Bug #79550 [Fbk->NoF]: Reloading the httpd service repeatedly causes a memory corruption error.

Edit report at https://bugs.php.net/bug.php?id=79550&edit=1 ID: 79550 Updated by: php-bugs@lists.php.net Reported by: kitagawa at sraoss dot co dot jp Summary: Reloading the httpd service repeatedly causes a memory corruption error. -Status: Feedback +Status: No Feedback Type: Bug Package: Apache2 related Operating System: Linux PHP Version: 7.3.17 Assigned To: cmb Private report: N New Comment: No feedback was provided. The bug is being suspended because we assume that you are no longer experiencing the problem. If this is not the case and you are able to provide the information that was requested earlier, please do so and change the status of the bug back to "Re-Opened". Thank you. Previous Comments: ------------------------------------------------------------------------ [2021-03-04 12:38:21] cmb@php.net If that memory corruption still happens for you with any of the actively supported PHP versions[1], we would at least need a stack backtrace[2] to have a chance to resolve the issue. [1] <https://www.php.net/supported-versions.php> [2] <https://bugs.php.net/bugs-generating-backtrace.php> ------------------------------------------------------------------------ [2020-05-07 02:10:48] kitagawa at sraoss dot co dot jp Thank you for your comment. Did a similar problem occur with "imap.so"? Currently, we continue to provide services including restart, but we want to restart and operate as safely as possible. Also, there are many servers that are graceful restart all over the world(including logrotate), so I would like to ask for a fundamental response. ------------------------------------------------------------------------ [2020-05-01 08:25:05] bugreports at gmail dot com chck out which extensions are loaded, on Fedora for some time "imap.so" loaded had a similar result so that i switched to restart instead of reload on machins which need it ------------------------------------------------------------------------ [2020-05-01 07:33:58] kitagawa at sraoss dot co dot jp Description: ------------ Hi, I'm running Apache 2.4.41 on CentOS 6.4 server. There was a problem (memory corruption), when graceful reloading the httpd service many times. The version of each app is as follows: - httpd-2.4.41 - apr-1.5.2 - apr-util-1.5.4 - php-7.3.4 The following steps will cause the problem. # cat /opt/httpd/htdocs/index.php ------ <?php phpinfo(); ?> ------ # /etc/init.d/httpd start Starting httpd: [ OK ] Display phpinfo() in your browser and check that it is working. http://localhost/index.php Now, when I run the following script, an error occurs (about 1023 times). # for i in $(seq 1030) ; do echo $i; /opt/httpd/bin/httpd -k graceful; done # less /opt/httpd/log/error_log ・・・・ [XXX XXX XX 16:05:23.261688 2020] [mpm_prefork:notice] [pid 2875] AH00171: Graceful restart requested, doing restart *** glibc detected *** /opt/httpd/bin/httpd: malloc(): memory corruption: 0x00000000021a4010 *** *** glibc detected *** /opt/httpd/bin/httpd: malloc(): memory corruption: 0x00000000021a4010 *** This bug occurs when loading the libphp7.so module. Different libphp7.so versions have the same result. But that doesn't happen with the libphp5.so module. It occurs on CentOS 7 as well as CentOS 6. Thanks, Regards. ------------------------------------------------------------------------ -- Edit this bug report at https://bugs.php.net/bug.php?id=79550&edit=1