<hertz>
Hey, can anyone help me diagnose this problem? I'm dropping to the grub rescue prompt and trying to figure out how to fix this. Since all of this happens on the ESP, I should be able to fix from macOS since it doesn't involve repairing an ext4 partition.
<hertz>
The error seems to be: "error: no such device: 725a9541–****-****-****-**********." (stars added by me)
<hertz>
But why would my root device's UUID have changed?
<hertz>
And how can I learn the current UUID for the root device?
swaggie has joined #asahi
<hertz>
(I did one weird thing which may have caused this. I noticed that my EFI\BOOT\BOOTAARCH.EFI file had changed compared to when I first installed, so I restored the original BOOTAARCH.EFI file. I guess BOOTAARCH.EFI hardcodes a certain UUID? That must've been what caused this.)
<artun>
hi, I'm trying to run FEX but apperently there is some issue with jemalloc. it prints this: <jemalloc>: Unsupported system page size . What could cause this? I'm running jemalloc that shipped in 'extra' repo, is that how it supposed to be?
<chadmed>
artun: FEX does not work with 16k pages and likely never will work
<chadmed>
it builds its own version of jemalloc with hard coded 4k paages
<chadmed>
this is required because x86 only supports 4k pages, and linux cannot do per-process page size like xnu can
<mps>
hertz: maybe you rerun grub-install unintentionally
<artun>
chadmed: is there any other alternatives to FEX then? my initial goal was to run x86 apps in wine.
<mps>
hertz: grub-install embeds boot partition in BOOTAARCH.EFI
<hertz>
Thanks, jannau ! But i don't think the keyboard works on M2's in current version of m1n1/uboot/grub. I'll go hunt for a USB keyboard... that might work.
cylm_ has quit [Ping timeout: 480 seconds]
<mps>
hertz: about few days ago I had similar problem with my old m1 mbp, I break u-boot and run 'run distro_bootcmd' and it find proper partition and boots normaly
<hertz>
Sorry I think I changed some chars, I thought it was unique to my machine
<hertz>
It is the one you just posted
<marcan>
lol
<hertz>
:-)
<marcan>
trying too hard :p
<hertz>
lol
<hertz>
yeah
swaggie has quit [Ping timeout: 480 seconds]
<marcan>
as for the HDD size, because U-Boot is silly and lies to you
<marcan>
it's not 233 GB, it's 233 GiB
<hertz>
Oh, ok, got it
<marcan>
(I will forever die on the hill that using binary powers of two for UI by default is dumb and pointless, doubly so if you don't actually use the right IEC suffixes that indicate so)
<marcan>
(except for RAM sizes)
<hertz>
So, marcan, I booted into linux with jannau's. Can I just hand edit \EFI\BOOT\BOOTAARCH64.EFI to put the right UUID in or do I need to reinstall grub?
<marcan>
if you can chroot into your system you can just run update-grub
<hertz>
(By the way, I agree about using GB vs GiB, I hate it when people don't do that)
<hertz>
Ok
<marcan>
oh if you're already booted into the real system just run update-grub
gladiac has joined #asahi
rvalue has quit [Remote host closed the connection]
rvalue has joined #asahi
<ar>
(I will forever die on the hill that using base10 for storage devices is dumb, pointless, and only forced by hdd manufacturers to market bigger numbers to consumers)
<mps>
ar: yes
<hertz>
The world would be a better place if everyone just used GiB for base2 and GB for base10, consistently. Then HDD manufacturers could use the bigger base10 number and it would be fine.
<mps>
'good marketing in technical world' (sorry for OT)
qdot has quit [Ping timeout: 480 seconds]
<mps>
in all old books, articles and technical docs you will read that 1KB is 1024 bytes
<hertz>
Yeah, that is true, mps
artun has quit [Remote host closed the connection]
<hertz>
Advice about trackpad not working in linux? It gets recognized in dmesg, but doesn't actually function?
<hertz>
input: Apple MP multi-touch as /devices/platform/soc/24eb14000.fifo24eb30000.input/0019:05AC:0351.0002/input/input6
<hertz>
Maybe because it gets recognized as a Magic Mouse?
<mps>
I doubt, also in my machine it magicmouse '1.009313] magicmouse 001C:05AC:0343.0002: input,hidraw1: SPI HID v3.52 Mouse [Apple Internal Keyboard / Trackpad] on spi0.0 (2)'
<mps>
but it works fine
<mps>
hertz: is you machine M2?
<hertz>
Yes
<mps>
ah, then idk
<hertz>
And I'm using a slightly old root.img, from the 0.4.1 era I think
<hertz>
Is M2 trackpad just not upstreamed in this kernel yet? I can use an external mouse
<hertz>
(0.4.1 installer era, I mean. Like from over the summer. I do an offline install and haven't updated my scripts.)
<hertz>
Also, I have one last problem. On all my macs, both M1 and M2, I get these errors in dmesg about loading the wifi/bluetooth firmware: https://pastebin.com/y292WDsd
<hertz>
However, the wifi usually works anyway. On this machine (M2 Air, 2022), the wifi is not working at all. I can't figure out why.
<hertz>
(Both Wifi and BT are not working at all.)
<waldi>
hertz: did a new installation on a M2 Air, 2022 a few days ago and all of that works. so you are likely missing firmware (they are extracted somehow) and m1n1 update (current device tree)
<hertz>
What is "error -2"? I don't get why it fails to load the firmware. Is this about loading the driver in linux, or about loading actual firmware into the BRCM microcontroller's memory?
<waldi>
hertz: ENOENT, aka file not found. so you are missing firmware files in /lib/firmware
<hertz>
waldi: ok, thanks, I can try an online install with the latest version of asahi installer.
<hertz>
waldi: you're right, those files are missing from /lib/firmware/brcm. Tons of other files, but not the ones listed in the ENOENT error messages. So that would explain it.
<hertz>
It's weird, firmware extractor seemed to go ok. I do a semi-manual, offline install, so I'm familiar with the firmware extraction process.
<hertz>
I'll try an online install with the latest.
leo60228 has quit [Ping timeout: 480 seconds]
swaggie has joined #asahi
swaggie has quit [Ping timeout: 480 seconds]
MajorBiscuit has joined #asahi
SSJ_GZ has joined #asahi
psykose has quit [Remote host closed the connection]
psykose has joined #asahi
kh has joined #asahi
<kh>
> for tethered boot we recommend a kernel with modules disabled (all built-in)
<kh>
I assume it simply means that all modules should be built in and not CONFIG_MODULES=n?
<j`ey>
kh: with everything =y
<j`ey>
doesnt really matter if MODULES=n or not
<kh>
jannau: Also thanks for that link, for some reason I couldn't find it and ended up wasting ~3 hours fiddling with m1n1 the other day
<kh>
j`ey: It wouldn't be able to load wifi and bt though?
<j`ey>
kh: if the firmware is on the initramfs it should be fine
Dcow has joined #asahi
Dcow has quit [Read error: Connection reset by peer]
leo60228 has joined #asahi
<kh>
j`ey: How? But I guess CONFIG_FIRMWARE_IN_KERNEL will do the job too?
<j`ey>
thats the old name, it's CONFIG_EXTRA_FITMWARE, that works too
possiblemeatball has joined #asahi
possiblemeatball is now known as Guest335
Guest335 has quit []
<j`ey>
but I think initramfs works because its loaded first before other drivers
<corion>
Is there any way to make it so one must press the power button to start the laptop? I've noticed all it takes is pressing any random key and it will boot. I've accidently booted it multiply times, causing to just idle and slowly drain the battery over night.
swaggie has quit [Ping timeout: 480 seconds]
<corion>
s/multiply/multiple/
<corion>
Guessing it might be outside of asahi's control.
<mps>
corion: it is probably somewhere in apple firmware. I'm trying to find this for months but didn't found
<mps>
tried with nvram settings but no one option worked for me
<corion>
mps: Ah, that makes sense. Glad to hear someone has looked into it. :)
<corion>
Happy holidays!
eiln has joined #asahi
<eiln>
mps: nice
<hertz>
corion : nvram autoboot=false did not work?
eiln has quit []
kh has quit [Remote host closed the connection]
merry has joined #asahi
c10l has quit [Quit: Bye o/]
c10l has joined #asahi
bcrumb has joined #asahi
<mps>
hertz: this didn't worked for me
<hertz>
mps: yeah, i just tried also, and it does not work for me (M2 Air 2022)
bcrumb has quit []
<hertz>
mps: sorry, I don't know the answer.
<hertz>
mps: if we found an SMC key for it, is there a good tool to change smc values on apple silicon? (either linux or macos)
<mps>
hertz: if someone find it I guess someone will make tool
<hertz>
There isn't currently any tool to change SMC keys? I think there are tools on Intel Macs...
<hertz>
Random question: what is SFR? I'm not finding an answer in Google. Context is when the Asahi installer says: "SFR Version: 21.x.y.z..."
<j`ey>
hertz: system firmware recovery or something
<handlerug>
system fallback recovery I think
<hertz>
j`ey So it's just the original versions of the component firmwares?
<hertz>
Oh, is that the alternate recovery OS? The one with its own partition?
<j`ey>
yeah
<j`ey>
handlerug: that makes more sense
<hertz>
There is something called "fallback recovery" though. I read about it on a blog (searching...)
<handlerug>
the docs repo is ought to be pinned in the GitHub org
<handlerug>
(hard to find it)
<hertz>
Thanks handlerug ! I'd seen this doc before but lost track of it.
Zopolis4 has quit [Quit: Connection closed for inactivity]
bluetail87 has joined #asahi
sdomi has quit [Quit: meow, Quitting.]
sdomi has joined #asahi
sdomi has quit []
bluetail8 has quit [Ping timeout: 480 seconds]
bluetail87 is now known as bluetail8
sdomi has joined #asahi
swaggie has joined #asahi
swaggie has quit [Ping timeout: 480 seconds]
bcrumb has joined #asahi
fallenchromium has joined #asahi
possiblemeatball1 has quit [Quit: Leaving]
bluetail80 has joined #asahi
bluetail8 has quit [Ping timeout: 480 seconds]
bluetail80 is now known as bluetail8
bcrumb_ has joined #asahi
bcrumb has quit [Ping timeout: 480 seconds]
Guest173 has quit [Quit: G-line: User has been permanently banned from this network.]
vx has joined #asahi
vx is now known as Guest345
<fallenchromium>
Hi everyone! I wanted to ask if there's some way to set an upper charging threshold on Asahi Linux, akin to what can be done with "Al Dente" app or Smart Charging on macOS. I really want this laptop to last me long, so I'm a bit worried about battery longevity. AFAIK SMC is responsible for this, is there a node for that setting exposed under /dev by any chance?