[PHP-DEV] Bug #10538 Updated: mcrypt_generic_init truncates key/iv upon first '\0' From: derick <email protected>
Date: 04/30/01

ID: 10538
Updated by: derick
Reported By: kettler <email protected>
Status: Closed
Bug Type: mcrypt related
PHP Version: 4.0.4pl1
Assigned To: derick
Comments:

It's fixed in CVS as you say (and it is fixed in 4.0.5 too BTW_

Previous Comments:
---------------------------------------------------------------------------

[2001-04-30 05:43:16] kettler <email protected>
Seems to be a duplicate of Bug #8839 and it is already fixed in CVS.

---------------------------------------------------------------------------

[2001-04-28 12:50:42] kettler <email protected>
Same happens in mcrypt_ecb, mcrypt_cbc, mcrypt_cfb and mcrypt_ofb too.

Script showing the bug:

<?php
$key1 = pack("H*", "FF00FF00000000000000000000000000000000000000000000000000000
00000");
$key2 = pack("H*", "FF000000000000000000000000000000000000000000000000000000000
00000");
$iv = pack("H*", "00000000000000000000000000000000");
$plain = pack("H*", "0000000000000000");

$handle = mcrypt_module_open(MCRYPT_TWOFISH, "", MCRYPT_MODE_CFB, "");
mcrypt_generic_init($handle, $key1, $iv);
$crypted1 = mcrypt_generic($handle, $plain);
mcrypt_generic_end($handle);

$handle = mcrypt_module_open(MCRYPT_TWOFISH, "", MCRYPT_MODE_CFB, "");
mcrypt_generic_init($handle, $key2, $iv);
$crypted2 = mcrypt_generic($handle, $plain);
mcrypt_generic_end($handle);

print bin2hex($plain)."nn";
print bin2hex($crypted1)."nn";
print bin2hex($crypted2)."nn";
?>

The two ciphertexts should NOT be the same as the key is different.

Proposed patch (also fixes a possible memory access problem, but only for the mcrypt_generic_init function, the I didn't fully understand php_mcrypt_do_crypt yet, when I do I will update the patch, see also Bug #10518):

--- php-4.0.4pl1/ext/mcrypt/mcrypt.c Wed Nov 22 22:40:15 2000
+++ php-4.0.4pl1-sk/ext/mcrypt/mcrypt.c Sat Apr 28 18:53:07 2001
@@ -463,14 +463,22 @@
                        Z_STRLEN_PP(key), key_size);
                php_error (E_NOTICE, dummy);
        }
- strncpy (key_s, Z_STRVAL_PP(key), key_size);
+ if (Z_STRLEN_PP(key) > key_size) {
+ memcpy (key_s, Z_STRVAL_PP(key), key_size);
+ } else {
+ memcpy (key_s, Z_STRVAL_PP(key), Z_STRLEN_PP(key));
+ }
 
        if (Z_STRLEN_PP(iv) != iv_size) {
                sprintf (dummy, "iv size incorrect; supplied length: %d, needed: %d",
                        Z_STRLEN_PP(iv), iv_size);
                php_error (E_WARNING, dummy);
        }
- strncpy (iv_s, Z_STRVAL_PP(iv), iv_size);
+ if (Z_STRLEN_PP(iv) > iv_size) {
+ memcpy (iv_s, Z_STRVAL_PP(iv), iv_size);
+ } else {
+ memcpy (iv_s, Z_STRVAL_PP(iv), Z_STRLEN_PP(iv));
+ }
 
        RETVAL_LONG (mcrypt_generic_init (td, key_s, key_size, iv_s));
        efree (iv_s);

---------------------------------------------------------------------------

ATTENTION! Do NOT reply to this email!
To reply, use the web interface found at http://bugs.php.net/?id=10538&edit=2 

-- 
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, e-mail: php-dev-unsubscribe <email protected>
For additional commands, e-mail: php-dev-help <email protected>
To contact the list administrators, e-mail: php-list-admin <email protected>