marcan changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | Not ready for end users / self contained install yet. Soon. | General project discussion | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Topics: #asahi-dev #asahi-re #asahi-gpu #asahi-stream #asahi-offtopic | Keep things on topic | Logs: https://alx.sh/l/asahi
babilen has joined #asahi
al3xtjames has quit [Quit: Ping timeout (120 seconds)]
al3xtjames has joined #asahi
axboe_ has joined #asahi
axboe_ is now known as axboe
chadmed has quit [Ping timeout: 480 seconds]
PhilippvK has joined #asahi
phiologe has quit [Ping timeout: 480 seconds]
kov has quit [Quit: Coyote finally caught me]
kov has joined #asahi
axboe has quit [Quit: Lost terminal]
axboe has joined #asahi
<marcan>
zv: if this still happens after setting the macOS volume as default, this is a monitor or Apple problem; nothing we can do about that unfortunately
<marcan>
all that boot options screen and recovery mode stuff is Apple code
<marcan>
we just install the code, but there's no way for us to screw that up since it's signed, and once you make macOS the default none of the stuff we touched comes into play any more
hays has quit [Remote host closed the connection]
chadmed has joined #asahi
eroux has joined #asahi
<Glanzmann>
zv: For me it happened a few times. (mac mini going black after reboot from u-boot.bin kmutil thing) - I always power it off and on again and everything works.
<Glanzmann>
jannau: Can you please write your command again to verify the dtb under Linux, I lost it.
willemml has joined #asahi
<mixi>
Glanzmann: the command you're looking for should be "dtc -I dtb -O dts /sys/firmware/fdt"
<Glanzmann>
mps: m a r c a n s merge window is 15 minutes every 4 weeks, it is easy to miss. ;-)
<mps>
Glanzmann: thanks
tomtastic_ has quit [Ping timeout: 480 seconds]
<povik>
j`ey: it's easier because of linux internal reasons
<j`ey>
povik: ah
<kettenis>
seems the setup that apple uses with a separate digital amp for each speaker is somewhat unique
<povik>
there's more than one speaker on one bus everywhere but the mac mini, and that took me some figuring out how to handle that correctly with the ASoC infrastructure
<povik>
probably related to what kettenis says
<j`ey>
so the mac mini has speaker support?
<povik>
sure, had for a while
<povik>
that worked first
<j`ey>
oh cool, I thought it was always just the jack that worked
<povik>
seems easier to add my repo to the official instance. let's see if it can not spam mailing lists about stuff there
<dottedmag>
apple-admac.c warnings on m68k are very useful.
<sven>
i don't think anyone minds that spam fwiw
<sven>
dottedmag: they actually point to a missing #include fwiw
<povik>
yeah. and in the report j`ey linked earlier they had pointed out an actual bug i missed
<_jannau_>
povik: I think nobody reads lkml unfiltered
<povik>
still it's bit funny some computer out there is test-building admac on obscure platforms
<povik>
_jannau_: good, seems the cc can be turned off in the end
<j`ey>
povik: m168k
<povik>
hm?
<j`ey>
povik: m1 + m68k = m168k :P
<povik>
ah :D
<j`ey>
or m69k nice
<sven>
that’s what you get when you either add COMPILE_TEST or don’t depend on arm64 or ARCH_APPLE ;)
<immychan[m]>
Hi, for some reason after installing Asashi Linux but later deleting the partitions I can’t install it again, the installer won’t run and macOS recovery tells me that it can’t find any users to authenticate on the volume
<immychan[m]>
Any ideas why this might be happening?
<Glanzmann>
immychan[m]: I would reboot into macos and try it again.
<Glanzmann>
Maybe the problem is that you have as default boot device a device selected which no longer exists?
<immychan[m]>
Already tried that, didn’t seem to work :/
<Glanzmann>
immychan[m]: Can you paste the output of the asahi installer from macos?
<immychan[m]>
And I’ve tried reselecting my startup disk and it didn’t work
<immychan[m]>
I’ll do that in a moment
<Glanzmann>
So your default startup disk is macos.
<Glanzmann>
You booeted into macos, and run the installer from there?
<immychan[m]>
Essentially it just tells me to run the installer from my primary macOS install, which I am doing
<Glanzmann>
Maybe change the startup disk to the macos, reboot and try again.
<immychan[m]>
Did that too
<Glanzmann>
Otherwise I'm out of my depth, but I'm sure m a r c a n, can help.
<Glanzmann>
Can you paste the output of 'diskutil list'?
<immychan[m]>
I’ll do that in a second
eroux has joined #asahi
thevar1able has joined #asahi
<immychan[m]>
Weird, running an update seemed to sort this issue out
<immychan[m]>
Gonna try again to install Asashi
<j`ey>
marcan: one thing Im not clear on, will we ever be able to have the main macOS install on a separate harddrive? such that I can use ~all the internal SSD for linux?
<dottedmag>
j`ey: m68001
<j`ey>
dottedmag: hehe
<Glanzmann>
immychan[m]: I had also a few times. I needed a reboot. But could not reproduce.
<marcan>
j`ey: nothing stops you from deleting macOS if you want post facto, and I guess I should make the installer be able to recognize a external install for auth purposes... though I'm not sure if that works with Apple's own tooling to begin with, need to test
grgy has left #asahi [WeeChat 3.4]
<Glanzmann>
What is the hdmi overscan on the mini?
<Glanzmann>
And how is it a problem?
<VinDuv>
Glanzmann: Some TVs cut the edges of the image when receiving an HDMI signal
thevar1able has quit [Ping timeout: 480 seconds]
<Glanzmann>
VinDuv: I see, thank you for the explanation.
<j`ey>
marcan: I thought it was said that a macOS install was needed to update SFR or something. Just that 70G is quite a big chunk if you have the 256G model (which I do)
<marcan>
j`ey: yes, I mean if you have an install in an external HDD nothing stops you from deleting the internal one
<marcan>
I just haven't tested if all that works properly from an external one yet
<marcan>
there's also the idevicerestore way of updating SFR, though it's a big hammer and probably not a good idea to recommend
<marcan>
I think that one ends up requiring internet activation due to its nature
<marcan>
eventually it's entirely possible we'll figure out how to trigger a run the updater ramdisks (I think it's a ramdisk?) from linux
<j`ey>
cool, it's something youve at least thought about then, so im happy :D
<zv>
marcan: thanks--it the behavior is still the same when the boot volume is macOS. I'll try to grab a monitor from work to test. Not sure why it worked only once for me.
<as400[m]>
Is it possible to boot macos on these new machines from external disk ?
yuyichao has quit [Ping timeout: 480 seconds]
mlq has quit [Ping timeout: 480 seconds]
mlq has joined #asahi
yuyichao has joined #asahi
<marcan>
as400[m]: indirectly, yes
<marcan>
the kernel always boots from internal disk
<as400[m]>
marcan: So 1TR would need to stay on internal disk all the time, right ?
<as400[m]>
This last partition.
<marcan>
that is not 1TR, that was a misunderstanding that became invalid with 12.x
klaus has joined #asahi
<marcan>
that is called System Recovery and is part of system firmware and nobody should ever delete it
<as400[m]>
oh, ok. I'm confused with all these recoveries.
<marcan>
1TR is not a thing, it's a mode
<marcan>
(yes the name is kind of terrible)
<as400[m]>
marcan: especially when you come from Linux. For me it's one big mess
<as400[m]>
I mean macos recoveries
<marcan>
it just means "the user held down the power button once and a proper recovery was booted"
<marcan>
you mean when you come from x86 UEFI :p
<marcan>
this is system firmware
<as400[m]>
marcan: probably ... :
<kettenis>
hey at least having all your firmware in flash is FSF compatible ;)
<as400[m]>
kettenis: :)
<mps>
on some chromebooks u-boot can be put in flash
<kettenis>
when does flash stop being flash?
<kettenis>
many arm SBCs have eMMC flash
<as400[m]>
marcan: does System Recovery have the kernel to boot with ?
<marcan>
yes
<mps>
kettenis: yes, I mean eeprom
<as400[m]>
mps: yes, I had uboot in SPI on Pinebook Pro.
<mps>
as400[m]: I thought PineBook Pro is made this default
<mps>
'flash' term nowadays is ambiguous
<as400[m]>
mps: no, by default it has uboot on eMMC
<mps>
as400[m]: I'll try to remember (and as usual I will forget ;) )
<mps>
I don't have PineBook Pro
<as400[m]>
mps: actually this thing is such a crap... Nothing to talk about :)
<as400[m]>
marcan: One more question about installer. Are you thinking of some mechanism of updating the stub macos partition from installer ?
handlerug has quit [Remote host closed the connection]
tdmm has quit [Remote host closed the connection]
raiaq_ has quit [Remote host closed the connection]
jleightcap has quit [Remote host closed the connection]
WhyNotHugo has quit [Remote host closed the connection]
coder_kalyan has quit [Remote host closed the connection]
handlerug has joined #asahi
coder_kalyan has joined #asahi
WhyNotHugo has joined #asahi
raiaq has joined #asahi
jleightcap has joined #asahi
tdmm has joined #asahi
JamieMagee[m] has quit [Quit: Reconnecting]
JamieMagee[m] has joined #asahi
JamieMagee[m] is now known as JamieMagee
willemml has joined #asahi
tdmm has quit [Remote host closed the connection]
handlerug has quit [Remote host closed the connection]
jleightcap has quit [Remote host closed the connection]
coder_kalyan has quit [Remote host closed the connection]
raiaq has quit [Remote host closed the connection]
WhyNotHugo has quit [Remote host closed the connection]
coder_kalyan has joined #asahi
handlerug has joined #asahi
WhyNotHugo has joined #asahi
raiaq has joined #asahi
tdmm has joined #asahi
jleightcap has joined #asahi
willemml has quit [Quit: willemml]
phiologe has joined #asahi
willemml has joined #asahi
PhilippvK has quit [Ping timeout: 480 seconds]
PhilippvK has joined #asahi
willemml has quit [Read error: Connection reset by peer]
phiologe has quit [Ping timeout: 480 seconds]
<landscape15[m]>
as400: I asked the same question some time ago. It is surely possible, but I don’t know if you will at some point avoid to set again the boot object in 1TR
Misthios has quit [Quit: Misthios]
artemist has quit [Read error: Connection reset by peer]
Misthios has joined #asahi
Misthios has quit []
Misthios has joined #asahi
Misthios has quit []
Misthios has joined #asahi
___nick___ has joined #asahi
___nick___ has quit []
Misthios has quit []
___nick___ has joined #asahi
Misthios has joined #asahi
roxfan2 is now known as roxfan
<as400[m]>
landscape15 - maybe macos install on external ssd would be enough ?
<as400[m]>
That would be cool.
the_lanetly_052 has joined #asahi
mlq has quit [Quit: WeeChat 3.0]
<as400[m]>
In case of my mbp 14 upower reports battery capacity lower than it really is - 54.5 Wh vs 60Wh. Anyone with the same symptom ?
<tpw_rules>
i wonder what coconutBattery under macOS says
<as400[m]>
Well when I look at /sys/class/power-supply/macsmc-battery/ and so on there is 60 Wh reported.
<tpw_rules>
the SMC reports both a "design capacity" and "full charge capacity"
<as400[m]>
Yep - both over 60Wh
mlq has joined #asahi
<VinDuv>
as400[m]: which upower metric are you talking about? energy-full?
<Glanzmann>
From the last output you pasted the identifier of the stub partition was disk0s3 but that can change after you reboot.
<David[m]123456789>
I booted it several time, the first time there were 2 flashes of the Apple logo followed by the recovery proposing me to change startup disk (which I did) bc the custom kernel was broken or sthg like this ; the only thing differeront from your video was to install KDE
<Glanzmann>
Okay, try to reinstall the stub. But I have no explanation how this can happen.
<Glanzmann>
Because the boot object will be copied somewhere else.
<Glanzmann>
Maybe you somehow destroyed the stub parition in a way that it is broken. But in order to see that I need a booted Linux and the output of https://pbot.rmdir.de/kCRi-yR4nW4D59POkQiulA
<Glanzmann>
This is how it looks on my system.
<Glanzmann>
But this was reordered when I updated the stub partition at some point.
<Glanzmann>
So in the beginning the esp and Linux partition had a partition number behind 1tr.
<Glanzmann>
Actually the above paste was from another installation method (dd) but this one was installed using the Debian installer: https://pbot.rmdir.de/Ct8mut-xMSBpz2CVcVn95A
<Glanzmann>
As you can see by the size of the ESP partition.
<David[m]123456789>
I rerun into a bootloop reinstalling the stub
<Glanzmann>
That can't be true.
<Glanzmann>
I think you know how to get rid of this.
<Glanzmann>
Once you cleaned up, you should start over.
<Glanzmann>
If you want we can do 1-on-1 session to get it right.
mini has quit [Quit: ZNC closing...]
mini has joined #asahi
isaac has joined #asahi
CME has joined #asahi
isaac has quit [Quit: WeeChat 3.4]
___nick___ has quit [Ping timeout: 480 seconds]
eroux has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Glanzmann>
mixi: You we're right. I really screwed up with the device tree. What happened is that I copied the right u-boot with the right device tree to /boot/efi and had a script 'ush' which than downloaded the wrong u-boot and installed that one.
<Glanzmann>
mixi: Backlight works. Problem is by default is off, because I forogt the defaulton thingy. I'm adding this now.
sirn- has joined #asahi
<Glanzmann>
I can turn off and on the backlight of the air now by doing: echo 1 > /sys/class/backlight/backlight/bl_power ; sleep 1; echo 0 > /sys/class/backlight/backlight/bl_power
<Glanzmann>
0 menas on; 1 means off.
<kov>
Glanzmann, where can I find your dt for backlight?
<j`ey>
Glanzmann: is it the screen, or is it the backlight?
sirn has quit [singleton.oftc.net coherence.oftc.net]
sirn- is now known as sirn
<Glanzmann>
j`ey: I don't really know. But I would say it is the backlight.
<Glanzmann>
So what happened now I booted the right device tree and linux made the screen go black.
<j`ey>
Glanzmann: well is it still visible?
<Glanzmann>
But x started and logged me in.
<Glanzmann>
j`ey: No I can't see shit.
<Glanzmann>
But I'm in a dark room right now.
<j`ey>
ok so its the full screen :)
<j`ey>
oh, well turn some lights on!
<Glanzmann>
Rest of the family is sleeping in the same room, but I turned on my mobile flash light, and I still can't see shit.
<Glanzmann>
So what I did, I used a second machine to log in via ssh and poked at stuff in the /sys/class/backlight/ thingy until I found that bl_power it is.
<Glanzmann>
I'm fixing the patch, testing it and than pasting it.
<Glanzmann>
j`ey: Where should I put that patch to the arch/arm64/boot/dts/apple/t8103.dtsi?
<Glanzmann>
Because I mean for the mini, you don't want or can turn of the screen.
<Glanzmann>
kov: One second.
<j`ey>
Glanzmann: in j313 j293
<j`ey>
bit of duplication is fine
<Glanzmann>
Okay let me fix that up and than post the patch.
<j`ey>
cool! I can't test, havent had the chance to resinstall m1n1 after I reformatted
<Glanzmann>
j`ey: I test it right now and I bet mps tests it as soon as he gets a chance.
<Glanzmann>
Than it is tested.
<Glanzmann>
x-tested-by: Glanzmann on air.
<mps>
Glanzmann: what, I just come back to home
<mps>
ah, bl_power
<Glanzmann>
Yep.
<j`ey>
I wonder if the smc gpio is used to turn the display off for sleep, and the dcp brightness control is for everything else
<mps>
ok will do, but have to finish some work first
<Glanzmann>
Sure, let me know if it works for me. I just tested it, it works on the air.
<kettenis>
so toggling gP12 saves about 1W of power
<Glanzmann>
j`ey: No idea, but it was really getting on my nerves that I could not turn off the display. I always shutdown my systems when I was not using it.
<Glanzmann>
kettenis: I see, how did you measure it?
<kettenis>
that's what the SMC power sensor reports
<j`ey>
Linux doesn't have that hooked up yet
<kettenis>
power consumption when idle drops from ~3.8W to ~2.8W
<kettenis>
Battery state: high, 99% remaining, 1355 minutes life estimate
<Glanzmann>
Wow.
<Glanzmann>
Is this on the air?
<kettenis>
13" mpb
<Glanzmann>
I see.
<kettenis>
j293
<Glanzmann>
Do you have it already hooked up to something in OpenBSD?
<kettenis>
no, I hacked the SMC driver for this
<Glanzmann>
I see.
<kettenis>
but I guess I'll write gpiobl(4) soon ;)
<Glanzmann>
hrhr.
<Glanzmann>
So when I press the power button on my air for a short time. I get in dmesg/syslog a key release event: [ 432.110637] macsmc-rtkit 23e400000.smc: RTKit: syslog message: apComms.cpp:373: SMC HID Event: 01 06 00
<Glanzmann>
Does someone know how I can hook that up to a userland script?
<Glanzmann>
Something i know which works out of the box is using syslog-ng and trigger a script everytime it sees that message, but I assume there is some event thingy I can subscribe to?
<kettenis>
not sure if we've figured out how that event gets reported "properly" yet
<kettenis>
you really don't want to parse the syslog output for that
<Glanzmann>
I also started a patch that does that for the kernel. But I did not know how to hook that up to smc, but it is ugly: https://tg.st/u/wip-displayblanking.patch
<j`ey>
fine for some hacks!
<Glanzmann>
I mean at some point, someone will do it properly. But until that. I like that.
<j`ey>
apple_smc_event_received in drivers/platform/apple/smc_core.c is a good place to start looking
<Glanzmann>
Thanks.
<kettenis>
so the lid is hooked up to gP01
<kettenis>
looks like you could try hooking that up using gpio-keys-polled
<Glanzmann>
kettenis: So gpio-keys-polled would poll gP01 and send a key event and than I could use my window manager to do something when that key event is received?
<kettenis>
look at arch/arm/boot/dts/imx6q-novena.dts
<zv>
update: it was indeed the monitor. asus vs278q-p does not like mac mini m1. a different dell monitor works fine.
<Glanzmann>
zv: Good.
<Glanzmann>
kettenis: I see.
<Glanzmann>
Btw. I just pressed the sleep button (F6) on air with debian. The system was freezing.
<Glanzmann>
Screen turned off than it came back probably because i pressed it again and than freezed.
<Glanzmann>
After booting up again the system was hanging on boot with some nvme messages, I think.
<Glanzmann>
than I powerecycled it again and it came up okay
<Glanzmann>
I can do a video.
<Glanzmann>
Maybe i should disable sleep.
<Glanzmann>
Debian folks: That disables any kind of sleep, I tested it: sudo systemctl mask sleep.target suspend.target hibernate.target hybrid-sleep.target
<j`ey>
Glanzmann: if youre still interested in looking: drivers/power/supply/macsmc_power.c apple_smc_register_notifier(power->smc, &power->nb);
<j`ey>
so this driver gets called, when an SMC notification happens. looks like all registered handlers would be called and its up to the callback to figure out if it needs to do something