Description of problem:
Issuing a restart command for the nginx daemon via the /etc/init.d/nginx script results in a crash.


Version-Release number of selected component (if applicable):
nginx.x86_64                               0.6.39-2.el5                installed
Linux XXXXXX 2.6.18-128.1.10.el5 #1 SMP Thu May 7 10:35:59 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux


How reproducible:
Every Time

Steps to Reproduce:
1. > /etc/init.d/nginx restart


Actual results:
    > /etc/init.d/nginx restart
    2010/01/12 16:28:43 [info] 10263#0: the configuration file /etc/nginx/nginx.conf was tested successfully
    Stopping nginx:                                            [  OK  ]
    Starting nginx:

    "Hey, where's my OK?"
    Log file shows "panic: MUTEX_LOCK (22) [op.c:352]."


    "Maybe a bad config?"
    > /etc/init.d/nginx start
    Starting nginx:                                            [  OK  ]

    "Nope, starts just fine"

  Expected results:
    nginx daemon should restart successfully.
    > /etc/init.d/nginx restart
    Stopping nginx:                                            [  OK  ]
    Starting nginx:                                            [  OK  ]


  Additional info:

    The MUTEX_LOCK error is interesting. Seems to be an issue with the daemon starting up before the cleanup is complete.

    Simple workaround until the root cause is identified is to insert a sleep in the restart section of the init script.
    Even a 1 second sleep is enough to avoid the MUTEX_LOCK issue:

    -/etc/init/d/nginx-
    ----------------------
    ...
    restart() {
        configtest || return $?
        stop
        # wait for cleanup to complete
        sleep 1
        start
    }
    ...
    ----------------------