Broadcom at least pay someone to develop a FOSS driver for the VideoCore IV and V, as found in the rPi.
Broadcom at least pay someone to develop a FOSS driver for the VideoCore IV and V, as found in the rPi.
Curiously Qualcomm have been submitting more and more, big, patches for the MSM subsystem, the Linux in-kernel part of the Freedreno graphics driver. I think someone has finally woken up and told management that Linux doesn't take two drivers for the same hardware, and once that reverse-engineered stuff was in the tree, you are royally fucked because you now have to play by the rules of whoever is maintaining it if you want to have any hope of upstreaming your proprietary driver.
This is pretty much the situation for the Vivante polluters. They have been so ridiculously ungrateful for the Linux stuff that makes the chips run for the SoC companies that want to integrate their silicon that by now there is a reverse-engineered driver (Etnaviv) in the tree. They are shut out. And on a longer timescale, the same is coming up for ARM and it's Mali (Lima project).
Everybody uses proprietary firmware for their 802.11ac adapters. Mediatek has gotten some interest because the firmware for some of their 802.11ac chips is fairly minimal and doesn't get in the way of substantially improving the open-source drivers.
Post qualcomm takover, their new chipsets are all heavily locked down, and only functional with a huge closed source driver-esque firmware megablob that is buggy and preforms poorly. You cant even set your mac address anymore.
Tying qualcomm to the atheros reputation for open source friendly chipsets is an insult.