I was surprised that I found there is /*resource leak*/ in the comment
It turns out that Beam complains about it and I thought BEAM made a
mistake there.
"
1657 : gshi 1.32 /* BEAM think this is a memory leak, but
actually it is handle nicely
1658 : * in cl_msg_replace: old value is freed and new
value is saved
1659 : */
"
Actually it was my mistake.
-Guochun
[email protected] wrote:
linux-ha CVS committal
Author : lars
Host :
Project : linux-ha
Module : lib
Dir : linux-ha/lib/clplumbing
Modified Files:
cl_msg_types.c
Log Message:
Bug #1162 (fix by gshi): cl_msg_replace() duplicate the value and assign
it to the field, frees the old value, i.e. in compress2uncompress, the
msg from wirefmt2msg() in line 1650, cl_msg_types.c is lost, therefore
memory leaks.
===================================================================
RCS file: /home/cvs/linux-ha/linux-ha/lib/clplumbing/cl_msg_types.c,v
retrieving revision 1.37
retrieving revision 1.38
diff -u -3 -r1.37 -r1.38
--- cl_msg_types.c 17 May 2006 23:22:33 -0000 1.37
+++ cl_msg_types.c 10 Jul 2006 19:09:57 -0000 1.38
@@ -1639,6 +1639,7 @@
char buf[MAXMSG];
size_t buflen = MAXMSG;
struct ha_msg* msgfield;
+ int err;
if (cl_decompress_field(msg, index, buf, &buflen) != HA_OK){
cl_log(LOG_ERR, "%s: compress field failed",
@@ -1656,7 +1657,11 @@
/* BEAM think this is a memory leak, but actually it is handle nicely
* in cl_msg_replace: old value is freed and new value is saved
*/
- return cl_msg_replace(msg, index, (char*)msgfield, 0, FT_UNCOMPRESS);/*
resource leak */
+ err = cl_msg_replace(msg, index, (char*)msgfield, 0, FT_UNCOMPRESS);/*
resource leak */
+
+ ha_msg_del(msgfield);
+
+ return err;
}
/*
_______________________________________________
Linux-ha-cvs mailing list
[email protected]
http://lists.community.tummy.com/mailman/listinfo/linux-ha-cvs
_______________________________________________________
Linux-HA-Dev: [email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
