Bug #70150 [Opn->Dup]: $iv = mcrypt_create_iv( $rem, MCRYPT_DEV_URANDOM ); times out
Edit report at https://bugs.php.net/bug.php?id=70150&edit=1
ID: 70150
Updated by: ab@php.net
Reported by: gessel at blackrosetech dot com
Summary: $iv = mcrypt_create_iv( $rem, MCRYPT_DEV_URANDOM
); times out
-Status: Open
+Status: Duplicate
Type: Bug
Package: mcrypt related
Operating System: FreeBSD 10.1-RELEASE #0 r285883
PHP Version: 5.5.27
Block user comment: N
Private report: N
New Comment:
@gessel, please check the latest 5.6 or 7 branch/snapshot (no release yet). This is likely a
duplicate of bug #69833.
Thanks.
Previous Comments:
------------------------------------------------------------------------
[2015-07-27 17:03:45] cmb@php.net
MCRYPT_DEV_URANDOM|MCRYPT_RAND has the same behavior as
MCRYPT_RAND, i.e. the else clause[1] will be processed. I have
deleted Cory's note.
[1] <https://github.com/php/php-src/blob/php-5.6.11/ext/mcrypt/mcrypt.c#L1426-L1469>
------------------------------------------------------------------------
[2015-07-27 07:24:50] gessel at blackrosetech dot com
Description:
------------
---
From manual page: http://www.php.net/function.mcrypt-create-iv
---
MediaWiki's MWCryptRand.php calls
$iv = mcrypt_create_iv( $rem, MCRYPT_DEV_URANDOM );
This times out on FreeBSD without returning a value (even with max_execution_time set to 3000).
modifying the call to
$iv = mcrypt_create_iv( $rem, MCRYPT_DEV_URANDOM|MCRYPT_RAND );
returns as expected, however, the MW dev is concerned that MCRYPT_RAND is not cryptographically
secure, and that CC's note at https://secure.php.net/manual/en/function.mcrypt-create-iv.php#117047
is incorrect and that this piped construction will return "3" (rather than 0,1,2 or as
Cory states, first attempt 1 and if that isn't available, try 2) and with "3" will
always initialize with the crytographically insecure MCRYPT_RAND.
Test script:
---------------
$iv = mcrypt_create_iv( $rem, MCRYPT_DEV_URANDOM );
------------------------------------------------------------------------
--
Edit this bug report at https://bugs.php.net/bug.php?id=70150&edit=1
Thread (3 messages)