Merge pull request #17399 from afq984/udev-escaped-string

Allow escaped string in udev rules

man/udev.xml

@@ -23,7 +23,8 @@
     <refpurpose>Dynamic device management</refpurpose>
   </refnamediv>
 
-  <refsect1><title>Description</title>
+  <refsect1>
+    <title>Description</title>
     <para>udev supplies the system software with device events, manages permissions
     of device nodes and may create additional symlinks in the <filename>/dev/</filename>
     directory, or renames network interfaces. The kernel usually just assigns unpredictable
@@ -44,7 +45,8 @@
     sources is provided by the library libudev.</para>
   </refsect1>
 
-  <refsect1><title>Rules Files</title>
+  <refsect1>
+      <title>Rules Files</title>
       <para>The udev rules are read from the files located in the system rules directories
       <filename>/usr/lib/udev/rules.d</filename> and <filename>/usr/local/lib/udev/rules.d</filename>, the
       volatile runtime directory <filename>/run/udev/rules.d</filename> and the local administration
@@ -68,9 +70,11 @@
       pointing to the device node, or run a specified program as part of
       the event handling.</para>
 
-      <para>A rule consists of a comma-separated list of one or more key-value pairs.
-      Each key has a distinct operation, depending on the used operator. Valid
-      operators are:</para>
+      <para>A rule consists of a comma-separated list of one or more key-operator-value expressions.
+      Each expression has a distinct effect, depending on the key and operator used.</para>
+
+      <refsect2>
+        <title>Operators</title>
         <variablelist>
           <varlistentry>
             <term><literal>==</literal></term>
@@ -115,7 +119,26 @@
             </listitem>
           </varlistentry>
         </variablelist>
+      </refsect2>
 
+      <refsect2>
+        <title>Values</title>
+          <para>Values are written as double quoted strings, such as ("string").
+          To include a quotation mark (") in the value, precede it by a backslash (\").
+          Any other occurrences of a character followed by a backslash are not further unescaped.
+          That is, "\t\n" is treated as four characters:
+          backslash, lowercase t, backslash, lowercase n.</para>
+
+          <para>The string can be prefixed with a lowercase e (e"string\n") to mark the string as
+          <ulink url="https://en.wikipedia.org/wiki/Escape_sequences_in_C#Table_of_escape_sequences">C-style escaped</ulink>.
+          For example, e"string\n" is parsed as 7 characters: 6 lowercase letters and a newline.
+          This can be useful for writting special characters when a kernel driver requires them.</para>
+
+          <para>Please note that <constant>NUL</constant> is not allowed in either string variant.</para>
+      </refsect2>
+
+      <refsect2>
+        <title>Keys</title>
         <para>The following key names can be used to match against device properties.
         Some of the keys also match against properties of the parent devices in sysfs,
         not only the device that has generated the event. If multiple keys that match
@@ -789,6 +812,7 @@
             </listitem>
           </varlistentry>
         </variablelist>
+      </refsect2>
   </refsect1>
 
   <refsect1>

src/fuzz/fuzz-udev-rule-parse-value.c

@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: LGPL-2.1+ */
+
+#include <string.h>
+
+#include "alloc-util.h"
+#include "fuzz.h"
+#include "udev-util.h"
+
+int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {
+        _cleanup_free_ char *str = NULL;
+        int r;
+        char *value = UINT_TO_PTR(0x12345678U);
+        char *endpos = UINT_TO_PTR(0x87654321U);
+
+        assert_se(str = malloc(size + 1));
+        memcpy(str, data, size);
+        str[size] = '\0';
+
+        r = udev_rule_parse_value(str, &value, &endpos);
+
+        if (r < 0) {
+                /* not modified on failure */
+                assert_se(value == UINT_TO_PTR(0x12345678U));
+                assert_se(endpos == UINT_TO_PTR(0x87654321U));
+        } else {
+                assert_se(endpos <= str + size);
+                assert_se(endpos > str + 1);
+        }
+
+        return 0;
+}

src/fuzz/meson.build

@@ -152,4 +152,8 @@ fuzzers += [
           'src/xdg-autostart-generator/xdg-autostart-service.c'],
          [],
          []],
+
+        [['src/fuzz/fuzz-udev-rule-parse-value.c'],
+         [libshared],
+         []],
 ]

src/shared/udev-util.c

@@ -6,13 +6,16 @@
 #include "alloc-util.h"
 #include "device-util.h"
 #include "env-file.h"
+#include "escape.h"
 #include "log.h"
+#include "macro.h"
 #include "parse-util.h"
 #include "path-util.h"
 #include "signal-util.h"
 #include "string-table.h"
 #include "string-util.h"
 #include "udev-util.h"
+#include "utf8.h"
 
 static const char* const resolve_name_timing_table[_RESOLVE_NAME_TIMING_MAX] = {
         [RESOLVE_NAME_NEVER] = "never",
@@ -320,3 +323,49 @@ bool device_for_action(sd_device *dev, DeviceAction action) {
 
         return a == action;
 }
+
+int udev_rule_parse_value(char *str, char **ret_value, char **ret_endpos) {
+        char *i, *j;
+        int r;
+        bool is_escaped;
+
+        /* value must be double quotated */
+        is_escaped = str[0] == 'e';
+        str += is_escaped;
+        if (str[0] != '"')
+                return -EINVAL;
+        str++;
+
+        if (!is_escaped) {
+                /* unescape double quotation '\"'->'"' */
+                for (i = j = str; *i != '"'; i++, j++) {
+                        if (*i == '\0')
+                                return -EINVAL;
+                        if (i[0] == '\\' && i[1] == '"')
+                                i++;
+                        *j = *i;
+                }
+                j[0] = '\0';
+        } else {
+                _cleanup_free_ char *unescaped = NULL;
+
+                /* find the end position of value */
+                for (i = str; *i != '"'; i++) {
+                        if (i[0] == '\\')
+                                i++;
+                        if (*i == '\0')
+                                return -EINVAL;
+                }
+                i[0] = '\0';
+
+                r = cunescape_length(str, i - str, 0, &unescaped);
+                if (r < 0)
+                        return r;
+                assert(r <= i - str);
+                memcpy(str, unescaped, r + 1);
+        }
+
+        *ret_value = str;
+        *ret_endpos = i + 1;
+        return 0;
+}

src/shared/udev-util.h

@@ -32,3 +32,5 @@ int device_wait_for_initialization(sd_device *device, const char *subsystem, use
 int device_wait_for_devlink(const char *path, const char *subsystem, usec_t deadline, sd_device **ret);
 int device_is_renaming(sd_device *dev);
 bool device_for_action(sd_device *dev, DeviceAction action);
+
+int udev_rule_parse_value(char *str, char **ret_value, char **ret_endpos);

src/test/meson.build

@@ -775,6 +775,10 @@ tests += [
           libselinux],
          '', 'manual', '-DLOG_REALM=LOG_REALM_UDEV'],
 
+        [['src/test/test-udev-util.c'],
+         [],
+         []],
+
         [['src/test/test-id128.c'],
          [],
          []],

src/test/test-udev-util.c

@@ -0,0 +1,202 @@
+/* SPDX-License-Identifier: LGPL-2.1+ */
+
+#include <stdlib.h>
+#include <string.h>
+
+#include "macro.h"
+#include "string-util.h"
+#include "udev-util.h"
+
+static void test_udev_rule_parse_value_one(const char *in, const char *expected_value, int expected_retval) {
+        _cleanup_free_ char *str = NULL;
+        char *value = UINT_TO_PTR(0x12345678U);
+        char *endpos = UINT_TO_PTR(0x87654321U);
+
+        assert_se(str = strdup(in));
+        assert_se(udev_rule_parse_value(str, &value, &endpos) == expected_retval);
+        if (expected_retval < 0) {
+                /* not modified on failure */
+                assert_se(value == UINT_TO_PTR(0x12345678U));
+                assert_se(endpos == UINT_TO_PTR(0x87654321U));
+        } else {
+                assert_se(streq_ptr(value, expected_value));
+                assert_se(endpos == str + strlen(in));
+        }
+}
+
+static void test_parse_value(void) {
+        /* input: "valid operand"
+         * parsed: valid operand
+         * use the following command to help generate textual C strings:
+         * python3 -c 'import json; print(json.dumps(input()))' */
+        test_udev_rule_parse_value_one(
+                "\"valid operand\"",
+                "valid operand",
+                0
+        );
+}
+
+static void test_parse_value_with_backslashes(void) {
+        /* input: "va'l\'id\"op\"erand"
+         * parsed: va'l\'id"op"erand */
+        test_udev_rule_parse_value_one(
+                "\"va'l\\'id\\\"op\\\"erand\"",
+                "va'l\\'id\"op\"erand",
+                0
+        );
+}
+
+static void test_parse_value_no_quotes(void) {
+        test_udev_rule_parse_value_one(
+                "no quotes",
+                0,
+                -EINVAL
+        );
+}
+
+static void test_parse_value_noescape(void) {
+        test_udev_rule_parse_value_one(
+                "\"\\\\a\\b\\x\\y\"",
+                "\\\\a\\b\\x\\y",
+                0
+        );
+}
+
+static void test_parse_value_nul(void) {
+        test_udev_rule_parse_value_one(
+                "\"reject\0nul\"",
+                0,
+                -EINVAL
+        );
+}
+
+static void test_parse_value_escape_nothing(void) {
+        /* input: e"" */
+        test_udev_rule_parse_value_one(
+                "e\"\"",
+                "",
+                0
+        );
+}
+
+static void test_parse_value_escape_nothing2(void) {
+        /* input: e"1234" */
+        test_udev_rule_parse_value_one(
+                "e\"1234\"",
+                "1234",
+                0
+        );
+}
+
+static void test_parse_value_escape_double_quote(void) {
+        /* input: e"\"" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\"\"",
+                "\"",
+                0
+        );
+}
+
+static void test_parse_value_escape_backslash(void) {
+        /* input: e"\ */
+        test_udev_rule_parse_value_one(
+                "e\"\\",
+                0,
+                -EINVAL
+        );
+        /* input: e"\" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\"",
+                0,
+                -EINVAL
+        );
+        /* input: e"\\" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\\\"",
+                "\\",
+                0
+        );
+        /* input: e"\\\" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\\\\\"",
+                0,
+                -EINVAL
+        );
+        /* input: e"\\\"" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\\\\\"\"",
+                "\\\"",
+                0
+        );
+        /* input: e"\\\\" */
+        test_udev_rule_parse_value_one(
+                "e\"\\\\\\\\\"",
+                "\\\\",
+                0
+        );
+}
+
+static void test_parse_value_newline(void) {
+        /* input: e"operand with newline\n" */
+        test_udev_rule_parse_value_one(
+                "e\"operand with newline\\n\"",
+                "operand with newline\n",
+                0
+        );
+}
+
+static void test_parse_value_escaped(void) {
+        /* input: e"single\rcharacter\t\aescape\bsequence" */
+        test_udev_rule_parse_value_one(
+                "e\"single\\rcharacter\\t\\aescape\\bsequence\"",
+                "single\rcharacter\t\aescape\bsequence",
+                0
+        );
+}
+
+static void test_parse_value_invalid_escape(void) {
+        /* input: e"reject\invalid escape sequence" */
+        test_udev_rule_parse_value_one(
+                "e\"reject\\invalid escape sequence",
+                0,
+                -EINVAL
+        );
+}
+
+static void test_parse_value_invalid_termination(void) {
+        /* input: e"\ */
+        test_udev_rule_parse_value_one(
+                "e\"\\",
+                0,
+                -EINVAL
+        );
+}
+
+static void test_parse_value_unicode(void) {
+        /* input: "s\u1d1c\u1d04\u029c \u1d1c\u0274\u026a\u1d04\u1d0f\u1d05\u1d07 \U0001d568\U0001d560\U0001d568" */
+        test_udev_rule_parse_value_one(
+                "e\"s\\u1d1c\\u1d04\\u029c \\u1d1c\\u0274\\u026a\\u1d04\\u1d0f\\u1d05\\u1d07 \\U0001d568\\U0001d560\\U0001d568\"",
+                "s\xe1\xb4\x9c\xe1\xb4\x84\xca\x9c \xe1\xb4\x9c\xc9\xb4\xc9\xaa\xe1\xb4\x84\xe1\xb4\x8f\xe1\xb4\x85\xe1\xb4\x87 \xf0\x9d\x95\xa8\xf0\x9d\x95\xa0\xf0\x9d\x95\xa8",
+                0
+        );
+}
+
+int main(int argc, char **argv) {
+        test_parse_value();
+        test_parse_value_with_backslashes();
+        test_parse_value_no_quotes();
+        test_parse_value_nul();
+        test_parse_value_noescape();
+
+        test_parse_value_escape_nothing();
+        test_parse_value_escape_nothing2();
+        test_parse_value_escape_double_quote();
+        test_parse_value_escape_backslash();
+        test_parse_value_newline();
+        test_parse_value_escaped();
+        test_parse_value_invalid_escape();
+        test_parse_value_invalid_termination();
+        test_parse_value_unicode();
+
+        return EXIT_SUCCESS;
+}

src/udev/udev-rules.c

@@ -990,8 +990,9 @@ static UdevRuleOperatorType parse_operator(const char *op) {
 }
 
 static int parse_line(char **line, char **ret_key, char **ret_attr, UdevRuleOperatorType *ret_op, char **ret_value) {
-        char *key_begin, *key_end, *attr, *tmp, *value, *i, *j;
+        char *key_begin, *key_end, *attr, *tmp;
         UdevRuleOperatorType op;
+        int r;
 
         assert(line);
         assert(*line);
@@ -1031,30 +1032,14 @@ static int parse_line(char **line, char **ret_key, char **ret_attr, UdevRuleOper
         key_end[0] = '\0';
 
         tmp += op == OP_ASSIGN ? 1 : 2;
-        value = skip_leading_chars(tmp, NULL);
+        tmp = skip_leading_chars(tmp, NULL);
+        r = udev_rule_parse_value(tmp, ret_value, line);
+        if (r < 0)
+                return r;
 
-        /* value must be double quotated */
-        if (value[0] != '"')
-                return -EINVAL;
-        value++;
-
-        /* unescape double quotation '\"' -> '"' */
-        for (i = j = value; ; i++, j++) {
-                if (*i == '"')
-                        break;
-                if (*i == '\0')
-                        return -EINVAL;
-                if (i[0] == '\\' && i[1] == '"')
-                        i++;
-                *j = *i;
-        }
-        j[0] = '\0';
-
-        *line = i+1;
         *ret_key = key_begin;
         *ret_attr = attr;
         *ret_op = op;
-        *ret_value = value;
         return 1;
 }