Before DMA Async updates changes for T3.6 and T4

This is before merge of the Pull request that changed the T4 to not copy the data from Frame buffer to smaller buffers that update the screen. This was the way earlier to handle how the T4 in DMAMEM and malloc memory cache works with DMA...