Page 1 of 1

Sliced front buffer rendering

Posted: Fri Sep 07, 2018 7:12 pm
by dave j
Qualcomm and ARM GPUs have the ability to render to part of the front buffer without requiring loading and saving the rest of it - which is slow. (Qualcom have their QCOM_tiled_rendering extension and ARM uses scissor boxes.) It also looks like the Eric Anholt's driver supports it using scissoring.

Do the Broadcom drivers for the VC4 provide such a facility?