#asahi-dev · 2021-02-11 18:31
marcananything that looks like PCI to software is PCI for all intents and purposes#asahi-dev · 2021-02-11 18:31
marcanPCI doesn't have to be a physical bus#asahi-dev · 2021-02-11 18:31
marcanI mean they aren't "fake"#asahi-dev · 2021-02-11 18:29
marcanlike any Ryzen etc#asahi-dev · 2021-02-11 18:29
marcanthat's how every x86 SoC works#asahi-dev · 2021-02-11 18:29
marcanI mean if you *actually* have embedded PCI devices that's fine#asahi-dev · 2021-02-11 18:29
marcanlol#asahi-dev · 2021-02-11 18:29
marcanarnd: oh cool!#asahi-dev · 2021-02-11 18:28
marcanthat's very good to know#asahi-dev · 2021-02-11 18:28
marcanah, that explains why I could've sworn I saw xHCI somewhere nested under that#asahi-dev · 2021-02-11 18:28
marcankettenis: I thought it was dwc3, which isn't xHCI? or is it?#asahi-dev · 2021-02-11 18:27
marcanlike they wrote a driver for a PCIe controller that doesn't exist, that just returns static data for everything#asahi-dev · 2021-02-11 18:27
marcanI don't know what made them think that was a good idea#asahi-dev · 2021-02-11 18:27
marcancorellium are... doing weird hacky stuff, and for some reason they decided to just make up a fake PCIe bus so the linux NVMe driver would "find" the NVMe device#asahi-dev · 2021-02-11 18:26
marcanyes#asahi-dev · 2021-02-11 18:26
marcanand both work fine#asahi-dev · 2021-02-11 18:26
marcane.g. USB xHCI supports both, so you can have PCI USB3.0 adapter cards, and built-in USB 3.0 support in ARM SoCs#asahi-dev · 2021-02-11 18:26
marcanand the solution is to just add support for that, just like every other device type that has to support both#asahi-dev · 2021-02-11 18:26
marcanthe "problem" is that the Linux NVMe driver does not support devicetree, because this is the first device ever that has NVMe that *isn't* PCIe#asahi-dev · 2021-02-11 18:26
marcanyes#asahi-dev · 2021-02-11 18:26
marcanyes#asahi-dev · 2021-02-11 18:25
marcanyes#asahi-dev · 2021-02-11 18:25
marcanon arm64 there's basically devicetree or ACPI, depending on the system, and PCI(e)#asahi-dev · 2021-02-11 18:24
marcanon x86 there's ACPI, there's PCI(e), there's legacy ISA PnP stuff we should avoid talking about, there's semi-fixed stuff like firmware, and there's a pile of one-off junk like the local APICs#asahi-dev · 2021-02-11 18:23
marcanthis isn't really about physical connections, it's about how you find devices#asahi-dev · 2021-02-11 18:22
marcanon x86 we have.... lots of things.#asahi-dev · 2021-02-11 18:22
marcanyeah the NVMe is not PCIe#asahi-dev · 2021-02-11 18:21
marcanyou're probably thinking of my comments about how corellium faked a PCIe bus instead of adding non-PCIe support to nvme#asahi-dev · 2021-02-11 18:21
marcan(and thunderbolt)#asahi-dev · 2021-02-11 18:21
marcanlike, physical PCIe devices#asahi-dev · 2021-02-11 18:21
marcanGlanzmann: only PCIe devices are on PCIe#asahi-dev · 2021-02-11 18:20
marcanHV-phys, HV-virt, guest-phys, guest-virt.#asahi-dev · 2021-02-11 18:20
marcanthere are 4 timers.#asahi-dev · 2021-02-11 18:20
marcanmodwizcode: it's.... complicated.#asahi-dev · 2021-02-11 18:18
marcan-s#asahi-dev · 2021-02-11 18:18
marcanaswesome#asahi-dev · 2021-02-11 18:18
marcanheh :)#asahi-dev · 2021-02-11 18:18
marcanhope I can send it tomorrow, and then if this moves forward well we should be able to increase our parallellism.#asahi-dev · 2021-02-11 18:17
marcanthanks :)#asahi-dev · 2021-02-11 18:16
marcanGlanzmann: I think #asahi-offtopic is probably a better place to discuss this#asahi-dev · 2021-02-11 17:57
marcanso we would probably be fine for that, *except* for the double mapping potentially causing issues#asahi-dev · 2021-02-11 17:57
marcanthat's good to know#asahi-dev · 2021-02-11 17:30
marcanI have no idea what they're doing and whatever was just discussed sounds... quite horrible.#asahi-dev · 2021-02-11 17:29
marcanI don't think we can really support the corellium kernel/distro here#asahi-dev · 2021-02-10 16:05
marcanthat would imply >1 stopbit, assuming you keep the buffer full#asahi-dev · 2021-02-10 02:08
marcanand the "1" in "8N1" means the stop bit ought to have the same duration as all the other bits too#asahi-dev · 2021-02-10 02:08
marcansegher: start bits have the same duration as all the other bits, otherwise the sampling would be off#asahi-dev · 2021-02-09 17:45
marcanwhich means https://twitter.com/marcan42/status/1043063673080569856 could've been using exactly that rate (but I actually ran it faster because buffers and stuff)#asahi-dev · 2021-02-09 17:44
marcanfun fact: https://www.sparkfun.com/products/12923 these are 64 * 48 pixels, monochrome. that means driving them at 30 FPS requires 64 * 48 * 30 / 8 = 11520 bytes per second, or precisely 115200 baud with standard 8N1 serial settings.#asahi-dev · 2021-02-09 17:42
marcanand a 3 because 3s