Re: svn: /php/php-src/ branches/PHP_5_3/NEWS branches/PHP_5_3/ext/pdo/pdo_dbh.c branches/PHP_5_3/ext/pdo/php_pdo_driver.h
branches/PHP_5_3/ext/pdo_pgsql/pdo_pgsql.c branches/PHP_5_3/ext/pdo_pgsql/pgsql_driver.c branches/PHP_5_3/ext/pdo_pgsql/php_pdo_pgsql_int.h
branches/PHP_5_3/ext/pdo_pgsql/tests/is_in_transaction.phpt trunk/ext/pdo/pdo_dbh.c trunk/ext/pdo/php_pdo_driver.h trunk/ext/pdo_pgsql/pdo_pgsql.c
trunk/ext/pdo_pgsql/pgsql_driver.c trunk/ext/pdo_pgsql/php_pdo_pgsql_int.h trunk/ext/pdo_pgsql/
| From: | Johannes Schlüter | Date: | Mon, 14 Jun 2010 14:26:47 +0000 |
| Subject: | Re: svn: /php/php-src/ branches/PHP_5_3/NEWS branches/PHP_5_3/ext/pdo/pdo_dbh.c branches/PHP_5_3/ext/pdo/php_pdo_driver.h branches/PHP_5_3/ext/pdo_pgsql/pdo_pgsql.c branches/PHP_5_3/ext/pdo_pgsql/pgsql_driver.c branches/PHP_5_3/ext/pdo_pgsql/php_pdo_pgsql_int.h branches/PHP_5_3/ext/pdo_pgsql/tests/is_in_transaction.phpt trunk/ext/pdo/pdo_dbh.c trunk/ext/pdo/php_pdo_driver.h trunk/ext/pdo_pgsql/pdo_pgsql.c trunk/ext/pdo_pgsql/pgsql_driver.c trunk/ext/pdo_pgsql/php_pdo_pgsql_int.h trunk/ext/pdo_pgsql/ |
||
| References: | 1 | Groups: | php.cvs php.internals php.pdo |
| Request: | Send a blank email to internals+get-48777@lists.php.net to get a copy of this message | ||
On Thu, 2010-06-10 at 12:11 +0000, Ilia Alshanetsky wrote:
> Modified: php/php-src/branches/PHP_5_3/ext/pdo/php_pdo_driver.h
> ===================================================================
> ---
> php/php-src/branches/PHP_5_3/ext/pdo/php_pdo_driver.h 2010-06-10
> 11:45:51 UTC (rev 300350)
> +++
> php/php-src/branches/PHP_5_3/ext/pdo/php_pdo_driver.h 2010-06-10
> 12:11:19 UTC (rev 300351)
> @@ -310,6 +310,7 @@
> pdo_dbh_check_liveness_func check_liveness;
> pdo_dbh_get_driver_methods_func get_driver_methods;
> pdo_dbh_request_shutdown persistent_shutdown;
> + pdo_dbh_txn_func in_transaction;
> };
>
> /* }}} */
Here you are changing a structure which is allocated and initialized in
a driver and then read from the PDO core. PDO core will therefore read
invalid memory when a driver compiled against 5.3.2 is used in 5.3.3
while we usually guarantee binary compatibility in bug fix releases.
This for instance affects distributors or MSFT's sqlsrv driver.
johannes