Closed macdudeuk closed 7 years ago
Hey,
i can't reproduce this yet, i tried the given pipeline on both linux and OSX; i'm monitoring memory usage with top/htop, and it seems constant while running for a minute or so.
Are you doing anything with the appsink buffers that might prevent them from getting garbage collected? Can you send an example script?
Edit: sorry, after re-reading your report, i tried with that exact pipeline in appsink.js. Still can't observe any leaking though. Can you tell me a bit about your environment? OS, Gstreamer version, node version?
Another edit: Seems i was looking in the wrong spot. I can see memory filling up now. I'll see if i can find anything..
turns out, if you gst_buffer_map(), you should also gst_buffer_unmap().
just as a sidenote: we could probably avoid the memcpy implied in Nan::CopyBuffer by deferring gst_buffer_unmap to a Nan::FreeCallback
thanks for reporting!
That was quick!
Thanks for that.
/A
Hi
I am getting a consistent memory leak when pulling data from appsink. This is true using the example apps too. For instance if you run appsink.js with the following pipeline:
videotestsrc is-live=true ! appsink max-buffers=1 drop=true name=sink
Then it quickly eats all available memory.
Using node inspector, the leak does not seem to be in the node.js environment, and checking the code in GObjectWrap.cpp there is a call to gst_sample_unref() so I am confused where it is leaking.
Any help appreciated.