Re: [vbox-dev] VirtualBox 5.x IDisplaySourceBitmap life cycle

2016-02-03 Thread Rūdolfs Bundulis
I could not even find this file in the SVN. Is the DisplaySourceBitmapWrap.h some kind of automatically generated file? 2016-02-02 17:36 GMT+02:00 Rūdolfs Bundulis : > I tried to find this by inspecting the source but found something weird. I > downloaded the VirtualBox-5.0.14 sources and initial

Re: [vbox-dev] VirtualBox 5.x IDisplaySourceBitmap life cycle

2016-02-03 Thread Michal Necasek
Yes, it is an automatically generated file, a COM wrapper based on the IDL description. It does make navigating the source code a bit harder. - Original Message - From: rudolfs.bundu...@gmail.com To: vbox-dev@virtualbox.org Sent: Wednesday, February 3, 2016 10:35:33 AM GMT +01:00 Ams

Re: [vbox-dev] VirtualBox 5.x IDisplaySourceBitmap life cycle

2016-02-03 Thread Rūdolfs Bundulis
> Yes, it is an automatically generated file, a COM wrapper based on the IDL description. It does make navigating the source code a bit harder. Ok thanks for clearing this out. I managed to find the place where the memory is actually handled - DisplaySourceBitmap::initSourceBitmap() in DisplaySo

Re: [vbox-dev] VirtualBox 5.x IDisplaySourceBitmap life cycle

2016-02-03 Thread Vitali Pelenjow
Hi, the VRAM pointer will be valid, because it points to the memory allocated for the VM graphics device. The memory is deallocated when VM shuts down. However the memory content might change of course. Vitali On 2/3/2016 11:01 AM, Rūdolfs Bundulis wrote: > > Yes, it is an automatically gener

Re: [vbox-dev] VirtualBox 5.x IDisplaySourceBitmap life cycle

2016-02-03 Thread Rudolfs Bundulis
Hi Vitali, Thanks for the clarifications. I was mostly worried about an access violation. If the pointer is valid then I can just make some invalidation mechanism that will drop all pending updates after a format change and skip any frames that were being encoded at the moment of notifiaction i