By Date: <-- -->
By Thread: <-- -->

Memory leak in SR 6.0 under d42_chnstatus

From: David LaRue


Hello,

I'm in the process of validating my Dialogic D/82JCT-U application on SR 6.0. My current customers use SR 5.1.1 SP1.

There was a problem in SR 5.1.1 SP1 that led to unreported disconnect events. My solution at the time was to have my service loop check the status of each line once a minute. The combination of my application, SR 5.1.1 SP1, and Windows 2000 Professional did not appear to have any memory leaks.

I've been retesting my application compiled for SR 6.0 and with Windows 2003 Server. I have noticed that the periodic call to d42_chnstatus() is leaking memory at the rate of perhaps 12-32KB every ten minutes. There are perhaps 24-36 calls to d42_chnstatus() during this time period.

I've not found any difference in the API documentation regarding this call. All of the parameters are statically allocated, so it appears to be a leak internal to the Dialogic API. Increasing the number of checks per minute greatly increases the memory loss. While I can reduce the number of calls, there would still be a noticeable memory leak after several weeks operation.

Have you seen a problem like this or read about a solution?

Thank you. 






To reply: mailto:General.31085 (at) pysdiscussext.py.intel.com
To start a new topic: mailto:General (at) pysdiscussext.py.intel.com
To login: http://membersresource.intel.com/webboardnet/wb.aspx?board=general