if (($stmts =~ tr/;/;/) == 1 &&
 				    $stmts !~ /^\s*(if|while|for|switch)\b/) {
@@ -5051,11 +5051,7 @@ sub process {
 			} elsif ($dstat =~ /^\+\s*#\s*define\s+$Ident.*;\s*$/) {
 				$ctx =~ s/\n*$//;
 				my $cnt = statement_rawlines($ctx);
-				my $herectx = $here . "\n";
-
-				for (my $n = 0; $n < $cnt; $n++) {
-					$herectx .= raw_line($linenr, $n) . "\n";
-				}
+				my $herectx = get_stat_here($linenr, $cnt, $here);
 
 				WARN("TRAILING_SEMICOLON",
 				     "macros should not use a trailing semicolon\n" . "$herectx");
@@ -5178,12 +5174,8 @@ sub process {
 				}
 			}
 			if ($level == 0 && $block =~ /^\s*\{/ && !$allowed) {
-				my $herectx = $here . "\n";
 				my $cnt = statement_rawlines($block);
-
-				for (my $n = 0; $n < $cnt; $n++) {
-					$herectx .= raw_line($linenr, $n) . "\n";
-				}
+				my $herectx = get_stat_here($linenr, $cnt, $here);
 
 				WARN("BRACES",
 				     "braces {} are not necessary for single statement blocks\n" . $herectx);
@@ -6026,12 +6018,9 @@ sub process {
 			}
 			if ($r1 !~ /^sizeof\b/ && $r2 =~ /^sizeof\s*\S/ &&
 			    !($r1 =~ /^$Constant$/ || $r1 =~ /^[A-Z_][A-Z0-9_]*$/)) {
-				my $ctx = '';
-				my $herectx = $here . "\n";
 				my $cnt = statement_rawlines($stat);
-				for (my $n = 0; $n < $cnt; $n++) {
-					$herectx .= raw_line($linenr, $n) . "\n";
-				}
+				my $herectx = get_stat_here($linenr, $cnt, $here);
+
 				if (WARN("ALLOC_WITH_MULTIPLY",
 					 "Prefer $newfunc over $oldfunc with multiply\n" . $herectx) &&
 				    $cnt == 1 &&
@@ -6114,12 +6103,9 @@ sub process {
 		if ($^V && $^V ge 5.10.0 &&
 		    defined $stat &&
 		    $stat =~ /^\+[$;\s]*(?:case[$;\s]+\w+[$;\s]*:[$;\s]*|)*[$;\s]*\bdefault[$;\s]*:[$;\s]*;/g) {
-			my $ctx = '';
-			my $herectx = $here . "\n";
 			my $cnt = statement_rawlines($stat);
-			for (my $n = 0; $n < $cnt; $n++) {
-				$herectx .= raw_line($linenr, $n) . "\n";
-			}
+			my $herectx = get_stat_here($linenr, $cnt, $here);
+
 			WARN("DEFAULT_NO_BREAK",
 			     "switch default: should use break\n" . $herectx);
 		}
Signature · 1 line
2.7.4
#3 lore raw

PATCH 3/4 checkpatch: warn for use of %px

In reply to Tobin C. Harding

Usage of the new %px specifier potentially leaks sensitive inforamtion. Printing kernel addresses exposes the kernel layout in memory, this is potentially exploitable. We have tools in the kernel to help us do the right thing. We can have checkpatch warn developers of potential dangers of using %px.

Have checkpatch emit a warning for usage of specifier %px.

---
 scripts/checkpatch.pl | 31 ++++++++++++++++++++++++-------
 1 file changed, 24 insertions(+), 7 deletions(-)
scripts/checkpatch.pl +24 -7
index 4534c9a9aefa..6df99d1ebca5 100755
@@ -5750,21 +5750,38 @@ sub process {
 		    defined $stat &&
 		    $stat =~ /^\+(?![^\{]*\{\s*).*\b(\w+)\s*\(.*$String\s*,/s &&
 		    $1 !~ /^_*volatile_*$/) {
-			my $bad_extension = "";
+			my $specifier;
+			my $extension;
+			my $bad_specifier = "";
+			my $stat_real;
+
 			my $lc = $stat =~ tr@\n@@;
 			$lc = $lc + $linenr;
 		        for (my $count = $linenr; $count <= $lc; $count++) {
 				my $fmt = get_quoted_string($lines[$count - 1], raw_line($count, 0));
 				$fmt =~ s/%%//g;
-				if ($fmt =~ /(\%[\*\d\.]*p(?![\WFfSsBKRraEhMmIiUDdgVCbGNOx]).)/) {
-					$bad_extension = $1;
-					last;
+
+				while ($fmt =~ /(\%[\*\d\.]*p(\w))/g) {
+					$specifier = $1;
+					$extension = $2;
+					if ($extension !~ /[FfSsBKRraEhMmIiUDdgVCbGNOx]/) {
+						$bad_specifier = $specifier;
+						last;
+					}
+					if ($extension eq "x" && !defined($stat_real)) {
+						if (!defined($stat_real)) {
+							$stat_real = get_stat_real($linenr, $lc);
+						}
+						WARN("VSPRINTF_SPECIFIER_PX",
+						     "Using vsprintf specifier '\%px' potentially exposes the kernel memory layout, if you don't really need the address please consider using '\%p'.\n" . "$here\n$stat_real\n");
+					}
 				}
+
 			}
-			if ($bad_extension ne "") {
-				my $stat_real = get_stat_real($linenr, $lc);
+			if ($bad_specifier ne "") {
+				$stat_real = get_stat_real($linenr, $lc);
 				WARN("VSPRINTF_POINTER_EXTENSION",
-				     "Invalid vsprintf pointer extension '$bad_extension'\n" . "$here\n$stat_real\n");
+				     "Invalid vsprintf pointer extension '$bad_specifier'\n" . "$here\n$stat_real\n");
 			}
 		}
 
Signature · 1 line
2.7.4
#4 lore raw

PATCH 4/4 checkpatch: add check for tag Co-Developed-by

In reply to Tobin C. Harding

From: Joe Perches <[email protected]>

Recently signature tag Co-Developed-by was added to the
kernel (Documentation/process/5.Posting.rst). checkpatch.pl doesn't know about it yet. All prior tags used all lowercase characters except for first character. Checks for this format had to be re-worked to allow for the new tag.

Cc Greg Kroah-Hartman <[email protected]>

---

Greg,

This patch has the same brief commit and achieves the same aim as the code you reviewed. I'm adding your tag here even though the code is totally different. I believe you were reviewing the behaviour of the code not the actual logic.

Yell if I'm wrong.

thanks,
Tobin.

 scripts/checkpatch.pl | 58 +++++++++++++++++++++++++++++++--------------------
 1 file changed, 35 insertions(+), 23 deletions(-)
scripts/checkpatch.pl +35 -23
index 6df99d1ebca5..36f3b00783a9 100755
@@ -461,16 +461,18 @@ our $logFunctions = qr{(?x:
 	seq_vprintf|seq_printf|seq_puts
 )};
 
-our $signature_tags = qr{(?xi:
-	Signed-off-by:|
-	Acked-by:|
-	Tested-by:|
-	Reviewed-by:|
-	Reported-by:|
-	Suggested-by:|
-	To:|
-	Cc:
-)};
+our @valid_signatures = (
+	"Signed-off-by:",
+	"Acked-by:",
+	"Tested-by:",
+	"Reviewed-by:",
+	"Reported-by:",
+	"Suggested-by:",
+	"Co-Developed-by:",
+	"To:",
+	"Cc:"
+);
+my $signature_tags = "(?x:" . join('|', @valid_signatures) . ")";
 
 our @typeListMisordered = (
 	qr{char\s+(?:un)?signed},
@@ -2183,6 +2185,17 @@ sub pos_last_openparen {
 	return length(expand_tabs(substr($line, 0, $last_openparen))) + 1;
 }
 
+sub get_preferred_sign_off {
+	my ($sign_off) = @_;
+
+	foreach my $sig (@valid_signatures) {
+		if (lc($sign_off) eq lc($sig)) {
+			return $sig;
+		}
+	}
+	return "";
+}
+
 sub process {
 	my $filename = shift;
 
@@ -2489,35 +2502,34 @@ sub process {
 			my $sign_off = $2;
 			my $space_after = $3;
 			my $email = $4;
-			my $ucfirst_sign_off = ucfirst(lc($sign_off));
+			my $preferred_sign_off = ucfirst(lc($sign_off));
 
-			if ($sign_off !~ /$signature_tags/) {
+			if ($sign_off !~ /$signature_tags/i) {
 				WARN("BAD_SIGN_OFF",
 				     "Non-standard signature: $sign_off\n" . $herecurr);
-			}
-			if (defined $space_before && $space_before ne "") {
+			} elsif ($sign_off !~ /$signature_tags/) {
+				$preferred_sign_off = get_preferred_sign_off($sign_off);
 				if (WARN("BAD_SIGN_OFF",
-					 "Do not use whitespace before $ucfirst_sign_off\n" . $herecurr) &&
+					 "'$preferred_sign_off' is the preferred signature form\n" . $herecurr) &&
 				    $fix) {
-					$fixed[$fixlinenr] =
-					    "$ucfirst_sign_off $email";
+					$fixed[$fixlinenr] = "$preferred_sign_off $email";
 				}
 			}
-			if ($sign_off =~ /-by:$/i && $sign_off ne $ucfirst_sign_off) {
+			if (defined $space_before && $space_before ne "") {
 				if (WARN("BAD_SIGN_OFF",
-					 "'$ucfirst_sign_off' is the preferred signature form\n" . $herecurr) &&
+					 "Do not use whitespace before $preferred_sign_off\n" . $herecurr) &&
 				    $fix) {
 					$fixed[$fixlinenr] =
-					    "$ucfirst_sign_off $email";
+					    "$preferred_sign_off $email";
 				}
-
 			}
+
 			if (!defined $space_after || $space_after ne " ") {
 				if (WARN("BAD_SIGN_OFF",
-					 "Use a single space after $ucfirst_sign_off\n" . $herecurr) &&
+					 "Use a single space after $preferred_sign_off\n" . $herecurr) &&
 				    $fix) {
 					$fixed[$fixlinenr] =
-					    "$ucfirst_sign_off $email";
+					    "$preferred_sign_off $email";
 				}
 			}
 
Signature · 1 line
2.7.4
#5 lore raw

Re: [PATCH 4/4] checkpatch: add check for tag Co-Developed-by

Quoted 12 lines · Tobin C. Harding

From: Joe Perches <[email protected]>

Recently signature tag Co-Developed-by was added to the
kernel (Documentation/process/5.Posting.rst). checkpatch.pl doesn't know about it yet. All prior tags used all lowercase characters except for first character. Checks for this format had to be re-worked to allow for the new tag.

Cc Greg Kroah-Hartman <[email protected]>
Quoted 8 lines

---

Greg,

This patch has the same brief commit and achieves the same aim as the code you reviewed. I'm adding your tag here even though the code is totally different. I believe you were reviewing the behaviour of the code not the actual logic.

Well, I did read the actual logic, I do know perl :)

Anyway, my reviewed-by still stands here, thanks.

greg k-h

#6 lore raw

Re: [PATCH 4/4] checkpatch: add check for tag Co-Developed-by

Quoted 15 lines · Greg Kroah-Hartman
Quoted 14 lines · Tobin C. Harding

From: Joe Perches <[email protected]>

Recently signature tag Co-Developed-by was added to the
kernel (Documentation/process/5.Posting.rst). checkpatch.pl doesn't know about it yet. All prior tags used all lowercase characters except for first character. Checks for this format had to be re-worked to allow for the new tag.

Cc Greg Kroah-Hartman <[email protected]>

---

Quoted 8 lines
Quoted 6 lines

Greg,

This patch has the same brief commit and achieves the same aim as the code you reviewed. I'm adding your tag here even though the code is totally different. I believe you were reviewing the behaviour of the code not the actual logic.

Well, I did read the actual logic, I do know perl :)

No offense meant, just trying to make your day easier.

Quoted 1 line

Anyway, my reviewed-by still stands here, thanks.

thanks,
Tobin.