[Linux-ha-dev] Re: [Linux-ha-cvs] Linux-HA CVS: heartbeat by sunjd from

[Lars Marowsky-Bree]

On 2004-09-01T09:43:51,
   linux-ha-cvs at lists.community.tummy.com said:

> linux-ha CVS committal
> 
> Author  : sunjd
> Host    : 
> Project : linux-ha
> Module  : heartbeat
> 
> Dir     : linux-ha/heartbeat
> 
> 
> Modified Files:
> 	hb_api.c 
> 
> 
> Log Message:
> minor fix to avoid severe memory leak ;-)
> 
> ===================================================================
> RCS file: /home/cvs/linux-ha/linux-ha/heartbeat/hb_api.c,v
> retrieving revision 1.109
> retrieving revision 1.110
> diff -u -3 -r1.109 -r1.110
> --- hb_api.c	31 Aug 2004 21:45:01 -0000	1.109
> +++ hb_api.c	1 Sep 2004 09:43:51 -0000	1.110
> @@ -1,4 +1,4 @@
> -/* $Id: hb_api.c,v 1.109 2004/08/31 21:45:01 alan Exp $ */
> +/* $Id: hb_api.c,v 1.110 2004/09/01 09:43:51 sunjd Exp $ */
>  /*
>   * hb_api: Server-side heartbeat API code
>   *
> @@ -1071,7 +1071,10 @@
>  	}
>  	api_send_client_msg(client, resp);
>  	/* What to do with 'resp'.  Do we need it any more?? FIXME!! */
> +	/* I think we don't need it anymore when go here, or memory leak
> +	   will be very severe. You don't test it;-)	
>  	goto del_msg;
> +	*/
>  
>  del_rsp_and_msg:
>  	if (resp != NULL) {

I'm pretty convinced this bugfix is wrong and creates memory corruption.
_If_ the message has been send successfully, the IPC library will
eventually free() it itself.

The caller is only supposed to free() received messages.


Sincerely,
    Lars Marowsky-Brée <lmb at suse.de>

-- 
High Availability & Clustering	   \\\  /// 
SUSE Labs, Research and Development \honk/ 
SUSE LINUX AG - A Novell company     \\//