feat: 450.66 release

[mmstick]

Currently testing on the oryp2

[leviport]

I'm seeing the same bug I saw with 450.57: https://github.com/pop-os/nvidia-graphics-drivers/pull/52#issuecomment-658199651

I'm only seeing it in Hybrid mode, after adding the 450.51 patch to enable reverse prime offloading. Could the issue be with reverse prime instead of the driver itself?

I'm going to remove 450.51 and test. If this driver is okay without reverse prime, we might be able to merge it and worry about reverse prime independently.

Edit: The same issue persists in hybrid mode without the 450.51 branch added. Any secondary displays arranged above or to the left of the primary display make the bug appear, but arranging secondary displays below or to the right of the primary works as expected. In Nvidia mode, all external display arrangements work as expected. In Integrated mode, external displays do not work at all, as expected.

[mmstick]

I wonder if NVIDIA is aware that their long-lived driver has this bug.

[malt3]

I wonder if NVIDIA is aware that their long-lived driver has this bug.

i think they noticed: https://forums.developer.nvidia.com/t/the-all-new-outputsink-feature-aka-reverse-prime/129828/10 https://forums.developer.nvidia.com/t/450-75-with-reverse-prime-xrandr-scale-2x2-does-not-work-renders-only-1-4-of-screen-problem-with-nvidia-settings/142424/2

[geodro]

Any updates on this, or how can we help test it?

[mmstick]

@geodro It's already been tested. The 450.66 driver is a bust, so we must wait until NVIDIA releases a new driver, and hope that it has resolved this regression.

[jackpot51]

Obsolete by #57