apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aaron Bannert <aa...@clove.org>
Subject Re: testglobalmutex.c
Date Mon, 15 Mar 2004 19:15:13 GMT

On Mar 15, 2004, at 10:56 AM, rbb@rkbloom.net wrote:
> Nope.  Take a look at the thread that is complaining about the API for
> global_mutex and proc_mutex.  Without that global_mutex_create call, 
> the program
> segfaults on any Unix machine.  This is a bug in the _child_init API, 
> because it
> should accept a lockmech_e argument so that the lock can be created 
> properly.

Actually the bug is that the unix implementation of 
apr_global_mutex_child_init
assumes the (apr_global_mutex **mutex) is valid, and it shouldn't.

So yes, we should pass the lock mechanism to the child_init so that
the struct can be allocated by the proc mutex implementation.

-aaron


Mime
View raw message