#asahi-dev · 2025-02-09 18:22
chaos_princessyea, that does sound suspicious#asahi-dev · 2025-02-09 18:22
chaos_princessah#asahi-dev · 2025-02-09 18:20
chaos_princessgive me like 5 mins, and will do#asahi-dev · 2025-02-09 17:38
chaos_princessexcept, well, i have tried to connect that port, and the display failed to lock on the signal until it finished booting#asahi-dev · 2025-02-09 17:37
chaos_princessi _think_ it was trying to show plymouth on an unconnected hdmi port#asahi-dev · 2025-02-09 17:37
chaos_princesshttps://pastebin.com/Emh5hGWg#asahi-dev · 2025-02-09 17:32
chaos_princesssec#asahi-dev · 2025-02-09 17:32
chaos_princessnot really?#asahi-dev · 2025-02-09 17:28
chaos_princesshv/no-hv is irrelevant, it also initializes if i enter the password blindly#asahi-dev · 2025-02-09 17:22
chaos_princessjannau: j314, booted it, dcp fails to update in plymouth/initrd (keeps u-boot messages), after entering disk password via m1n1 shell, dcp initializes and works#asahi-dev · 2025-02-09 11:38
chaos_princessah, yea, aic1#asahi-dev · 2025-02-09 11:38
chaos_princessi don't think so, iirc m4 also has the same thing where the boot core is not 0#asahi-dev · 2025-02-09 11:04
chaos_princessthis is most likely a wrong opinion, but i like the home-button-less ones more, app switching with gestures feels more natural/speedy (i was helping someone with their iphone 8 recently, and wow, do not want to go back)#asahi-dev · 2025-02-09 11:01
chaos_princessi think at least some generations of SEs can do this exploit#asahi-dev · 2025-02-09 10:59
chaos_princess(and each time i see that kind of activity i am sad that my iphone x died)#asahi-dev · 2025-02-09 10:59
chaos_princesscertain old devices have a bootrom exploit (checkra1n), and can load arbitrary os-es#asahi-dev · 2025-02-08 10:39
chaos_princessmakes sense then#asahi-dev · 2025-02-08 10:36
chaos_princesswhy adt btw, it is not an untrusted input, no?#asahi-dev · 2025-02-08 10:36
chaos_princessthe best part is when you supposedly have an independent "test" environment, and after you push a release to there, you somehow break prod :P#asahi-dev · 2025-02-07 17:42
chaos_princessthen it should be much less annoying#asahi-dev · 2025-02-07 17:35
chaos_princessthough if you have hw uart, you can chainload from there, but you need to first progress far enough that m1n1 can speak to your dev machine over uart#asahi-dev · 2025-02-07 17:33
chaos_princessfor early init - you will have to go via recovery every time#asahi-dev · 2025-02-07 17:33
chaos_princessonce you have usb working, you will be able to chainload. #asahi-dev · 2025-02-07 10:55
chaos_princessand it does slice::from_raw_parts or something internally#asahi-dev · 2025-02-07 10:54
chaos_princessmore fn get_value() -> &'static [u8]#asahi-dev · 2025-02-07 10:53
chaos_princesssince we do not rewrite the adt, probably &[u8], with probably even the static lifetime?#asahi-dev · 2025-02-07 10:49
chaos_princesswhat'cha planning?#asahi-dev · 2025-02-07 10:40
chaos_princessno, we have an allocator, https://github.com/AsahiLinux/m1n1/blob/main/rust/src/chainload.rs#L8#asahi-dev · 2025-02-06 14:01
chaos_princessjannau: iirc they were interested in development#asahi-dev · 2025-02-06 14:00
chaos_princessjust have backups, no matter whether you want to experiment with one or not#asahi-dev · 2025-02-06 14:00
chaos_princessget backups. #asahi-dev · 2025-02-06 13:58
chaos_princessthe use a non-mac computer to run m1n1 proxy client#asahi-dev · 2025-02-06 13:58
chaos_princessno#asahi-dev · 2025-02-06 13:57
chaos_princessbut yes, m1n1 requires two machines, and no, you can't use vms on one#asahi-dev · 2025-02-06 13:57
chaos_princessthe second machine does not need to be a mac#asahi-dev · 2025-02-06 13:31
chaos_princessfor use over virtual serial (the way you should do it), the other machine can be on any os, though linux or macos are preferred#asahi-dev · 2025-02-06 12:58
chaos_princessk#asahi-dev · 2025-02-06 12:58
chaos_princesssame inbox as in your sign-off?#asahi-dev · 2025-02-06 12:45
chaos_princessjannau: do you want to be cc-d as reviewer on adpdrm? i'm filling out the separate maintainers entry#asahi-dev · 2025-02-05 22:05
chaos_princessdone#asahi-dev · 2025-02-05 11:17
chaos_princessfrom what i got, it is mostly for new components, like usb support, secure boot, etc#asahi-dev · 2025-02-05 08:26
chaos_princessyea, i noticed the d-bus thing#asahi-dev · 2025-02-05 08:23
chaos_princesswill do this evening#asahi-dev · 2025-02-05 08:23
chaos_princessah, i see#asahi-dev · 2025-02-05 08:22
chaos_princessdoes it need a new release? i thought the only change was adding the -d switch to bless#asahi-dev · 2025-02-03 15:17
chaos_princesshttps://github.com/AsahiLinux/linux/blob/asahi/drivers/gpu/drm/apple/iomfb_template.h - ts1 and ts2 are timestamps, unk_10 also contains at least two timestamps, and they are set by zero by getting the memory from kzalloc and never writing anything there.#asahi-dev · 2025-02-03 15:05
chaos_princessno, we just set all timestamps to zero, and it works for non-vrr modes, but causes the vrr mode to only refresh at 60 hz. The suspicion is that setting timestamps correctly will make it go to 120#asahi-dev · 2025-02-03 14:53
chaos_princessnot at all, you need them to know when to present the frame.#asahi-dev · 2025-02-03 14:47
chaos_princesshttps://github.com/AsahiLinux/m1n1/blob/main/proxyclient/hv/trace_dcp.py - this is the dcp call tracer, and try looking into what the timestamps do in the dcp swap message, there is a suspicion that they are responsible for it somehow#asahi-dev · 2025-02-03 14:44
chaos_princessasahi only does 60hz on those, since the 120hz mode hard-requires vrr to work, and that is not supported