#asahi-dev · 2022-01-31 07:45
marcanwe shouldn't need any of the optional features#asahi-dev · 2022-01-31 07:45
marcanwe don't need the batch feature#asahi-dev · 2022-01-31 07:02
marcananyway poke around and see if you can show me how things would look; I'll do that myself too and decide at some point, but first I've got more pressing things on my plate :)#asahi-dev · 2022-01-31 07:02
marcan(fwiw: I know cargo makes life easier for more complex projects, especially around dependency updates... but I'm not too worried about that because the scope of this part of the code is so simple, I don't see us ever really needing to upgrade the deps, unless there is some catastrophic security fail in the ed25519 package or something)#asahi-dev · 2022-01-31 07:01
marcanright, so ideally we'd build that vendor dir out of submodules; alternatively we could just vendor in the deps in the main git tree, which we already do for tinf/lzma/vsprintf/dlmalloc/chunks of pdlibc anyway#asahi-dev · 2022-01-31 06:59
marcanI mean what would the tarball contain?#asahi-dev · 2022-01-31 06:58
marcanhm, what's different between that and submodules?#asahi-dev · 2022-01-31 06:58
marcan(this is partially driven by me having had, er, bad experiences with language-native package manager stuff when doing things like building distro packages... the tool is fine, going out to the internet to download stuff isn't)#asahi-dev · 2022-01-31 06:57
marcanif we can get cargo to use pre-downloaded assets for all the deps and never talk to the network, then using cargo to build the rust chunk is fine#asahi-dev · 2022-01-31 06:55
marcanI've been meaning to learn meson, but I don't know if doing that on top of the rust stuff is really called for right now :p#asahi-dev · 2022-01-31 06:55
marcanis it really that bad? I thought it was just a compiler down behind the scenes#asahi-dev · 2022-01-31 06:50
marcanwhich I think is worth it; as cool as rust is, it's worth keeping m1n1 builds as friction free as possible for folks, and the big part that benefits from it is not something needed for most use cases#asahi-dev · 2022-01-31 06:49
marcanalso, since this code is only for chainloads, it means we can support the use case of people installing "official" m1n1 in non-chainload mode (from a binary) and then doing local m1n1 builds for development, without depending on rust at all#asahi-dev · 2022-01-31 06:46
marcanand for sigs, https://github.com/dalek-cryptography/ed25519-dalek (with all the optional stuff turned off)#asahi-dev · 2022-01-31 06:46
marcanfor fat32 I was thinking of using https://crates.io/crates/fat32#asahi-dev · 2022-01-31 06:45
marcanif you want something to investigate, I'm curious whether we can bypass cargo and just git submodule in the things we need, to make the whole system more self contained (one recursive clone like we already do, no extra network access or anything for builds)#asahi-dev · 2022-01-31 06:44
marcanadt isn't an attack surface; if the adt is compromised then the whole system is#asahi-dev · 2022-01-31 06:38
marcancatches up on a pile of email...#asahi-dev · 2022-01-31 06:37
marcanthings should calm down now though#asahi-dev · 2022-01-31 06:37
marcansorry for yet another week mostly offline, it's been a bit hectic in my life#asahi-dev · 2022-01-31 06:37
marcanalright, time to catch up with everything...#asahi-dev · 2022-01-24 16:12
marcanmaz: ack, thanks#asahi-dev · 2022-01-23 13:17
marcanbut I think Normal WB is likely to be fine for any ASC comms via carveouts#asahi-dev · 2022-01-23 13:16
marcanI imagine the GPU will have some way to blast out framebuffers as write-through so that doesn't happen there#asahi-dev · 2022-01-23 13:16
marcanthat's why m1n1 switched to uncached and a shadow copy#asahi-dev · 2022-01-23 13:16
marcanso on t6000 the memory controller throws a hissy fit if you end up with framebuffer cache lines in CPU cache (even though it still works)#asahi-dev · 2022-01-23 13:15
marcanthe framebuffer has to be mapped non-cacheable because while the SoC is cache-coherent, apparently that coherency does not come with realtime guarantees#asahi-dev · 2022-01-23 10:31
marcanI guess at the very least we could just say "outside the main RAM area" or something...#asahi-dev · 2022-01-23 10:31
marcanso we need something else#asahi-dev · 2022-01-23 10:30
marcanthe carveout-memory-map regions include all of RAM :-)#asahi-dev · 2022-01-22 17:31
marcanhow on earth are you two both missing things#asahi-dev · 2022-01-22 17:31
marcanalyssa: AVE is there#asahi-dev · 2022-01-22 17:30
marcan(clearly I haven't been corrupted by Japan enough yet)#asahi-dev · 2022-01-22 17:30
marcanmy brain just goes AV = Audio/Video#asahi-dev · 2022-01-22 17:30
marcanand yeah, I think Apple is probably more correct? since they don't do audio lol#asahi-dev · 2022-01-22 13:05
marcan| AVD | SoC | Audio/Video Decoder | No | Embedded in XNU | OS | OS | ~40 KB#asahi-dev · 2022-01-22 13:05
marcanit isn't?#asahi-dev · 2022-01-22 12:57
marcanalyssa: okay, a mythological ARM SystemReady machine#asahi-dev · 2022-01-22 12:40
marcanI'm basically trying to condense info about how these things work (to provide background) and then talk about how we think other OSes should interop#asahi-dev · 2022-01-22 12:39
marcanin case anyone wants to take a look :)#asahi-dev · 2022-01-22 12:39
marcangoing to get dinner, but this is what I've been typing up: https://mrcn.st/p/u89nqBhg (still incomplete, I just got to the interesting bit...)#asahi-dev · 2022-01-22 11:30
marcannot what I'd do for regular installs, but it would work for you#asahi-dev · 2022-01-22 11:29
marcanI could add some kind of knob to make the proxy itself time out#asahi-dev · 2022-01-22 11:07
marcanis 3800 words into a wiki one-stop article about how these machines work and what you need to know as a distro/bootloader/etc dev#asahi-dev · 2022-01-22 11:05
marcanmaz: that's what would happen with -v (once I implement it), it would wait a few seconds for USB enumeration#asahi-dev · 2022-01-22 05:19
marcanalright, I feel like documentation today, let me work on that#asahi-dev · 2022-01-22 05:04
marcanprobably easiest is to just throw a signature on it and have some "official" asahi signed m1n1 binaries or distros can build their own with their own key (or users can build their own etc) and the installer just installs whatever signed build is appropriate for $distro#asahi-dev · 2022-01-22 05:03
marcanwill have to come up with some secureboot scheme for it though. I've been thinking of making the next progress report also a call out to distros to talk about stuff like that...#asahi-dev · 2022-01-22 05:02
marcanso nvme will be useful for that#asahi-dev · 2022-01-22 05:02
marcanmostly so we can update devicetrees from distros