ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | Non-development talk: #asahi | General development | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-dev
<mischa85>
24D60 kernel stops at 0x21a7eb0/0x6e3c000
zumi_ has joined #asahi-dev
zumi has quit [Remote host closed the connection]
zumi_ is now known as zumi
nesnas has quit [Ping timeout: 480 seconds]
<mischa85>
m1n1 is not responding anymore fwiw
<mischa85>
could this be related to mmc?
<mischa85>
*mcc
nesnas has joined #asahi-dev
<fl0_id>
mischa85 btw, didn't think about it at the time, but does the adt change between firmware versions? as the one I sent you was still for sonoma as I'm stlll using that. Not sure if I mentioned that at the time...
<mischa85>
fl0_id: not really. it's mainly useful to see where m3max and m4max are similar and give me some hints
<fl0_id>
mischa85 ok good
<fl0_id>
just wanted to make sure I didn't accidentally sabotage you
<mischa85>
t6041 seems more similar to t6031 than to t8132 in many ways
<mischa85>
no problem, it's fine
<fl0_id>
mischa85 yeah based on the code numbers that makes sense I guess. Like also on m3, m3 has an 8x number, pro and max have a 6x. and the pro and max always need extra things
<mischa85>
i still have a tz_regs section which i'm unsure about in mcc. should be translatable from region-id-2 and region-id-4 but nobody told anyone *how* they translate
nesnas has quit [Ping timeout: 480 seconds]
<mischa85>
so i might very well have those wrong
<mischa85>
fl0_id: yeah it's going to be a bit of a mix here and there. i have a branch from yu-re-ka which got some info on t8132, now with t6031 i'm trying to fill in the gaps for t6041
<mischa85>
at least enough to get hv running which might supply some missing values and chickens
<fl0_id>
mischa85 ah yeah btw their nick on irc is yuka probably
<fl0_id>
(or so I assume)
<mischa85>
yeah i noticed on my irc searches
<fl0_id>
yeah makes sense
<fl0_id>
ah ok :)
<mischa85>
So dberlin wrote this: Note: For unhandled machines, the TZ regions can be found (on m1, m2, m3) by looking at region-id-2 and region-id-4 on a booted macos, in the /chosen/carveout-memory-map DT node. This can be used along with dumping the mcc reg space to find the correct start/end/enable above.
<mischa85>
but this still doesn't explain how two big hex values translate to tz.count/stride/start/end/enable
nesnas has joined #asahi-dev
<mischa85>
anyway, disabling mmc_init() doesn't make a difference
<mischa85>
so worries for later i guess
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
pb17 has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
pb17 has joined #asahi-dev
nesnas has joined #asahi-dev
dylantaylor has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
greguu has joined #asahi-dev
nesnas has joined #asahi-dev
dylantaylor has quit [Quit: Konversation terminated!]
rrendec has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
ipatch has quit [Read error: Connection reset by peer]
ipatch has joined #asahi-dev
pb17 has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
Extern has joined #asahi-dev
tobhe_ has joined #asahi-dev
nesnas has joined #asahi-dev
tobhe has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
nora_ has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
nora has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
Extern has quit [Quit: WeeChat 4.1.1]
nesnas has joined #asahi-dev
<chadmed>
lesgoo figured out the silly alsa client thing
pb17 has joined #asahi-dev
mischa85 has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
mischa85 has joined #asahi-dev
nesnas has quit [Remote host closed the connection]
nesnas has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
zerozerozero has joined #asahi-dev
mischa85 has joined #asahi-dev
nesnas has joined #asahi-dev
MountainRose has quit [Remote host closed the connection]
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
JayBeeFOSS has quit [Ping timeout: 480 seconds]
JayBeeFOSS has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
mischa85 has joined #asahi-dev
nesnas has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
pb17 has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
mischa85 has joined #asahi-dev
zumi_ has joined #asahi-dev
zumi has quit [Remote host closed the connection]
zumi_ is now known as zumi
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
mischa85 has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
mischa85 has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
nimprod3l has joined #asahi-dev
johey has quit [Read error: Connection reset by peer]
johey has joined #asahi-dev
<nickchan>
rebased my a7-a11 pmu series above the feat_pmuv3 support, will send when the merge window closes
pb17 has joined #asahi-dev
mischa85 has joined #asahi-dev
nimprod3l has quit [Ping timeout: 480 seconds]
nimprod3l has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
<sven>
does that one also require changes to the dt? if yes please directly add those to the series
nesnas has joined #asahi-dev
<nickchan>
sven: I will for that version, thanks for the heads up!
<nickchan>
next version
<sven>
makes testing and picking them up easier :)
aditya has joined #asahi-dev
pb17 has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
mischa85 has joined #asahi-dev
nesnas has quit [Remote host closed the connection]
nesnas has joined #asahi-dev
mischa85 has quit [Ping timeout: 480 seconds]
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
pb17 has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
mischa85 has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
nesnas has joined #asahi-dev
nesnas has quit [Ping timeout: 480 seconds]
<mischa85>
alright, i now tried with a way smaller release kernel. it stops at 0x5d2360/0x3ed4000, does show the dots this time. but throws another serialexception
<mischa85>
this time in total early bringup m1n1 mode at 115200 baud
pb17 has quit [Ping timeout: 480 seconds]
<sven>
iirc there are some control chars that will be caught by the cs software. have you tried disabling them?
<mischa85>
i was thinking the same thing, but 0x1f doesn't seem to be involved. but going to disable them anyway just to be sure
nesnas has joined #asahi-dev
<sven>
it's pretty likely 0x1f will be somewhere in the xnu binary and mess things up
<mischa85>
yes but the character gets passed, it requires a combination with another specific char to do anything
<mischa85>
but lemme try without any handling at all
nesnas has quit [Ping timeout: 480 seconds]
<mischa85>
you might have been right. steady supply of dots now.......
<LarstiQ>
chadmed: what was the silly alsa client thing?
<mischa85>
i'm not going to touch anything and make myself coffee
<LarstiQ>
mischa85: smakelijk :)
<sven>
nice :)
<mischa85>
stuff is happening
<mischa85>
LarstiQ: :-)
<mischa85>
woah it almost booted. the python script tried setting some cpu rvbars which of course SError'd. lemme remove those too
nesnas has joined #asahi-dev
<mischa85>
it seems the ecores are actually powered on, the pcores cause SErrors
<mischa85>
i'll just disable cpustart_wh in the script for now
<sven>
iirc the chickens are some of the very first things xnu does anyway
<LarstiQ>
sven: the way you phrased that makes me think xnu plays Stardew Valley ;)
pb17 has joined #asahi-dev
zumi_ has joined #asahi-dev
zumi has quit [Remote host closed the connection]
zumi_ is now known as zumi
Mary has quit [Quit: .]
Mary has joined #asahi-dev
<fl0_id>
maybe it does. needs to get its protein in
<sven>
lol
<chadmed>
LarstiQ: wireplumber persisted the hidden alsa node as the default because the filter chain was not of a higher session priority
<chadmed>
thats why folks were getting "quiet" audio in alsa clients like audacity, and also why we had to vendor the dsp script for a little while
<chadmed>
ive fixed this in main and will tag releases tomorrow
aditya has quit [Quit: Connection closed for inactivity]
<nickchan>
mischa85: have you tried turning on the cpu power domains?
pb17 has quit [Ping timeout: 480 seconds]
<nickchan>
they are on at iboot -> kernel handoff on previous chips
Extern has joined #asahi-dev
pb17 has joined #asahi-dev
<mischa85>
nickchan: i cannot do that without knowing the ps-regs right?
<chaos_princess>
something went wrong with mmu, but idk what
rrendec has joined #asahi-dev
<mischa85>
ok, lemme try disabling mmu totally again. which means i have to upload the kernel at 115200 baud
<sven>
broken MMU can happen with incorrect chicken bits unfortunately
<mischa85>
yeah, with a bit of luck this is going to supply us with the correct chicken bits, so we can enable mmu and be back at 1.5 mbaud again for further development
<mischa85>
i'm trying with the release kernel at the moment as it's way smaller than the development kernel
<mischa85>
i had to take out page tables update, now it crashes after jumping to entrypoint. nothing from guest, SYNC exception. ESR: 0x96000035 (data abort (current))
<mischa85>
Right after: Jumping to entrypoint at 0x100199f8000
<chaos_princess>
what is the instruction at elr?
<mischa85>
that only gets printed at guest exception
<chaos_princess>
could it be that you are jumping to a no-exec page then?
<mischa85>
my bad
pb17 has quit [Ping timeout: 480 seconds]
<mischa85>
well, seeing how much i had to cut from the process
<mischa85>
i don't know
zerozero1 has joined #asahi-dev
zerozero2 has joined #asahi-dev
zerozerozero has quit [Ping timeout: 480 seconds]
zerozero1 has quit [Ping timeout: 480 seconds]
cow123 has quit [Ping timeout: 480 seconds]
zerozero2 has quit [Quit: WeeChat 4.5.2]
zerozerozero has joined #asahi-dev
<mischa85>
how can i find out?
<mischa85>
the pt_update function doesn't crash instantly, i can try to let it run partially
pb17 has joined #asahi-dev
<mischa85>
pt_update causes an Exception:SYNC on HW:RAM-HIGH
<mischa85>
which is indeed the range it tries to jump to: PT[10016428000:12000000000] -> HW:RAM-HIGH
okb has joined #asahi-dev
<mischa85>
my best guess is that mcc_init() has something to do with it
johey has quit [Ping timeout: 480 seconds]
sting has quit [Quit: doei]
<mischa85>
nope. so enabling mmu_init makes pt_update pass without crashing, but crashes later as a guest exception
<mischa85>
is it even possible to launch hv without mmu?
<chaos_princess>
from what i remember from reading arm64 spec, you can disable translation in el2
sting has joined #asahi-dev
john-cabaj has joined #asahi-dev
<mischa85>
HCR_VM?
johey has joined #asahi-dev
zumi_ has joined #asahi-dev
zumi has quit [Remote host closed the connection]
zumi_ is now known as zumi
john-cabaj has quit [Ping timeout: 480 seconds]
pb17 has quit [Ping timeout: 480 seconds]
bjoto has quit [Remote host closed the connection]
zeezie01 has joined #asahi-dev
<mischa85>
taking HCR_VM); // Enable stage 2 translation out doesn't make a difference sadly
<mischa85>
i'm out of ideas for today
pb17 has joined #asahi-dev
yuyichao_ has quit [Ping timeout: 480 seconds]
Extern has quit [Quit: WeeChat 4.1.1]
<mischa85>
interestingly, i also don't get a rainbow apple logo. it just crashes before
<mischa85>
something something paging related, but what...
HitokiriKincho has joined #asahi-dev
pb17 has quit [Ping timeout: 480 seconds]
nimprod3l has quit [Quit: Leaving]
pb17 has joined #asahi-dev
zerozerozero has quit [Quit: WeeChat 4.5.2]
HitokiriKincho has quit [Quit: Leaving...]
<mischa85>
chaos_princess: translation should be disabled when mmu_unit() is not called. the hv_translate() function actually checks for the SCTLR_M bit (mmu) in the SCTLR_EL12 (SCTLR_2?) register and acts accordingly
<mischa85>
and that bit should be 0 according to arm64 specs
<mischa85>
i should, however, take a closer look at what happens to HCR_EL2
<mischa85>
it's going to be painful tho transferring the kernel at 115200 bauds everytime. if anyone has a trick to get 1.5 mbaud without m1n1 denying it without mmu i'd be glad to hear of it