Bug #75107 [Opn]: opcache_reset will cause running FPM process stop with SIGKILL

Edit report at https://bugs.php.net/bug.php?id=75107&edit=1 ID: 75107 Updated by: nikic@php.net Reported by: greatdeer at gmail dot com Summary: opcache_reset will cause running FPM process stop with SIGKILL Status: Open Type: Bug Package: opcache Operating System: CentOS6.5 PHP Version: 5.6.31 Block user comment: N Private report: N New Comment: opcache_reset() triggers an opcache restart. If a process holds onto an SHM lock for more than opcache.force_restart_timeout seconds after the restart is initiated, it will be killed. See also http://php.net/manual/en/opcache.configuration.php#ini.opcache.force-restart-timeout. Previous Comments: ------------------------------------------------------------------------ [2017-08-23 06:58:53] greatdeer at gmail dot com Description: ------------ I use Nginx and PHP-FPM, Occasionally, php-fpm.log will report SIGKILL ("WARNING: [pool www] child 12197 exited on signal 9 (SIGKILL) after 2046.724398 seconds from start") with no more messages. I look for the reason for a long time, and finally I found it. It was opcache_reset. When I update the php codes, opcache_reset() will be triggered to make the codes validate immediately. And sometimes some scripts were executing just the same time and would crash. Is it a bug? or I should avoid useing opcache_reset()? ------------------------------------------------------------------------ -- Edit this bug report at https://bugs.php.net/bug.php?id=75107&edit=1