start to enforce plain ascii or valid utf8

No device node or symlink can have other characters as plain
readable ascii or valid utf8. The /dev/disk/by-label/* symlinks
can no longer contain weird stuff read from untrusted sources.

Signed-off-by: Kay Sievers <kay.sievers@suse.de>

test/udev-test.pl

@@ -1324,7 +1324,25 @@ EOF
 		devpath		=> "/block/sda/sda1",
 		exp_name	=> "sane",
 		rules		=> <<EOF
-BUS=="scsi", KERNEL=="sda1", PROGRAM=="/bin/echo -e name; (/sbin/badprogram)", RESULT="name_ _/sbin/badprogram_", NAME="sane"
+BUS=="scsi", KERNEL=="sda1", PROGRAM=="/bin/echo -e name; (/sbin/badprogram)", RESULT=="name_ _/sbin/badprogram_", NAME="sane"
+EOF
+	},
+	{
+		desc		=> "untrusted string sanitize (don't replace utf8)",
+		subsys		=> "block",
+		devpath		=> "/block/sda/sda1",
+		exp_name	=> "uber",
+		rules		=> <<EOF
+BUS=="scsi", KERNEL=="sda1", PROGRAM=="/bin/echo -e \\xc3\\xbcber" RESULT=="\xc3\xbcber", NAME="uber"
+EOF
+	},
+	{
+		desc		=> "untrusted string sanitize (replace invalid utf8)",
+		subsys		=> "block",
+		devpath		=> "/block/sda/sda1",
+		exp_name	=> "replaced",
+		rules		=> <<EOF
+BUS=="scsi", KERNEL=="sda1", PROGRAM=="/bin/echo -e \\xef\\xe8garbage", RESULT=="[?][?]garbage", NAME="replaced"
 EOF
 	},
 	{

udev_rules.c

@@ -362,6 +362,7 @@ static void apply_format(struct udevice *udev, char *string, size_t maxsize,
 	char *head, *tail, *pos, *cpos, *attr, *rest;
 	int len;
 	int i;
+	int count;
 	unsigned int next_free_number;
 	struct sysfs_class_device *class_dev_parent;
 	enum subst_type {
@@ -544,7 +545,9 @@ found:
 			i = strlen(temp2);
 			while (i > 0 && isspace(temp2[i-1]))
 				temp2[--i] = '\0';
-			replace_untrusted_chars(temp2);
+			count = replace_untrusted_chars(temp2);
+			if (count)
+				info("%i untrusted character(s) replaced" , count);
 			strlcat(string, temp2, maxsize);
 			dbg("substitute sysfs value '%s'", temp2);
 			break;
@@ -812,9 +815,13 @@ try_parent:
 			if (rule->program.operation != KEY_OP_NOMATCH)
 				goto exit;
 		} else {
+			int count;
+
 			dbg("PROGRAM matches");
 			remove_trailing_char(result, '\n');
-			replace_untrusted_chars(result);
+			count = replace_untrusted_chars(result);
+			if (count)
+				info("%i untrusted character(s) replaced" , count);
 			dbg("result is '%s'", result);
 			strlcpy(udev->program_result, result, sizeof(udev->program_result));
 			dbg("PROGRAM returned successful");
@@ -960,6 +967,7 @@ int udev_rules_get_name(struct udev_rules *rules, struct udevice *udev, struct s
 			if (!udev->symlink_final && rule->symlink.operation != KEY_OP_UNSET) {
 				char temp[PATH_SIZE];
 				char *pos, *next;
+				int count;
 
 				if (rule->symlink.operation == KEY_OP_ASSIGN_FINAL)
 					udev->symlink_final = 1;
@@ -969,7 +977,10 @@ int udev_rules_get_name(struct udev_rules *rules, struct udevice *udev, struct s
 				}
 				strlcpy(temp, key_val(rule, &rule->symlink), sizeof(temp));
 				apply_format(udev, temp, sizeof(temp), class_dev, sysfs_device);
-				dbg("rule applied, added symlink '%s'", temp);
+				count = replace_untrusted_chars(temp);
+				if (count)
+					info("%i untrusted character(s) replaced" , count);
+				dbg("rule applied, added symlink(s) '%s'", temp);
 
 				/* add multiple symlinks separated by spaces */
 				pos = temp;
@@ -993,9 +1004,13 @@ int udev_rules_get_name(struct udev_rules *rules, struct udevice *udev, struct s
 
 			/* set name, later rules with name set will be ignored */
 			if (rule->name.operation != KEY_OP_UNSET) {
+				int count;
 				name_set = 1;
 				strlcpy(udev->name, key_val(rule, &rule->name), sizeof(udev->name));
 				apply_format(udev, udev->name, sizeof(udev->name), class_dev, sysfs_device);
+				count = replace_untrusted_chars(udev->name);
+				if (count)
+					info("%i untrusted character(s) replaced", count);
 
 				info("rule applied, '%s' becomes '%s'", udev->kernel_name, udev->name);
 				if (udev->type != DEV_NET)

udev_utils.h

@@ -42,9 +42,9 @@ extern int add_matching_files(struct list_head *name_list, const char *dirname,
 /* udev_utils_string.c */
 extern int strcmp_pattern(const char *p, const char *s);
 extern int string_is_true(const char *str);
-extern void replace_untrusted_chars(char *string);
 extern void remove_trailing_char(char *path, char c);
 extern int utf8_encoded_valid_unichar(const char *str);
+extern int replace_untrusted_chars(char *str);
 
 /* udev_utils_file.c */
 extern int create_path(const char *path);

udev_utils_string.c

@@ -232,12 +232,35 @@ int utf8_encoded_valid_unichar(const char *str)
 	return len;
 }
 
-void replace_untrusted_chars(char *string)
+/* replace everything but whitelisted plain ascii and valid utf8 */
+int replace_untrusted_chars(char *str)
 {
-	size_t len;
+	size_t i = 0;
+	int replaced = 0;
 
-	for (len = 0; string[len] != '\0'; len++) {
-		if (strchr(";,~\\()\'", string[len]))
-			string[len] = '_';
+	while (str[i] != '\0') {
+		int len;
+
+		/* valid printable ascii char */
+		if ((str[i] >= '0' && str[i] <= '9') ||
+		    (str[i] >= 'A' && str[i] <= 'Z') ||
+		    (str[i] >= 'a' && str[i] <= 'z') ||
+		    strchr(" #$%+-./:=?@_", str[i])) {
+			i++;
+			continue;
+		}
+		/* valid utf8 is accepted */
+		len = utf8_encoded_valid_unichar(&str[i]);
+		if (len > 1) {
+			i += len;
+			continue;
+		}
+
+		/* everything else is garbage */
+		str[i] = '_';
+		i++;
+		replaced++;
 	}
+
+	return replaced;
 }