Emma Anholt
Emma Anholt
URL: https://www.96boards.org/documentation/guides/ db820c board-specific guides link goes to the db410c link too, looks like just copy and paste fail as I can s/41/82/ and get the right link (though that...
These are going to be tricker than the other panels: We need to use the transposer to write back into memory, then when the transposer is done we need to...
Atomic check should be making sure we can actually allocate the LBM space. This will probably require making the LBM allocation at check time in the plane_state, and using similar...
Need to be able to get hang state back out for debug. Maybe not merge blocker
The bin bo isn't being included any more, missing much of the useful hang state.
When we're scanning out (for example) to a RGB565 panel from an RGB888 framebuffer, we should enable HVS's dithering. The firmware does this for DPI panels.
danvet said about our clock_select detection in `vc4_crtc.c`: > Hm, so the usual way we solve the "crtc needs information from the encoder > problem" is to add bits to...
It would be nice if we used our FIFO underrun interrupts to detect when we've screwed something up in our modesetting sequence. Enable the `DSPnEISLUR` interrupts, and debug all instances...
If we read out the current state of our PV/HVS/HDMI registers at boot, we could potentially recognize that the initial mode that fbdev chooses matches what the firmware had configured,...
When we're only doing 1:1 RGBA scaling and aren't using the LBM memory, we could power it down (bit 30 of `SCALER_DISPCTRL`. Note that it takes 1000 cycles of the...