Hi,<div><br></div><div>please keep the discussion on the list.<br><br><div class="gmail_quote">On Tue, May 8, 2012 at 11:29 AM, <span dir="ltr"><<a href="mailto:juddsim@gmail.com" target="_blank">juddsim@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Each instance created allocates 40mb (I'm just looking at the task manager) and eventually the app gets to like 2 gigs and crashes.<br>
</blockquote><div><br></div><div>The task-manager is not necessarily a good leak-detector. Can you provide a small self-contained example which demonstrates this? I've </div><div>just tried here with a small video thats replayed and each 5 seconds the player instance is stopped, removed from the layout and del'ed and</div>
<div>the memory consumption doesn't even reach 20M - thats minus the shared memory already. The important bit seems to be to stop the playing</div><div>url, otherwise the player-object seems to not be garbage collected.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
I'm not familiar with a valgrind log file.<br></blockquote><div><br></div><div>Valgrind is a multi-purpose tool, among others it can help you detect memory leaks in applications. Its Linux-Only though, but there are similar tools for other platforms afaik. A task-manager's memory-column is not always a good measurement for detecting actual memory leaks.</div>
<div><br></div><div>Andreas </div></div></div>