#asahi-dev · 2021-05-15 14:17
modwizcodemarcan: that's crazy amounts of travel to me wow#asahi-dev · 2021-04-29 15:49
modwizcodeI'm planning to pickup the patches again in the next month or two since I have time again.#asahi-dev · 2021-04-29 15:48
modwizcodeI just ended up going with the workaround for now.#asahi-dev · 2021-04-29 15:48
modwizcodeLike I implemented what I could. The way some of the peripherals operate (core local aliases) is fairly opposed to the model that qemu supports nicely so I got a bit stuck on how to make that code not be overly complex while not also using a workaround that's actually technically unsafe.#asahi-dev · 2021-04-29 15:47
modwizcodeThe interrupt controller should mostly work? I've not been keeping up with the project so I'm going to guess the latest linux patches don't work at all just as a pessimistic look.#asahi-dev · 2021-04-29 15:45
modwizcodeI have the patches on my github branch, they're in pretty rough shape but I do have interesting in cleaning them up and getting them upstream. I stopped having time just before adding USB so they're like I said pretty rough. You can interact through the serial console and there's some instructions on the Wiki for the Asahi project that I think are mostly up to date.#asahi-dev · 2021-04-29 15:43
modwizcodeillya: I do have some patches sorry I've been away from the project for awhile due to school but I'm back on task now#asahi-dev · 2021-03-08 17:44
modwizcodeheh#asahi-dev · 2021-03-08 17:42
modwizcodeah#asahi-dev · 2021-03-08 17:42
modwizcodeYou don't think full sleep is useful?#asahi-dev · 2021-03-08 17:42
modwizcodethat's what I thought, but I figured that iBoot could temporarily init the MMU then shut it down...#asahi-dev · 2021-03-08 17:41
modwizcodeyeah#asahi-dev · 2021-03-08 17:41
modwizcodehas to be physical right?#asahi-dev · 2021-03-08 17:41
modwizcodepossibly dumb question but is it set to the physical address or the virtual address?#asahi-dev · 2021-03-08 17:41
modwizcodefun!#asahi-dev · 2021-03-08 17:40
modwizcodewhat's the deal with #3 there? Where does the cpu0 reset vector point?#asahi-dev · 2021-03-08 17:39
modwizcodeyeah so I'm going to have to do something qemu side then to get what I want#asahi-dev · 2021-03-08 17:36
modwizcoderight#asahi-dev · 2021-03-08 17:27
modwizcodeYeah it's very minimal, and all the logic is currently in the python script, if it was in C I would have just included it into qemu with appropriate licensing#asahi-dev · 2021-03-08 17:25
modwizcodeI care about that because faking a concatenated image by dynamically loading the -kernel and -initrd qemu parameters after m1n1 allows hassle free boot straight into the kernel#asahi-dev · 2021-03-08 17:24
modwizcodeWhat I meant is that concatenated m1n1 + kernel does not work without loading it correctly#asahi-dev · 2021-03-08 17:23
modwizcodeer sorry#asahi-dev · 2021-03-08 17:23
modwizcodeit does not if you do not load m1n1 correctly#asahi-dev · 2021-03-08 17:15
modwizcodeMaybe you're just going to use your python script and flesh it out enough to load darwin?#asahi-dev · 2021-03-08 17:13
modwizcodemarcan: does making m1n1 boot darwin imply adding mach-o loading to m1n1? (I can never track down the early boot files in the xnu source dump so I don't quite remember how much it needs)#asahi-dev · 2021-03-04 16:39
modwizcodeSounds like some nice updates though, I miss having time to work on stuff#asahi-dev · 2021-03-04 16:39
modwizcodeI swear the goal of arm assembly is to create new positions to understand the manual itself.#asahi-dev · 2021-02-23 15:54
modwizcodethat's neat#asahi-dev · 2021-02-20 18:23
modwizcodesven: understandable, it doesn't seem important at the moment anyway.#asahi-dev · 2021-02-20 15:38
modwizcodeif you just search it up you can find some rather similar errors https://gist.github.com/modwizcode/493380e2406fd0cea18880614700af4b#asahi-dev · 2021-02-20 14:27
modwizcodeSo that leads me to think it's completely unrelated again, especially because even if it does match up it's only "Global register space 0" not anything else, which would seem useless to a guest or anything.#asahi-dev · 2021-02-20 14:20
modwizcodeSome of the zero registers also line up perfectly with "write only" registers. Although others... don't?#asahi-dev · 2021-02-20 14:20
modwizcodeso that would line up.#asahi-dev · 2021-02-20 14:19
modwizcode"implementation defined space" which at first I thought was a deadend but actually that's where ARM encourages you to provide a view of the TLB#asahi-dev · 2021-02-20 14:19
modwizcodeThe SMMU spec for ARM is difficult because it has a few different memory spaces not a single map, but the values that you see line up with "#asahi-dev · 2021-02-20 14:18
modwizcodeI thought I was making some progress on trying to determine how the SMMU thing works corellium has a blog post that is... almost helpful :p. I've shifted back to thinking it might be just the TLB we're looking at#asahi-dev · 2021-02-20 14:16
modwizcodesven: Earlier something called DirectHW was mentioned https://github.com/PureDarwin/DirectHW#asahi-dev · 2021-02-19 19:53
modwizcode:p I don't even own any macs#asahi-dev · 2021-02-19 19:46
modwizcodethanks#asahi-dev · 2021-02-19 19:41
modwizcodesven: Can I get a full dump from the SMMU range?#asahi-dev · 2021-02-19 19:33
modwizcodeMy first thought here though is it's some kind of GARt#asahi-dev · 2021-02-19 19:30
modwizcodeI think you're probably right it's not a cache, based on the fact that not every device has one#asahi-dev · 2021-02-19 19:29
modwizcodeSuspicious MMU#asahi-dev · 2021-02-19 19:29
modwizcodeThe S has so many possibilities :p#asahi-dev · 2021-02-19 19:27
modwizcodeIs it actually doing some kind of memory op or is it just more granular mapping? (If I'm understanding things right)#asahi-dev · 2021-02-19 19:26
modwizcodeInteresting#asahi-dev · 2021-02-19 19:21
modwizcodethat's exciting#asahi-dev · 2021-02-19 19:00
modwizcodeThere's a lot of versions, I think it's kind of like a v1#asahi-dev · 2021-02-19 18:52
modwizcodeI could be just seeing what I want to see#asahi-dev · 2021-02-19 18:52
modwizcodewhat ARM's SMMU calls context bank is possibly in the correct place but dynamic constants make it hard to tell