1. 27
  1. 5

    I’m impressed by the effort that went into this, but I’ve got to ask: Why?

    This is the sort of thing that ARM could be paying someone to do. Or at the very least, ARM should be giving these guys a development board so that they can properly do development. Or if they don’t want to help, you can contact the chip vendor… though dev boards are in general in short supply.

    1. 3

      I’d love to know why that’s happening for Arm. I kind of get it for NVIDIA with its secrecy (even though I think it’s silly), but it’s not like Arm has some crazy private optimisations that they need to keep market advantage. It’s literally in Mediatek’s interest to throw money and hardware at Collabora people.

      1. 3

        Oh, I would think Collabora is sponsored by many of Arm’s direct customers. As you say, it’s in their interest. Apparently, Arm thinks it has too many secrets. I get it if they worry about their IP, since they don’t sell chips, but they got to realize they aren’t selling drivers either. Or are they? I think Arm will have to re-evaluate which secrets to worry about if Collabora continues to out-reverse them. Which btw. I think is a fantastic job worth supporting.

    2. 3

      Apple’s input/output memory management unit uses larger pages, 16384 bytes (16K) large. As a consequence, when we run Linux bare metal on Apple platforms, we configure Linux to use 16K pages everywhere to keep life simple.

      I did not know this about M1. That’s wild.