marcan changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | https://asahilinux.org/2022/03/asahi-linux-alpha-release/ | 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
<LuigyLeon[m]> any hints for debugging this? or should I just try nixos-install again from usb and partitions mounted?
<tpw_rules> did you run out of space on the ESP?
<LuigyLeon[m]> probably
<LuigyLeon[m]> let me check
<megalodon42[m]> I am trying to use bless --setBoot --device /dev/diskXX but it rejects all the disk numbers I am trying :(
<tpw_rules> i think you give it the preboot partition in the apfs?
<hellosway[m]> So does the asahi installer just 1. Download the tarballs from the servers. 2. make a fat ESP and system partition 3. get firmware from the recoveryos 4. move firmware to the ESP 4. untar the os tarball into the root partition 5. bless the ESP 6. install m1n1
<tpw_rules> ok ,looks like you've got plenty of space. you could try nixos-rebuild boot --install-bootloader
<j`ey> hellosway[m]: yes
yzz has joined #asahi
<tpw_rules> i think the root partition is a disk image rather than an untar
<hellosway[m]> j`ey: how exactly is m1n1 installed
<j`ey> with kmutil
<megalodon42[m]> I did bless --setBoot --device /dev/disk2s3, it asks for Local owner username and then password. Then it replies: Bootability failed error Domain=ByServiceErrorDomain Code-602 "(null)"
<LuigyLeon[m]> tpw_rules last message for me? rebooting brb
<tpw_rules> yes
<LuigyLeon[m]> oh easy to fix? xD I must have done something wrong during the stub upgrade
<tpw_rules> what do you mean?
<tpw_rules> mine does that too, it only shows two entries and you have to scroll for the rest
<tpw_rules> not sure why tbh
<tpw_rules> but it does list all of them for me...?
<LuigyLeon[m]> mine got stuck at 31 even though it should be at 41 by now
<tpw_rules> so that install-bootloader flag didn't fix it?
<LuigyLeon[m]> tpw_rules nope, that didn't change the original behavior
<hellosway[m]> Isnt the installer a bit overcomplicated for runnning a bunch of file writes and calling some programs.
<j`ey> if you can simlpify it, go ahead!
<tpw_rules> LuigyLeon[m]: hm. i don't think i've got that many generations before. let me try to replicate it in a bit. my m1 is not my daily driver
<tpw_rules> can you manually delete some older generations? i wonder if there's too many conf files or something
<hellosway[m]> I could turn it into a very simple linear shell/pythion script but first I have to figure it out compleetly
yzz has quit [Ping timeout: 480 seconds]
<j`ey> that would be hell to maintain?
<tpw_rules> hellosway[m]: notice there's a whole lot of weird nonsense necessary to interact with diskutil and the firwmare
<j`ey> also it streams the IPSW file, by using HTTP range requests
<LuigyLeon[m]> tpw_rules that's the surprising part as I don't have that many anymore, but somehow systemd-boot still shows me 1 - 31
<hellosway[m]> tpw_rules: what confuses me is why writing to the ESP takes almost 50 lines of code, most of with is redundant becuase the os information already verified before
<j`ey> hellosway[m]: which bit?
<tpw_rules> LuigyLeon[m]: oh... i notice systemd-boot has that "warning: boot loader reports a different ESP UUID than detected" do you have multiple of those?
<tpw_rules> so you do have two linux OSes
<tpw_rules> actually, that should be 8 partitions. what does lsblk say
<hellosway[m]> j`ey: its not really anything sspecificaly, just a line of code randojmly scatered
<j`ey> hellosway[m]: I dont agree really.
<tpw_rules> maybe /dev/nvme0n1p4 should be mounted at /boot .
<j`ey> hellosway[m]: I only see a few places related to esp/efi
<tpw_rules> what is in /etc/nixos/hardware-configuration.nix? and `cat /proc/device-tree/chosen/asahi,efi-system-partition` or so
<Lucy[m]> <tpw_rules> "choose a UEFI environment..." <- aw, I have to do it manually?
<tpw_rules> manually with regard to what
<tpw_rules> oh
<tpw_rules> yes
<tpw_rules> i suppose you could download the TAR and customize the installer-data.json
<kov> hellosway[m], doing the stuff in a particular way is simple, most of us who are here from the start were doing it completely manually before the installer
yzz has joined #asahi
<kov> hellosway[m], doing it in a way that is safe, handles several different configurations and corner cases, and is relatively flexible is not as simple
<tpw_rules> LuigyLeon[m]: what is in /dev/disk/by-uuid/ ?
<tpw_rules> ok, so your hardware-configuration.nix is wrong
<tpw_rules> it's mounting /dev/nvme0n1p5 as /boot but u-boot is reading /dev/nvme0n1p4
<j`ey> hellosway[m]: https://github.com/AsahiLinux/asahi-installer/blob/e9803d34be2d7c927d9a0717ec89e4a32dc90603/src/osinstall.py is the only file that deals with efi, apart from main.py that just prints it out
<tpw_rules> so you need to change the UUID in there to 97DE-08ED
<tpw_rules> and probably delete /nvme0n1p5
<tpw_rules> did you rerun the installer, not delete the old ESP, then not rerun `nixos-generate-config`?
<j`ey> hellosway[m]: and its about 6 lines of code in there
<hellosway[m]> kov: I would verify all information in a function at the start of execution, then get all data. Then the user would be asked everything, after this point there should be no questions to the user. But this might end up witha lot of copy pasted code
kjm99 has joined #asahi
<tpw_rules> didn't you just say you would make it have less code?
<j`ey> lol
<kov> hellosway[m], we'll be happy to be proven wrong, I can tell you =) if you're confident you can do better, why not do it?
<tpw_rules> use pathlib while you're at it, it's good™
<j`ey> I suppose you could do that, but Im not sure how it would help with the complexity
<LuigyLeon[m]> tpw_rules: I must have done something like that, I think I skipped both of those steps x.x. Rebuilding and rebooting after the hardware-configuration
<j`ey> still need to handle all the different cases
chadmed has joined #asahi
ngnboone has joined #asahi
THX1984 has quit [Ping timeout: 480 seconds]
yzz has quit [Ping timeout: 480 seconds]
<brentr123[m]> hopefully rosetta can somehow be implemented into asahi
<hellosway[m]> If qemu gets working with good performance, then you can use user space emulation with binutils for it
<brentr123[m]> hellosway[m]: I tried to use box64 wasn't working so hopefully that would be lit
kjm99 has quit [Ping timeout: 480 seconds]
<brentr123[m]> j`ey: i belive it says on aashi website that fex doesnt work cuz 16 pages
ryan_nupp[m] has joined #asahi
<chadmed> none of that will work until sven's patches for 4K pages on top of the IOMMUs get merged
<chadmed> nowhere near ready for use in user machines
<ngnboone> hey all - just heard about asahi through mental outlaw's youtube video
<ryan_nupp[m]> hey I just watched that video as well ngnboone
<chadmed> how soy and cringe and bloated and glowing did he say we are :P
<ngnboone> he said y'all are the real heros of our generation
<ngnboone> I was wondering what sort of progress/knowledge gathering has been made as far as the neural engine
<kode54> Glanzmann: what did you mean upgrade my stub?
<chadmed> well we know what it does and where it lives in MMIO but we cant really do anything with it yet
<kode54> I fresh installed for the first time with the script a few days ago
<kode54> then I wanted to wipe my macOS install
<chadmed> its not that high up on anyones priority list tbh, somewhere down between bluetooth and a vulkan driver
<kode54> macOS refused to create the user for my account because the stub existed
<kode54> so I wiped my machine, which deleted the stub, but kept the EFI partition and the Linux partition
<kode54> so now I have macOS 12.3 again, but no stub, just EFI and Linux
<tpw_rules> how did you wipe it?
<kode54> using Recovery, resetpassword, menu, Erase
<ryan_nupp[m]> just curious, is it possible to install asahi on an external ssd?
yzz has joined #asahi
<kode54> no
<tpw_rules> oh, that's a little weird
<kode54> I also had to delete my machine from iCloud like three times
<kode54> each time, it successively dropped a version number, keeping the same machine
<kode54> it went through 12.3, 12.1, 11.4
<tpw_rules> ryan_nupp[m]: you have to have at least 3GB on the internal ssd, but you could put the rootfs on an external ssd. that's of somewhat limited utility at this point because the maximum access speed you'd get is USB 2
<ngnboone> I can respect that chadmed... I guess it doesn't really come into play in the course of normal os operation
<chadmed> yeah priorities right now are stabilising what we already have and focusing on stuff that will improve daily driving (alyssa's mesa driver for example)
<kode54> I assume the bluetooth isn't working because all new Broadcom chip that requires new firmware blob and specific support in the Broadcom bt driver
<chadmed> it's not just the firmware
<kode54> I said driver too
ReaganMcFarland[m] has joined #asahi
<chadmed> well kinda
<kode54> device tree too?
<chadmed> the main problem is that broadcom spun out an entirely new protocol for it
<kode54> oh crud
<chadmed> like its not UART, its not USB
<kode54> so it will need an all new driver
<ReaganMcFarland[m]> Hey guys! Been following the project for a while now - just ran into a bad error during install
<ryan_nupp[m]> tpw_rules: Oh ok. What causes the USB 2 speed limitations?
<chadmed> its some weird UART-over-USB-over-PCIe crap
<ReaganMcFarland[m]> Is there a support channel?
<matthewayers[m]> We might need one in the future, but for now, what’s your question?
<j`ey> ReaganMcFarland[m]: no usb3 driver yet
<matthewayers[m]> marcan: see above ^^
<tpw_rules> ryan_nupp[m]: the USB3 and thunderbolt stuff just hasn't been fully reverse engineered and developed and made ready for end users yet. it will get there eventually. usb3 should be pretty soon
ngnboone has quit [Quit: Leaving...]
<ReaganMcFarland[m]> It actually created the OS on the drive, but only got to 2.5 gigs worth then the installer crashed
<ReaganMcFarland[m]> I didn't want to try re-installing just yet, since it made the stub on the drive already
<marcan> at least we have the output this time around, after I fixed that... I have no idea why that fails
<marcan> looks like a diskutil bug
<marcan> I almost want to say it's some sort of race condition, like it created the volume but something along the line didn't notice it yet
<ReaganMcFarland[m]> How much is left in the script after this stage
<marcan> most of it
<ReaganMcFarland[m]> crap
<marcan> you need to delete the volume and try again
<marcan> diskutil apfs deleteContainer disk4
<nico_32> asahi linux: finding bug in Apple' code since 2021
yzz has quit [Ping timeout: 480 seconds]
<marcan> and yes, the past few days things have been quickly shifting from "fixing installer bugs" to "working around apple bugs"
<ReaganMcFarland[m]> failed again
<marcan> same problem?
<ReaganMcFarland[m]> Yea
<marcan> does "diskutil mount disk4s2" work?
<ReaganMcFarland[m]> Yes
<marcan> looks like I need a workaround for this nonsense then
<ReaganMcFarland[m]> damn - I was really looking forward to trying it out
<ReaganMcFarland[m]> Been following you guys since June last year
<ReaganMcFarland[m]> From the shadows though :)
<matthewayers[m]> marcan: what if the installer asks for a volume to prevent having to delete the volume and start over?
<marcan> that was a thing in the past but it's confusing and error-prone
<marcan> I will be adding a delete option instead in the future
<matthewayers[m]> Like a cleanup script of some sort?
<marcan> we already have a cleanup script, but it's kind of brute force
<marcan> I mean a proper installer option to delete an OS install, and also an expert option to just delete an arbitrary partition
<marcan> ReaganMcFarland[m]: can you do "diskutil apfs listVolumeGroups"?
<marcan> need to make sure the grouping worked
<matthewayers[m]> Gotcha, thanks for the info.
<ReaganMcFarland[m]> I deleted the disk again - let me try re-installing to see
<ReaganMcFarland[m]> I'm doing "max" when asked for how much space should be allocated - anyway that's involved?
<ReaganMcFarland[m]> * be allocated to the new OS - anyway
<marcan> not at this point, you're far enough yet for that to make a difference
<marcan> *not far enough yet
<ReaganMcFarland[m]> wait its working now
<marcan> lol.
<marcan> race condition, I'm sure.
yzz has joined #asahi
skipwich has joined #asahi
<LuigyLeon[m]> tpw_rules: thanks! vfinally my generations make sense again 😅
<marcan> just pushed a potential workaround deployed as v0.3.20: https://github.com/AsahiLinux/asahi-installer/commit/9effb6f898376fd251917ff258a737ffb4554373
caef^ has joined #asahi
<marcan> let's see if people stop reporting this failure mode :)
<marcan> argh, the f
<marcan> v0.3.21 it is then :')
<marcan> I always forget the f
<tpw_rules> lol me too
yuyichao_ has quit [Ping timeout: 480 seconds]
<nico_32> releng works is always fixing this kind of bugs
<nico_32> real world user data is always surprising
yzz has quit [Ping timeout: 480 seconds]
<M12destroyer21[m]> While updating to 12.3, my macbook air decided to nuke itself. I have a secondary Linux pc, on which I have compiled idevicerestore from main. I have my MacBook showing up in lsusb as a dfu device. When running the idevicerestore command it installs a bunch of things and a progress bar shows up on the MacBook screen. The screen then goes blank for a while when the Linux pc prints “ERROR: Device failed to enter restore mode”. The
<M12destroyer21[m]> MacBook then lights up with an exclamation mark and a link to apple.com/mac/restore and it shows up in lsusb as in recovery mode. Have any of you had this problem and do you have any ideas on how to fix this?
<nicolas17> (lol the link doesn't work? gg apple)
<M12destroyer21[m]> Sorry, It’s support.apple.com/mac/restore
<marcan> M12destroyer21[m]: you'll have to manually download the 12.2.1 IPSW and restore that one
<marcan> 12.3 is known broken on a small fraction of macs, for some unknown raeson
<marcan> restoring 12.3 again won't work
<nicolas17> ah okay
<marcan> download that and explicitly pass it to idevicerestore
<marcan> and you'll have to wait until apple fixes this then, unfortunately
<marcan> hopefully they'll release a 12.3.1 soon...
yuyichao_ has joined #asahi
yzz has joined #asahi
* marcan should sleep
creechy has quit [Ping timeout: 480 seconds]
yzz has quit [Ping timeout: 480 seconds]
deflax has quit []
deflax has joined #asahi
vyoid has joined #asahi
<chadmed> jeez your sleep schedule really has taken a beating. stay healthy!
creechy has joined #asahi
<vyoid> Hey, quick question. I tried deleting Asahi from my disk utility but whenever I'm trying to do a erase command, it spits out that the volume is being used by the kernel and cannot perform the erase. Anybody got to this state? Thanks -
<ReaganMcFarland[m]> Isn't the default root password just blank?
<chadmed> setting it to blank means the system will always reject any password
<chadmed> you cant log in as root
<chadmed> sudo -s
<ReaganMcFarland[m]> What's default sudo password
<chadmed> your user account password, just like any other system with sudo on it
<nicolas17> your user password
<ReaganMcFarland[m]> I'm so dumb
<ReaganMcFarland[m]> Thanks for being patient
<ReaganMcFarland[m]> :)
<nicolas17> "unfamiliar with Linux's peculiarities" is not "dumb" :)
yzz has joined #asahi
<chadmed> +1
<chadmed> lurking on IRC when you should be writing a patient case study? now *that's* dumb
creechy has quit [Ping timeout: 480 seconds]
yzz has quit [Ping timeout: 480 seconds]
<alden4[m]> i'm unable to format my drive, i don't have any time machine snapshots, just a update snapshot
creechy has joined #asahi
<tpw_rules> you'll have to either delete the update snapshot or shrink it by 3.3GB less
hl`` has joined #asahi
j`ey has quit [Remote host closed the connection]
j`ey has joined #asahi
hl` has quit [Ping timeout: 480 seconds]
nicolas17 has quit [Ping timeout: 480 seconds]
minecrell has quit [Quit: Ping timeout (120 seconds)]
minecrell has joined #asahi
nicolas17 has joined #asahi
THX1984 has joined #asahi
<Lucy[m]> Nice, I got the Pop-Asahi image working.
Kevin5 has joined #asahi
Kevin5 has quit []
kylealanhale has joined #asahi
THX1984 has quit [Quit: Leaving]
riker77_ has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
riker77 has quit [Ping timeout: 480 seconds]
riker77_ is now known as riker77
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
creechy has joined #asahi
eroux has joined #asahi
pzcz has joined #asahi
snek has quit [Server closed connection]
snek has joined #asahi
pzcz has quit [Remote host closed the connection]
sjg1 has quit [Server closed connection]
sjg1 has joined #asahi
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
creechy has quit [Ping timeout: 480 seconds]
creechy has joined #asahi
pzcz has joined #asahi
yzz has joined #asahi
pzcz has quit [Remote host closed the connection]
EvadingEvasions[m] has joined #asahi
<EvadingEvasions[m]> How long does extracting root.img usually take?
<nicolas17> I think it's not only extracting but also downloading concurrently, so it would depend on your internet connection
phiologe has joined #asahi
rann has quit [Server closed connection]
rann has joined #asahi
<EvadingEvasions[m]> nicolas17: All right, thanks for the information.
<nicolas17> someone said each + in the progress bar is another 10MB
creechy has quit [Ping timeout: 480 seconds]
pzcz has joined #asahi
<ReaganMcFarland[m]> Any tips on debugging audio issues? I have the original M1 macbook pro so I should be able to use my headphones
pzcz has quit [Remote host closed the connection]
PhilippvK has quit [Ping timeout: 480 seconds]
yzz has quit [Remote host closed the connection]
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
creechy has joined #asahi
Techcable has quit [Ping timeout: 480 seconds]
Gaelan has quit [Server closed connection]
<kode54> How do I get it back up and running again? I already have an EFI and Linux partition pair
Gaelan has joined #asahi
<kode54> and I so
<kode54> *don't know why Glanzmann was asking me about updating the efi image, it should already be fairly up to date, it was just created a couple of days ago
kov has quit [Quit: Coyote finally caught me]
kov has joined #asahi
yzz has joined #asahi
kenzie35 has quit [Server closed connection]
kenzie35 has joined #asahi
yzz has quit [Quit: leaving]
jn has quit [Read error: Connection reset by peer]
jn has joined #asahi
kylealanhale has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kode54 has quit [Server closed connection]
kode54 has joined #asahi
bpye has quit [Server closed connection]
bpye has joined #asahi
vyoid has quit [Quit: Page closed]
aratuk has quit [Quit: Leaving]
sheepgoose_ has joined #asahi
sorear has quit [Server closed connection]
sorear has joined #asahi
eric_engestrom has quit [Server closed connection]
eric_engestrom has joined #asahi
nicolas17 has quit [Ping timeout: 480 seconds]
sheepgoose has quit [Ping timeout: 480 seconds]
strangefield[m] has joined #asahi
Techcable has joined #asahi
tardyp has quit [Server closed connection]
tardyp has joined #asahi
nkaretnikov has quit [Server closed connection]
nkaretnikov has joined #asahi
strangefield[m] has left #asahi [#asahi]
sheepgoose_ has quit [Ping timeout: 480 seconds]
creechy has quit [Ping timeout: 480 seconds]
marvin24_ has joined #asahi
creechy has joined #asahi
marvin24 has quit [Ping timeout: 480 seconds]
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
<Glanzmann> kode54: Sorry, I wnated to know if your stub upgrade worked?
pzcz has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
pzcz has quit [Remote host closed the connection]
axboe has quit [Server closed connection]
axboe has joined #asahi
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
m2231puppy has quit [Quit: m2231puppy]
creechy has joined #asahi
<ReaganMcFarland[m]> Accidentally set my brightness to 0 in a Plasma widget - login to black screen every time now
WhyNotHugo has quit [Server closed connection]
WhyNotHugo has joined #asahi
handlerug has quit [Server closed connection]
handlerug has joined #asahi
<aem64[m]> How can i set macos back to my default boot volume?
<aem64[m]> also where can i see the asahi disk in disk utility?
raiaq has quit [Server closed connection]
raiaq has joined #asahi
coder_kalyan has quit [Server closed connection]
coder_kalyan has joined #asahi
BenPierce[m] has joined #asahi
<nametable[m]> <aem64[m]> "also where can i see the asahi..." <- Not sure about how to set the default boot volume, but I had a similar issue not seeing the disk in the disk utility. Although you can see if it you run `diskutil list` in the terminal.
tdmm has quit [Server closed connection]
tdmm has joined #asahi
<aem64[m]> nametable[m]: all i have is one nameds "Asahi" which takes up 2.5gb
<aem64[m]> oh here it is
<aem64[m]> probably Linux Filesystem
<chadmed> you can set macos as the default by going into System Preferences and choosing it as the Startup Disk
<aem64[m]> another thing that would be really nice is being able to read apfs drives (aka my mac installation)
<aem64[m]> chadmed: thanks
<chadmed> theres also an experimental apfs kernel driver floating around somewhere but i wouldnt trust it tbh
<aem64[m]> hmm
<aem64[m]> i might consider really using this once i can have video codec and gpu acceleration support
<aem64[m]> and figure out how to install box64
jleightcap has quit [Server closed connection]
jleightcap has joined #asahi
BenPierce[m] is now known as bgpierc[m]
<aem64[m]> damn can't verify atm
handlerug has quit [Remote host closed the connection]
handlerug has joined #asahi
handlerug has quit [Remote host closed the connection]
handlerug has joined #asahi
Ry_Darcy has joined #asahi
nyx_o has quit [Server closed connection]
nyx_o has joined #asahi
damien_vancouver has quit [Server closed connection]
<Ry_Darcy> @Glanzmann As discussed on Sunday, the M1 now boots from your Debian installation using a 2K monitor (Bern). On returning to Zurich yesterday, I tried booting the M1 again, this time using my 4K monitor - "Black screen" again. Are there any issues here with screen resolutions? Very strange.
damien_vancouver has joined #asahi
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
handlerug has quit [Remote host closed the connection]
handlerug has joined #asahi
<immychan[m]> <aem64[m]> "and figure out how to install..." <- Box64 isn’t really as useful as you’d think tbh
<immychan[m]> But I imagine it’d be as simple as installing a precompiled deb
Guest2753 has quit [Remote host closed the connection]
<lkvrsfld[m]> Like qemu-user is
<chadmed> i will say again
<chadmed> neither fex nor box64 work at all in asahi right now
<chadmed> and they likely never will on a native 16k pagesize kernele
<chadmed> you will have to install a patched 4K pagesize kernel yourself once sven's patches are merged
ihaveamac has quit [Server closed connection]
ihaveamac has joined #asahi
balrog has quit [Server closed connection]
balrog has joined #asahi
<chadmed> the tradeoff there is significantly reduced performance in some workloads
<chadmed> but if running x86 software is that important to you then its something you will have to deal with
<kode54> Glanzmann: I don't know what a stub upgrade is
kylealanhale has joined #asahi
Erus_Iluvatar has quit [Server closed connection]
Erus_Iluvatar has joined #asahi
<aem64[m]> how will upgrading work? will ther be a built in upgrade script or will we have to reinstall it from scratch?
<Chainfire> aem64[m]: so far, if you stick to Arch, pacman -Syu
<aem64[m]> ok
<aem64[m]> does the kyboard backlight work yet?
<Chainfire> Not for me
<aem64[m]> ok
creechy has quit [Ping timeout: 480 seconds]
creechy has joined #asahi
Techcable has quit [Remote host closed the connection]
jeffmiw has quit [Ping timeout: 480 seconds]
Techcable has joined #asahi
gvchmvxbmm[m] has joined #asahi
<gvchmvxbmm[m]> hello
<ptitSeb[m]> <chadmed> "neither fex nor box64 work at..." <- box64 works. I even made a blog entry about that https://box86.org/2022/03/box64-running-on-m1-with-asahi/
bdju has quit [Server closed connection]
bdju has joined #asahi
<chadmed> ohhh you guys already solved the page alignment on loongson
<chadmed> nice
<chadmed> wasnt aware of that!
Simonx22 has quit [Server closed connection]
Simonx22 has joined #asahi
<ptitSeb[m]> yeah, I already had some case of 16k page
<ptitSeb[m]> I need to clean up the patch, but it's already working
<ptitSeb[m]> (it has trouble to quit stuff, so still some bugs somewhere, but I suspect its purely internal to box64)
<Glanzmann> Ry_Darcy: Yes, there are problems with the dcp code and all resolutions bigger than 1080p are blacklisted.
<Glanzmann> Ry_Darcy: So at the moment the boot loader is initializing the dcp but blacklisting all modes bigger than 1080p. So what type of monitor are you trying? Do you have the exact model?
<Glanzmann> Ry_Darcy: I have two 4k screens in my home (TVs). I can try the mini with the tv under linux and let you know if they work.
<Glanzmann> kode54:
<Glanzmann> 07:03 <kode54> Glanzmann: if I still have the EFI and the Linux filesystem, can I just install the m1n1 + UEFI mode?
winter has quit [Server closed connection]
<Glanzmann> A stub is the fake macos the installer puts on a 2.5GB partition.
winter has joined #asahi
<Ry_Darcy> @Glanzmann - that might be (my) the problem indeed. My 4K monitor is a Samsung U28E590D.
<Glanzmann> I thought you asked me yesterday how to get rid of the old stub and install a new stub. My answer was: https://tg.st/u/upgrade_stub.txt
<kode54> no
<kode54> I was asking how to install a stub when I have none to begin with
<kode54> a system wipe erased the stub, but left the Linux partitions
Chainfire has quit [Server closed connection]
<kode54> I have no access to a Linux partition to do the first parts
<Glanzmann> kode54: The answer to that is: https://tg.st/u/upgrade_stub.txt
<kode54> I have no bootable Linux on this machine
<Glanzmann> kode54: Send me the output of 'diskutil list'
<Glanzmann> kode54: You can paste it here: tg.st/p
<kode54> oops, I used 0x0.st
<Ry_Darcy> Glanzmann - Strange, this monitor works with normal linux, MacOS and Pipcet's Linux. Are there any workarounds?
ptudor has quit [Server closed connection]
ptudor has joined #asahi
<Ry_Darcy> @Glanzmann Having said that, Pipcet had an option at initial installation to supress 4K. Even though 4K did work.
<chadmed> what version of macos were you using
<chadmed> the DCP firmware changed and broke some stuff with 12.0
<Ry_Darcy> 12.1 works perfectly works perfectly wit this monitor.
<Ry_Darcy> eith
<Ry_Darcy> with - all thumbs.
<chadmed> does it still work with macos?
<Ry_Darcy> yes.
<chadmed> Glanzmann: does the kernel you provide have the dcp patches in?
artemist has quit [Server closed connection]
<chadmed> if so try with the vanilla linux-asahi kernel which uses the framebuffer set up by iboot/m1n1 which should work
artemist has joined #asahi
<Ry_Darcy> I cannot install 12.3 on the M1. Has bricked my machine already at the weekend.
linearcannon has quit [Server closed connection]
<Ry_Darcy> Also, I would prefer to stick with Debian.
<chadmed> you dont need to move from debian
<Ry_Darcy> Still, 12.3 is a no-go for me at the moment.
linearcannon has joined #asahi
<chadmed> sure ok, but you can install a different kernel
<Glanzmann> chadmed: Nope.
<chadmed> oh well then that theory is irrelevant
<Glanzmann> kode54: So I would evacuate the efi partition, make space for the stub and create stub+efi.
<Glanzmann> kode54: Can you paste me the output of the 'mount' command?
<Glanzmann> kode54: Okay, next step is to mount the efi partition: mkdir /Volumes/efi; mount -t msdos /dev/disk0s3 /Volumes/efi
<Glanzmann> kode54: Than cd /Volumes/efi and take a backup of the EFI stuff: tar cfz efi.tar.gz EFI/
<Glanzmann> kode54: Than cd /Volumes/efi and take a backup of the EFI stuff: tar cfz ~/efi.tar.gz EFI/
<Glanzmann> Once you have done that and veriried that you have it by executing: tar tfz ~/efi.tar.gz
<Glanzmann> Than we get rid of the esp partition.
<Glanzmann> umount /Volumes/efi
<Glanzmann> diskutil eraseVolume free free disk0s3
<kode54> the EFI folder in that mount?
<kode54> it contains only one thing, BOOT/BOOTAA64.efi
<kode54> there are other folders, though
<Glanzmann> kode54: Yep, that's fine.
<Glanzmann> kode54: You can paste me the output of 'find /Volumes/efi' but we don't need them.
<as400[m]> Glanzmann: do I remember correctly that you have created debian installer image that can be booted from usb ?
<Glanzmann> as400[m]: Yes, it is here: https://alx.sh/w/Debian Please remeber to not choose 'use entire disk' so only use unpartitioned space.
<Glanzmann> Ry_Darcy: So I'll try now a patch to m1n1, if it works. for my 4k monitor, I send it to you.
<Glanzmann> Ry_Darcy: Are you able to access your mini using ssh?
<as400[m]> Glanzmann: thx man.
<Glanzmann> as400[m]: You're welcome.
<Ry_Darcy> Glanzmann, not at the moment, I am in the office. I will be able to work with the M1 in the afternoon.
<Glanzmann> Ry_Darcy: I see, I'll try it now, if I succeed, I send it to you.
<Glanzmann> So I'm trying this: https://tg.st/u/0001-dont-castrate-modes.patch
<Ry_Darcy> Glanzmann, you have my email address thougg. Thanks for the assistance.
<Ry_Darcy> though
<Glanzmann> you're welcome.
<Glanzmann> Ry_Darcy: I send you an email with my findings, but also put it in irc for the others.
creechy has quit [Ping timeout: 480 seconds]
<Ry_Darcy> Will do. Avec plaisir.
<Glanzmann> Ry_Darcy: I'm not sure if m1n1 stage2 reinitalized the display or leaves it alone, but I'm about to find out.
<Ry_Darcy> Glanzmann good luck.
Darkmatter66 has quit [Server closed connection]
Darkmatter66 has joined #asahi
kylealanhale has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Chainfire has joined #asahi
kahanis[m] has joined #asahi
<kode54> it won't let me install to just 2.94Gb of space
<kode54> how much space do I need to free up for this?
<chadmed> at least 3
<chadmed> 2.5gb for stub partition, 500mb for efi
creechy has joined #asahi
AndreasArayaOsorio[m] has joined #asahi
realmoot has joined #asahi
<Glanzmann> kode54: ^
<realmoot> Hi everyone
<realmoot> I am currently trying to run Asahi as a daily driver
<realmoot> Just facing one small issue of installing a text editor
<realmoot> VS Code gives a segfault
<kode54> all electron has the same problem as chromium
<roxiun[m]> Cos it’s electron
<roxiun[m]> I think that why at least
<roxiun[m]> Sublime should work right?
<kode54> 16k pages don't work well with that
<realmoot> and Sublime Text says:
<realmoot> <jemalloc>: Unsupported system page size
<realmoot> <jemalloc>: Unsupported system page size
<realmoot> GLib (gthread-posix.c): Unexpected error from C library during 'malloc': Cannot allocate memory. Aborting.
<realmoot> [1] 61448 IOT instruction (core dumped) subl
<roxiun[m]> Oh
skipwich has quit [Server closed connection]
skipwich has joined #asahi
<roxiun[m]> kode54: Yeah that makes sense
<realmoot> Thanks for responding roxiun[m]
Major_Biscuit has joined #asahi
<roxiun[m]> Np
<roxiun[m]> Hmm so what editors do work?
juancri has joined #asahi
<realmoot> Are you also running Asahi?
<roxiun[m]> Aside from vim and nano ig
<realmoot> I didn't know about any other GUI editors...
<roxiun[m]> realmoot: nah not yet
<roxiun[m]> I’m planning on once I can be confident battery life and gpu will work well enough
<roxiun[m]> Cos of school
<realmoot> But apparently these packages work on Rpi4, so i was hoping they work on Mac
<realmoot> Its a 2020 M1 mbp btw
<AdryzzOLEDEdition[m]> the pi uses 4K pages
<AdryzzOLEDEdition[m]> e.g. they split up the memory in 4KB chunks
giskard has joined #asahi
<AdryzzOLEDEdition[m]> while M1 has 16K pages
<realmoot> Oh okay
<realmoot> Is that why sublime also probably wont work?
<kode54> jemalloc has the same problem
<AdryzzOLEDEdition[m]> there's a list of apps that currently don't work with 16K pages
<kode54> which also breaks all of Rust
<kode54> so you can't use paru if you're on Arch
<realmoot> Oops alright..
<AdryzzOLEDEdition[m]> kode54: aaaa
<kode54> but you can use yay, because that's in go
<realmoot> Would be interested in suggestions on text editors
<AdryzzOLEDEdition[m]> paru > yay
<realmoot> yes Im using go
<kode54> I know
<AdryzzOLEDEdition[m]> i use nano
<kode54> but paru is no-go right now
<realmoot> Also I cant play any video at all in firefox
<kode54> I can play video in Firefox here
<AdryzzOLEDEdition[m]> realmoot: audio issue
<realmoot> kode54: it just spins for me
<kode54> it needs a sound device
<realmoot> how do you give it a sound device
<kode54> headphone jack is mostly non-working even though it should be
<kode54> you plug in a USB sound device
<AdryzzOLEDEdition[m]> "switch the audio profile to input and back to duplex"
<realmoot> I tried headphone jack and idn't work
<Sobek[m]> Is there any fix in the pipeline for jemalloc ?
<AdryzzOLEDEdition[m]> it should play even without an audio device
<Sobek[m]> (I wonder whether IntelliJ stuff can run on Asahi, I don’t think they are electron)
<realmoot> I can download the video using something like youtube-dl and it plays (without sound) but firefox just spins all the time
Major_Biscuit has quit [Ping timeout: 480 seconds]
<AdryzzOLEDEdition[m]> try switching the audio profile to input and back to duplex
<Dcow[m]1> Sobek, do they provide builds for arm?
<AdryzzOLEDEdition[m]> if it spins it's because it can't start audio playback (yes, even without an audio device)
<Glanzmann> Ry_Darcy: The experiment did not work.
<Glanzmann> Ry_Darcy: So we have to wait for jannau/marcan to give us proper gpu.
Glanzmann has quit [Quit: leaving]
<realmoot> AdryzzOLEDEdition[m]: The second I change to Input youtube video plays but back to duplex it pauses as if its buffering
<AdryzzOLEDEdition[m]> i have no clue then, just relaying an old troubleshootiung step
<AdryzzOLEDEdition[m]> s/troubleshootiung/troubleshooting/
<realmoot> ok now it works AdryzzOLEDEdition[m]
<realmoot> Thanks
<AdryzzOLEDEdition[m]> i don't have a machine (yet?) so i just lurk here
<realmoot> kode54: what mac are you using Asahi on?
<kode54> Mac mini
<kode54> and now I'm back in it
<Sobek[m]> Speaking of jemalloc https://github.com/jemalloc/jemalloc/pull/2242 appears to have been merged which ought to fix the issue once things are released / propagate downstream ?
<AdryzzOLEDEdition[m]> :( not dynamic
<j`ey> Sobek[m]: no because we have generic aarch64 builds
<Sobek[m]> (And given the way rust is bootsrapped it won't work unless generic builds are fixed)
<Ry_Darcy> @Glanzmann in other words, concerning the M1, my 4K monitor is useless at the moment. Schade!
<chadmed> Sobek[m]: thats not really going to fix anything
<chadmed> oh yeah what j`ey said
<Sobek[m]> :/ I wonder why there’s no link to a bug report for jemalloc on the page tracking those issues ? Should I like open one ?
kylealanhale has joined #asahi
<kode54> unfortunately, not that useful
<j`ey> most of the comments are from 2017, so it could be worth bringing up again
<kode54> as most packagers bake in a page size of 4KB
<kode54> so it would require packagers to bake in the page size of 64KB and allow it to work for 16 and 4
<kode54> it only works for smaller, not larger
<jannau> I think most packagers do nothing and get the default page size of the compile host which is usually 4k
kylealanhale has quit [Ping timeout: 480 seconds]
<Sobek[m]> <j`ey> "most of the comments are from 20..." <- Especially given the state of hardware has changed and page size shenanigans are happenning more and more. The idea of compiling 3 copies for 16k, 64k and 4k and selecting one implementation at runtime might be the more intelligent solution, provided the selection can work reliably. (Right now kernel change, but it may become per process given there are reason to support 4k and 16k on asahi linux,
<Sobek[m]> eventually (4k being pretty useful to make x86 translation work))
<Sobek[m]> The issue has been mentioned like twice this year btw
<AdryzzOLEDEdition[m]> it's not even a check in a hot path
<AdryzzOLEDEdition[m]> it is a one time thing at startup
<Sobek[m]> If you set-it up this way yes. But they pointed out that some compile time structure relied on page size, so you have to do it that way and you can't just have a rather synamic implementation where the same code handles the three case.
<Sobek[m]> The main cost is obviously making the library take up three time the space it used to (which will feed the troll about rust binaries sizes obviously ^^')
<chadmed> shipping fat binaries for page size bs is just straight up never going to happen, especially for stuff that embededd systems rely on
<chadmed> and good luck getting packagers to compile multiple copies of things for various page sizes, and then good luck getting users to configure their package managers to get the right one
<chadmed> jemalloc already has the infrastructure for this to not be a problem with a single command line switch in packagers' build scripts
<AdryzzOLEDEdition[m]> aaaa this is insane
<Sobek[m]> Hmm, embedded people are the one who can likely add in the switch to specify which arm page size they want. (They know their hardware).
<Sobek[m]> It sounds like the page size should be part of the target triple or something though, given the impact it has on portability. Is it deemed a part of the ABI or not btw ?
jackhill has quit [Server closed connection]
<AdryzzOLEDEdition[m]> are custom memory allocators that much better in the first place?
<kode54> would anyone know what I've done differently to alarm to make it crash sddm on boot?
<Sobek[m]> jemalloc is generally linked statically / vendored or is it a package that asahi linux could override on M1s with a version suitable for it ?
<AdryzzOLEDEdition[m]> statically afaik
<kode54> telegram-desktop in the arch repos links it dynamically
<kode54> as does some mariadb thing that KDE installs
<kode54> everything else will need to be built with larger page sizes by default
jackhill has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
<kode54> there we go
<Sobek[m]> (Larger page size or page size detection)
<kode54> larger page size should be forced by the build scripts by default
<kode54> since the library supports arbitrarily smaller page sizes at runtime
ptudor_ has joined #asahi
<Sobek[m]> This is where we should probably have a chat with distro maintainers to see what's the favoured solution for them ?
<kode54> probably
realmoot has quit [Quit: Konversation terminated!]
<Sobek[m]> (At least it appears M1 is not the only aarch64 chip out there that has some page size subtleties, Loongson is another one from what appears earlier in the chat)
tbodt has quit [Server closed connection]
tbodt has joined #asahi
MajorBiscuit has joined #asahi
ptudor__ has joined #asahi
creechy has joined #asahi
<j`ey> Sobek[m]: RHEL used 64K page sizes on arm for a while.. but they build the entire distro so can control that
ptudor has quit [Ping timeout: 480 seconds]
ptudor_ has quit [Ping timeout: 480 seconds]
<ptitSeb[m]> <Sobek[m]> "(At least it appears M1 is not..." <- Loongson is not aarch64, it's longarch64, based on mips (but different)
<Sobek[m]> Ah, well, I got confused ><
<ptitSeb[m]> but it do use 16k pages
<ptitSeb[m]> s/do/does/
<gsora[m]> 4k kernel i bet
<enchiladasconpinguino[m]> vscode working on latest alpine kernel (4k page)
<enchiladasconpinguino[m]> of course yeah, but working great fwiw
creechy has quit [Ping timeout: 480 seconds]
<gsora[m]> those little m1 chips are a joy to work with
<enchiladasconpinguino[m]> <1s ide startup time, off usb even..
<gsora[m]> the bootup time of the whole system is so fast i can't believe it's a cold boot
<enchiladasconpinguino[m]> my rough benchmark so far, maybe 10% perfomance loss from 16 -> 4k pages
<enchiladasconpinguino[m]> but now software works so meh worth it
rbenua has quit [Server closed connection]
rbenua has joined #asahi
<AdryzzOLEDEdition[m]> exactly as running with mitigations=off on intel lol
pent1ckel has quit [Remote host closed the connection]
<megalodon42[m]> hello, I still cannot figure out how to finalize the install with the bputil fail at step2... should I restart the install from scratch or it is generally not a good idea?
creechy has joined #asahi
<megalodon42[m]> I've been craving for this install for so long haha :)
pent1ckel has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
hell__ has quit [Server closed connection]
hell__ has joined #asahi
<nilsi[m]> enchiladasconpinguino: can i do `apk add` with the 4k patch that someone linked here before after installinging the KDE distro with the installer in order to get 4k support?
<nilsi[m]> * 4k support? Or how did you do it_
<nilsi[m]> * 4k support? Or how did you do it?
<j`ey> apk add is for alpine, not arch
ptudor__ has quit [Ping timeout: 480 seconds]
<nilsi[m]> Oh, okay. I guess not then :)
creechy has joined #asahi
ptudor has joined #asahi
<mps> nilsi[m]: you can download kernel for alpine (apk package) and untar it
<mps> this is built with 4K page size
<mps> so `tar -xf linux-asahi-5.17.rc7.20220318-r0.apk -C /` should install it on Arch, but you have manually to add it in bootloader (grub)
<mps> though not sure will it work with systemd as pid 1, never tested this
<nilsi[m]> Cool, I might give it a try after work tonight.
ptudor_ has joined #asahi
hitmoon has joined #asahi
<megalodon42[m]> thank you to all of you for your work with Asahi, especially marcan. I can't get it installed but I will continue to follow your progress from alpha to beta and so on! Cheers!
<Chainfire> Does the M1 support 64kb and does anyone have a kernel for that?
<maz> no, and no.
<j`ey> megalodon42[m]: you should make a github issue, with the logs
<megalodon42[m]> j`ey: thank you
<jn> maz: i'm curious, what is missing in hardware for 64k pages?
pzcz has joined #asahi
<maz> jn: I guess support in the PTW, for example.
<AdryzzOLEDEdition[m]> 64K pages sounds like a waste too if the lowest memory size is 8GB
pzcz has quit [Remote host closed the connection]
creechy has quit [Ping timeout: 480 seconds]
<maz> AdryzzOLEDEdition[m]: it has nothing to do with the memory capacity, but with the way you use that memory.
<AdryzzOLEDEdition[m]> yes but you can't allocate half a page
ptudor has quit [Ping timeout: 480 seconds]
<Chainfire> I have dynamic Chromium patch for 4kb/16kb working (busy with cleanup), it'll break on 64kb though
<AdryzzOLEDEdition[m]> so if you don't need the full 64k it is going wasted
<j`ey> marcan: you didnt bother testing the installer with spaces as your password?? :P
<maz> AdryzzOLEDEdition[m]: with 64kB, I want 512M blocks. it's not about individual page allocation, but efficiency of the TLBs.
megalodon42[m] has left #asahi [#asahi]
<AdryzzOLEDEdition[m]> isn't it the same issue of cluster size on persistent storage?
<maz> i.e. waste of memory is not a concern if my workload is about squeezing out every bit of performance.
<AdryzzOLEDEdition[m]> o yea ofc
<arnd> AdryzzOLEDEdition[m]: https://docs.google.com/spreadsheets/d/1Y-eeXEHr8Tud2ul4iApn6hLEQoMsl3Bsud-vlW3V2t8/edit#gid=0 has some numbers from an experiment I did on my typical workload (building kernels with gcc).
<arnd> using 64KB more than doubles the memory usage in this case, but as long as you have enough memory, it's still slightly faster than 16KB pages
<maz> I'm not condoning the use of 64k pages on arm64 (I wish we never had the option), but there *are* workloads that benefit (a bit) from them on some implementations.
<arnd> I think what makes the 64K pages so tempting is that the workloads that benefit most are exactly the things that get used in CPU benchmarks. The same thing is true for ilp32 mode, which is something we managed to dodge for arm64
<arnd> "x32" mode on arch/x86 serves mostly to make Intel look good in some benchmarks, but causes way too many problems for the limited practical use
creechy has joined #asahi
hays has quit [Server closed connection]
hays has joined #asahi
zv has quit [Server closed connection]
zv has joined #asahi
Glanzmann has joined #asahi
<Glanzmann> Ry_Darcy: Good news, I think it works.
<Glanzmann> marcan[m]: Is it safe to initialize the dcp on the mini from the stage2 m1n1 or should it be done from the stage1 m1n1?
<Glanzmann> Ry_Darcy: I can provide you with a m1n1 rust binary which you can install in 1tr that sets the display correctly.
<sven> it's safe from stage 2 as well
<Glanzmann> sven: Than I have to try that.
<sven> it's no different that iboot first initializing the display and then m1n1 trying it again
creechy has quit [Ping timeout: 480 seconds]
<Glanzmann> sven: I thought we're only initalizing the display, if it is not initalized: https://pbot.rmdir.de/3Wg6NBVLr7-muf34oAGwkw
<Glanzmann> But let me try that.
<sven> right, so for stage 2 there's no different if stage 1 and/or iboot did anything
<sven> *difference
<Glanzmann> Okay.
Z750 has quit [Server closed connection]
Z750 has joined #asahi
<Glanzmann> Wow there is now a percentage in the installer. How fancy. :-)
<enchiladasconpinguino[m]> mps: so tried a fresh install with qemu to physical usb like we discussed
<enchiladasconpinguino[m]> but your kernel doesn't seem able to boot "crypt+sys" install ?
reillyeon has quit [Server closed connection]
reillyeon has joined #asahi
joske has joined #asahi
<enchiladasconpinguino[m]> mount: mounting /dev/mapper/root on /sysroot failed: No such file or directory
<enchiladasconpinguino[m]> Mounting root failed.
<enchiladasconpinguino[m]> I also tried manually editing the grub.cfg to use UUID and match lts kernel setup, but same error
<mps> enchiladasconpinguino[m]: huh, so I didn't enabled some of device mapper modules, I guessing
<mps> enchiladasconpinguino[m]: ah you mean in qemu VM
<enchiladasconpinguino[m]> yeah some kernel module config for cryptsetup missing I think
<enchiladasconpinguino[m]> I haven't even gotten to booting new install off bare metal yet
<Glanzmann> mps / enchiladasconpinguino[m] Here is a blaoted config which enables almost anything also stuff you don't need on the m1: https://tg.st/u/config-2022-03-17-distro-sven-jannau.txt
<enchiladasconpinguino[m]> making sure it can boot in QEMU first
<mps> enchiladasconpinguino[m]: I didn't tested with crypt setup. you can join #alpine-linux on OFTC and ask there
<enchiladasconpinguino[m]> oh so your daily driver install no crypt setup? I guess I just assumed that worked
<mps> Glanzmann: thanks, but I'm trying to keep alpine slim and not bloated :)
<enchiladasconpinguino[m]> yeah, I doubt its more than one or two missing kernel module
creechy has joined #asahi
<mps> enchiladasconpinguino[m]: join #alpine-linux or #alpine-devel and ask
<enchiladasconpinguino[m]> okay I will investigate further and ask around, thanks
<mps> (I'm not on these channels because there are a lot of annoying and sometimes bad people)
<enchiladasconpinguino[m]> lol same
<enchiladasconpinguino[m]> otherwise we are pretty close to standard alpine installer working
<enchiladasconpinguino[m]> actually I will try just plain "sys" install next with this method, to prove that it does boot bare metal correctly
<mps> enchiladasconpinguino[m]: I'm busy with upgrading kernel to 5.17 on alpine now, and will be probably all the day
<enchiladasconpinguino[m]> gotcha no worries, thanks mps
<mps> enchiladasconpinguino[m]: by the chance, do you know rust building, cargo (cult ;) )
<enchiladasconpinguino[m]> nope I don't do any rust stuff sorry, totally ignorant on that one
<mps> enchiladasconpinguino[m]: np, also I don't
<enchiladasconpinguino[m]> for native stuff, c/c++/go/zig only
<AdryzzOLEDEdition[m]> i know just basic stuff about cargo
<mps> though I ported rust to alpine for armv7 and aarch64 2 or 3 years ago, but after that I lost interest in rust
<mps> AdryzzOLEDEdition[m]: hah, I introduced zig to alpine about 2 years ago, and I still maintain it but also loosing interest on it
<AdryzzOLEDEdition[m]> all i use rust for is M0+ on the pi pico
<mps> starting again to love plain C and perl
<mps> though bought book about crystal and learning it for financial programming
<mps> and posix shell ofc
<gsora[m]> crystal for financial programming? wow!
<gsora[m]> does it fit well in the financial ecosystem?
<mps> gsora[m]: it is modeled like ruby and ruby is strong in this field
<j`ey> I have this untested APKBUILD for m1n1 that should work with rust https://termbin.com/3d0v
chir4gm has quit []
chir4gm has joined #asahi
<mps> gsora[m]: there are some financial apps (big ones) in crystal so I think it is good, and as compiled lang it is fast, lot faster than ruby (and faster than rust)
<mps> j`ey: will test it
<j`ey> it needs changing to add all the new sources
<Sobek[m]> <mps> "enchiladasconpinguino: by the..." <- What exactly are you interested in ?
<mps> j`ey: I see
<j`ey> and probably should set CARGO_HOME and RUSTUP_HOME to somewhere in /tmp
<gsora[m]> mps: interesting! didn't even know ruby had a following in the financial environment
<mps> gsora[m]: not mainstream financial apps (microsofot ...) but in some knowledgeable shops it is
<mps> Sobek[m]: ?
<mps> Sobek[m]: don't understand question
<j`ey> Sobek[m]: it's alpine specific really
<Sobek[m]> Matrix IRC might have mangled who I was replying to, I was asking enchiladas, wrt his rust / cargo question.
<mps> j`ey: m1n1 still doesn't build with your patch on my environment
emptynine has quit [Server closed connection]
emptynine has joined #asahi
<j`ey> mps: what's the error?
<j`ey> I don't have aports on my m1 yet
<mps> j`ey: ehm, let me rerun
joske has quit [Remote host closed the connection]
joske has joined #asahi
chir4gm has quit []
chir4gm has joined #asahi
<mps> j`ey: running with ash as shell I don't get error, just warnings. will investigate later and report to you
<j`ey> mps: ok
Ariadne has quit [Server closed connection]
Ariadne has joined #asahi
<mps> j`ey: at the rust git submodules --init should be added in APKBUILD, probably in prepare()
<mps> s/at/and/
<Ry_Darcy> @Glanzmann Will I be receiving some stuff?
<j`ey> mps: that wont work since the APKBUILD downloads a tarball
<mps> j`ey: or there is way to downoad this as tarball
<j`ey> mps: I don't know of a way to download it all
<kaprests> Probably a noob question, but will 'all' the software available for Arch Arm just work on the M1 macs with Asahi, or will the individual packages have to be adapted individually by maintainers/developers/community? Just curious since the latter scenario was the case for software for macOS for Apple silicon at least, yet I can
<mps> j`ey: ok, will ask later if someone know and would help to us
<j`ey> kaprests: most will work yes
<j`ey> currently asahi comes with a 16K kernel which means it has issues with software such as https://github.com/AsahiLinux/docs/wiki/Software-known-to-have-issues-with-16k-page-size
<kaprests> Wow, that's pretty great! How come software for macOS hade to be ported individually then?
<j`ey> because that was from x86 to arm
tmlind has quit [Server closed connection]
tmlind has joined #asahi
<j`ey> ArchLinuxArm is already arm :P
<kaprests> Oh yes of course!
<kaprests> Excited to try out the alpha then :D
<Sobek[m]> kaprests: Mostly because a lot of the mac software is closed source + macos never ran on ARM before, whereas linux is a. open source so recompile is quite a bit easier, and linux already runs on ARM so software have already been ported to ARM.
<EvadingEvasions[m]> I got a question. Can you transfer files on the mac to the linux OS?
<kaprests> Sobek[m]: Yeah I see! Makes alot of sense :))
<Sobek[m]> (But the page size stuff is likely going to cause headache to a bunch of people before it gets better I fear)
<j`ey> EvadingEvasions[m]: not easily, currently
<EvadingEvasions[m]> j`ey: And difficultly?
<Sobek[m]> EvadingEvasions[m]: So, I think there are some tools that allow mount ext2 on macos (using fuse or something like that), if they have made the jump to ARM land. And the other way round, there's a fuse APFS fs that can be used (but in readonly, be careful).
<EvadingEvasions[m]> I'll see
<j`ey> Sobek[m]: macOS can also read FAT natively, so you can make a small FAT partition for transferring files if you really wanted
<Sobek[m]> Or you can dump the files on a dumb USB key with a fs that both OS like (FAT32 for small files, I don't remember off the top of my head what to use for big files).
<enchiladasconpinguino[m]> ^ this is safe to use to copy from macos to linux
<EvadingEvasions[m]> Oh and can you play video on Asahi atm?
<chadmed> yes
<EvadingEvasions[m]> I can't seem to though.
<chadmed> your machine will scream at you since its software rendered but it does work
<EvadingEvasions[m]> Whenever I go to youtube.com it just goes at one by one.
<Glanzmann> sven: Do you consider this patch sane? https://tg.st/u/0001-Always-initialize-display-and-allow-modes-upto-width.patch it works.
<EvadingEvasions[m]> And ideas on when Chromium will be done? Looks like there's some progress going on already.
<Glanzmann> Ry_Darcy: So, I'll make you a patch ready. You need to install it and reboot and than let me know if it works.
<Glanzmann> Ry_Darcy: I'll also provide you with the binary.
<chadmed> no idea on chromium since i dont use it and its a problem for upstream
<chadmed> honestly my advice is simply do not use electron apps at all
<Glanzmann> Ry_Darcy: And thank you for asking me this question, now I can finally use my mini as workstation. Because before I never bothered because dcp was too slow for me and with the framebuffer I got only 1080p.
<gsora[m]> are macbook trackpads handled through libinput?
<chadmed> yes
<EvadingEvasions[m]> Maybe.
<gsora[m]> i can't seem to find a scrolling speed config
<_jannau_> Glanzmann: I don't think we want to always re-initilize the display. at least not on devices with integrated display
<EvadingEvasions[m]> gsora[m]: I did it.
<EvadingEvasions[m]> you meant the sensitivity ?
<gsora[m]> i think? it scrolls too much for the amount of actual finger travel i use
<EvadingEvasions[m]> Ok wait
<EvadingEvasions[m]> first run `xinput --list`
<EvadingEvasions[m]> And send screenshot
<Glanzmann> jannau: I'm not planning to distribute it besides to Ry_Darcy and me. I'm using this now on my mini and it appears to work perfectly.
<chadmed> if youre on kde, just go system settings -> input devices -> touchpad -> scrolling speed
maxung[m] has joined #asahi
<_jannau_> I think you're change will also break on macbook pro 14/16 and imac since the native/only mode for the display has a width > 2560
<Glanzmann> jannau: I see.
<chadmed> Glanzmann: you know that simply posting it here means 300 people will pick it up and uncritically apply it and then sook when it breaks something :P
<gsora[m]> > > <@gsora:awake.noskills.club> i can't seem to find a scrolling speed config... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/EdkkpDZKKIcukHzIpIsIZAxC)
<Glanzmann> chadmed: Okay, I remove the file.
<EvadingEvasions[m]> Ok wait
<EvadingEvasions[m]> gsora: now run
<EvadingEvasions[m]> `xinput --list-props 8`
<EvadingEvasions[m]> 8 is the ID of your track pad
<gsora[m]> ohhhhh
<_jannau_> Glanzmann: what we could do with chainloading is to append a variable 'preferred_mode' to allow selection of higher res / wide screen moodes
<gsora[m]> yeah it's 100% libinput
<EvadingEvasions[m]> Yeah.
<gsora[m]> i can use any libinput config thing then
<EvadingEvasions[m]> cool
<EvadingEvasions[m]> I run a startup script
<gsora[m]> can you share it here?
<EvadingEvasions[m]> xinput --set-prop 7 279 0.6`
<EvadingEvasions[m]> `
<gsora[m]> OT: fonts look gorgeous on plasma wth, it's been years since i tried it properly last time
<Glanzmann> jannau: That would make sense, yes.
<EvadingEvasions[m]> Currently in the process of ricing it.
<Glanzmann> jannau: I wanted it to make it as easy as possible for Ry_Darcy to apply it.
<EvadingEvasions[m]> Catpuccin theme, gonna use EWW widgets next.
<chadmed> plasma rocks, idk how people use any other DE these days
<EvadingEvasions[m]> KDE good.
<gsora[m]> thanks for the script! it's much better now
<gsora[m]> chadmed: i've been a lifelong gnome/i3 user, kde 4 scared me away tbh
<mps> Sobek[m]: few days ago I compiled linux-apfs-rw kernel module and tested it, it works though upstream say it is not safe yet
<kode54> I figured out why Gnome Terminal may have failed
<chadmed> incidentally gnome has sucked for about as long as kde has rocked
<mps> in meantime I use exfat
<maxung[m]> Glanzmann: Could you send me the patch in private, I also have a Mac Mini but with 1080p it's not worth considering for me
<EvadingEvasions[m]> gsora[m]: change ids respectively
<gsora[m]> yup ofc
<Glanzmann> Ry_Darcy: Run this command as root and reboot. Let me know if your screen works. If it does not work at first. Try to turn the mini off again and on again. This patch works for me on a 2k display. curl -sLo /boot/efi/m1n1/boot.bin https://tg.st/u/rydarcy.bin
<kode54> basic arch linux arm doesn't set up locale.gen
<kode54> and sets the default locale to "C" too
<EvadingEvasions[m]> Rice coming out well though.
<Glanzmann> maxung[m]: done.
<EvadingEvasions[m]> https://github.com/catppuccin
<EvadingEvasions[m]> In case anyone wants to use it
<Glanzmann> maxung[m]: Edit the number for your maximul resolution.
<Glanzmann> IIRC marcan put the 1080p limit there because some monitors offered a virtual 4k resolution while they did not have one.
<chadmed> yeah there are monitors with screwed up EDID/whatever the new extended version is called
<maxung[m]> Thanks, I'm using 1440p, so should fit
<maxung[m]> s/fit/be fine/
<gsora[m]> anybody using a native matrix client for plasma?
<chadmed> eventually obviously we will tear down and reinitialise a framebuffer in the dcp kernel driver so it shouldnt be an issue long term
<chadmed> but we need SOME display output for m1n1 and u-boot
<Glanzmann> maxung[m]: I would edi the resolution exactly to your resolution, because on one of my editors it does not work with the version for what I send you.
_alice has quit [Server closed connection]
_alice has joined #asahi
<maxung[m]> Yeah I noticed, the other version you posted before had the correct one ^^
<Ry_Darcy> @Glanzmann - will try this afternoon. I will report back. Thanks.
klange has quit [Server closed connection]
klange has joined #asahi
<Glanzmann> maxung[m]: I corrected it.
<Glanzmann> maxung[m]: So this patch as you see it when you press reload, works for me on the eizo, the one with the 4k resolution does _not_ work for me on an eizo.
<Ry_Darcy> @Glanzmann you say run as root, where as root? MacOS?
<Glanzmann> Ry_Darcy: On Linux.
<Glanzmann> Ry_Darcy: And for me this does not work on a 2k monitor, only on a 4k.
<Glanzmann> However if you installed openssh-server and have a non-root user, you can login remote and exchange it.
skoobasteeve has quit [Server closed connection]
skoobasteeve has joined #asahi
<Glanzmann> jannau: Maybe adding the parameter is not such a bad idea. So we could overwrite the parameter without touching m1n1 stage 1, can we?
<Glanzmann> jannau: If I add your dcp driver, but do not use xrandr, should it work with performance and hdmi hotplug? If that is the case, I'll try that later.
creechy has quit [Ping timeout: 480 seconds]
<_jannau_> yes, I would expect that just appending the param to ESP:m1n1/boot.bin will work
the_lanetly_052 has joined #asahi
pzcz has joined #asahi
<_jannau_> dcp is complicted, let me think about it tonight
pzcz has quit [Remote host closed the connection]
<Glanzmann> jannau: I see. I'll try it once you give me a go ahead.
<gsora[m]> ugh, neochat from [community] also doesn't support 16k pages
<EvadingEvasions[m]> Why is chrome crashing for me?
<EvadingEvasions[m]> Is chrome unusable?
<AdryzzOLEDEdition[m]> chromium doesnt support 16K pages
<EvadingEvasions[m]> Yeah but why can't it launch?
<AdryzzOLEDEdition[m]> because it doesn't support 16K pages
<EvadingEvasions[m]> So it can't even launch... R.I.P.
<AdryzzOLEDEdition[m]> it can't allocate memory
<gsora[m]> does 4k-only software terminates with SIGTRAP?
<Glanzmann> jannau:
creechy has joined #asahi
<gsora[m]> mprotect(0xffff63202000, 8192, PROT_NONE) = -1 EINVAL (Invalid argument) < crashes on this
Hotswap has quit [Server closed connection]
Hotswap has joined #asahi
<Ry_Darcy> Glanzmann will do. Thanks.
kameks has joined #asahi
robher has quit [Server closed connection]
robher has joined #asahi
bradfier has quit [Server closed connection]
bradfier has joined #asahi
pulsar[m] has quit [Server closed connection]
pulsar[m] has joined #asahi
brudda[m] has joined #asahi
<brudda[m]> Hello, is anyone else having issues downloading some packages? I can't download xorg because libepoxy always fails to download
ave3 has quit [Server closed connection]
ave3 has joined #asahi
<brudda[m]> ```
<brudda[m]> ```
<brudda[m]> Failed to connect to mirror.archlinuxarm.org port 80 after 3050ms: No route to host
<brudda[m]> Bruh
<brudda[m]> Oh well, that was the error
<chadmed> works fine for me
<chadmed> might be your isp
<chadmed> or your network configuration
acez[m] has quit [Server closed connection]
<brudda[m]> It probably is, though plenty of other packages install fine
acez[m] has joined #asahi
<brudda[m]> I'm on university WiFi its a pain to get connected anyway lol
KDDLB has quit [Server closed connection]
KDDLB has joined #asahi
WindowPain has quit [Server closed connection]
<brudda[m]> Also, I'm guessing its not yet possible, but can I replace systemd with openrc?
Skirmisher has quit [Server closed connection]
Skirmisher has joined #asahi
WindowPain has joined #asahi
<_jannau_> brudda[m]: general arch linux (arm) instructions should apply for switching to openrc but it's offtopic here
<brudda[m]> Okay, thank you : )
povik has quit [Server closed connection]
povik has joined #asahi
jpnurmi has quit [Server closed connection]
jpnurmi has joined #asahi
chadmed has quit [Remote host closed the connection]
subject38[m] has quit [Server closed connection]
subject38[m] has joined #asahi
branon has quit [Server closed connection]
ChanServ changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | https://asahilinux.org/2022/03/asahi-linux-alpha-release/ | 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
branon has joined #asahi
<enchiladasconpinguino[m]> mps: quick update with some good news for you, official 3.15 alpine.iso "sys" install working with your kernel :)
vx has quit [Server closed connection]
vx has joined #asahi
<enchiladasconpinguino[m]> only required small tweaks to grub.cfg and your .apk
<opticron> I really hate that Android stole the .apk extension from Alpine
NoahSchill[m] has quit [Server closed connection]
NoahSchill[m] has joined #asahi
<mps> opticron: :)
<mps> opticron: openwrt also plan to start use apk
pfdub[m] has quit [Server closed connection]
pfdub[m] has joined #asahi
<ReaganMcFarland[m]> Hey guys! Accidentally set my screen brightness in my KDE plasma desktop to 0 so now I log into a black screen
<mps> enchiladasconpinguino[m]: ok, that is expected, not everything is fixed
<ReaganMcFarland[m]> Any tips?
<opticron> mps, that'll at east be the same package format (probably)
<j`ey> ReaganMcFarland[m]: can you press F2 I think it is? to turn it back up
<opticron> Android's .apk files are entirely unrelated to Alpine's
<ReaganMcFarland[m]> How do the function keys work with touch bar disabled?
<mps> opticron: yes, pkg format and tool (apk)
<j`ey> ReaganMcFarland[m]: fn+number row
<marcan> (but that won't help for media keys...)
<mps> opticron: right, android just took (stole) name of pkg extension
pwg has quit [Server closed connection]
pwg has joined #asahi
<marcan> ReaganMcFarland[m]: you can ctrl-alt-f1 (option-ctrl-Fn-1) to a text terminal, log in from there, and there should be some file you can delete to remove that preference from KDE... though I don't know what it is off the top of my head
<marcan> and yes the brightness situation is kind of dodgy right now, especially on the MBP with no touch bar
<mps> opticron: openwrt working activelly to adopt apk but they wait for apk version 3
<marcan> that will be fixed with DCP
dianshi has quit [Server closed connection]
dianshi has joined #asahi
<marcan> actually I think you need tty2, so ...-2 instead of ...-1
<gsora[m]> the backlight thing you mean?
<marcan> maybe I should just hack gpio-backlight to ignore the brightness attribute entirely for the time being...
<marcan> yes
<gsora[m]> oh nice
<mps> marcan: for me works on tty1 also
<marcan> mps: is this arch?
<mps> marcan: alpine
<marcan> then your experience is irrelevant
<mps> :D
<gsora[m]> oof
os has quit [Server closed connection]
<marcan> let's stop mixing together distros, please
<marcan> stuff like what tty a login manager spawns on is not the same everywhere
os has joined #asahi
<enchiladasconpinguino[m]> speaking of brightness, anyone working on gamma control in DCP yet?
<mps> marcan: I didn't noticed that the question is for specific distro
<j`ey> enchiladasconpinguino[m]: I dont think so
<marcan> mps: I think it's safe to assume that if someone comes asking for help and they don't mention a different distro, they're probably on ALARM :p
<mps> marcan: heh, you are right
caef^ has quit [Remote host closed the connection]
the_real_briel[m] has quit [Server closed connection]
the_real_briel[m] has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
pzcz has joined #asahi
pzcz has quit [Remote host closed the connection]
limegorilla[m] has quit [Server closed connection]
limegorilla[m] has joined #asahi
plaer1[m] has quit [Server closed connection]
plaer1[m] has joined #asahi
<enchiladasconpinguino[m]> so its possible to set brightness from macos, and then persists in linux? but not the gamma control?
<enchiladasconpinguino[m]> DCP state survives reboot?
creechy has joined #asahi
<Glanzmann> enchiladasconpinguino[m]: Nope.
<Glanzmann> enchiladasconpinguino[m]: I think someone set the brightness from macos and rebooted and the brightness state is unmodifed.
<mps> yes, brightness persist
<mps> but not gamma
megalokafes has quit [Server closed connection]
megalokafes has joined #asahi
linuxgemini9 has quit [Server closed connection]
linuxgemini9 has joined #asahi
<enchiladasconpinguino[m]> okay thanks, brightness is better than nothing
<enchiladasconpinguino[m]> my tired old eyes can't look at blue lit full brightness for too long..
HagenSchrder[m] has quit [Server closed connection]
HagenSchrder[m] has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
al3xtjames has quit [Server closed connection]
al3xtjames has joined #asahi
creechy has joined #asahi
turo has quit [Server closed connection]
turo has joined #asahi
<maxung[m]> <Glanzmann> "maxung: So this patch as you see..." <- patch works perfectly, now i’m tempted to try 144hz as well
<maxung[m]> patch works perfectly, now i’m tempted to try 144hz as well (although i don’t know if that makes sense with software rendering)
<maxung[m]> * > <@_oftc_Glanzmann:matrix.org> maxung: So this patch as you see it when you press reload, works for me on the eizo, the one with the 4k resolution does _not_ work for me on an eizo.
<Glanzmann> maxung[m]: I just run the dcp experiment on my monitor.
<Glanzmann> I see no modes with more than 60hz. But I see 4k modes, but my monitor does not support 4k.
<Glanzmann> I assume that is why marcan blacklisted it.
vmeson has quit [Server closed connection]
vmeson has joined #asahi
creechy has quit [Ping timeout: 480 seconds]
realmoot has joined #asahi
<realmoot> Hey All
realmoot has quit []
joske has quit [Remote host closed the connection]
<marcan> the iBoot DCP interface is too dumb to do proper mode selection
hspak has quit [Server closed connection]
<marcan> former iBoot and m1n1 do it differently and each gets it right and wrong in different cases
hspak has joined #asahi
<marcan> apple could fix this, but since they aren't doing third party monitors at all any more... I doubt they ever will
msteffen has quit [Server closed connection]
msteffen has joined #asahi
nametable[m] has quit [Server closed connection]
nametable[m] has joined #asahi
<marcan> in fact I dread that thing disappearing, if it does we'll be forced to implement full-fat DCP in m1n1 which will *suck*. but I guess if it comes to that at least we'll be able to do proper mode selection...
<Glanzmann> marcan: I see. This also explains why I don't get a picture in iboot. They probably try to set a 4k resolution on a 2k screen.
<marcan> no, they do not try anything
<Glanzmann> Oh, I see.
<marcan> iBoot stopped initializing HDMI monitors entirely
<marcan> with 12.0
<marcan> 4K modes *do* work, they are just *scaled* which is obviously silly
<marcan> the monitor still gets 1080p
<marcan> there is no flag in the mode table telling you it's a scaled mode
<marcan> that's why I just cap it at 1080p
<marcan> the full DCP interface does have flags for non-native
<maxung[m]> <Glanzmann> "I see no modes with more than..." <- tried as well, same result, but i see exactly 1 mode
<Glanzmann> povik: Bad news, now that I started using the mini I can make my mini hang by going to alsamixer and press f6. System still pings, but I can't log in via ssh.
realmoot has joined #asahi
snnw[m] has quit [Server closed connection]
adamjm has quit [Server closed connection]
adamjm has joined #asahi
snnw[m] has joined #asahi
<Glanzmann> marcan: When I modified the m1n1 for a 4k resolution, I had no visiable output, if I modify it for 2k it works.
<realmoot> Glanzmann: Whats m1n1?
<realmoot> Sorry im a noob at this
<Sobek[m]> realmoot: It is the thing marcan wrote that pretends to be a macOS kernel and then goes around and loads the rest of the open OS boot chain
<Glanzmann> realmoot: A bootloader/hypervisor/firmware/debugger that marcan wrote for m1 machines.
creechy has joined #asahi
<realmoot> Ah okay thx
<Sobek[m]> It can also contains a fun hypervisor that is used to help reverse engineering what macos does.
<Glanzmann> realmoot: So out boot chain is iboot1 > iboot 2> m1n1 stage 1> m1n1 stage 2> u-boot > grub > kernel.
<realmoot> Oh wow that's complicated
<Glanzmann> realmoot: Yes it is, but also very flexbibel.
<realmoot> I am liking asahi tbh, very cool what was achieved
<marcan> Glanzmann: we also do not reallocate framebuffers; it is possible 4K does not fit
<realmoot> Only wish sublime-text / jemalloc issue was fixed
<realmoot> But i understand that's due to 16K buffers instead of 4K
<Glanzmann> marcan: I see.
<Sobek[m]> And each piece has a simpler job to do. Also U-boot and grub are pieces that are more or less off the open source shelve that required few patches. The big work is m1n1 to interface with how macs want to boot stuff (they expect something that looks like macOS and quacks like macOS), and linux kernel to have all the drivers.
<marcan> u-boot did get a ton of driers
<marcan> *drivers
<marcan> grub has zero patches
<marcan> it is completely off the shelf since it uses EFI services
<j`ey> (provided by U-boot)
<Sobek[m]> Oh well, I was wrong about U-boot (hopefully it is not also a Yellow kind of submarine, unless I'm very wrong :D)
<Sobek[m]> realmoot: Pages, not buffers.
<_jannau_> Glanzmann: I think the frame buffer allocated by iboot is not large enough for 4k
<marcan> and I *know* I can ship a 4K kernel and "fix" all those things for y'all but I want y'all to instead complain upstream because not supporting 16K pages is an upstream package problem and the solution isn't to force everyone to lose performance to go 4K
<marcan> because I know if we start shipping 4K kernels this will never get fixed, like it hasn't until now
<marcan> you're all going to have to endure the main ;)
<marcan> *pain
<_jannau_> I was speaking about a 4k display mode
<marcan> I mean about realmoot's page size comment
<Sobek[m]> Plus, don't you need 16k pages for some PCIe / hardware stuff because of the IOMMU ?
<marcan> Sobek[m]: sven has a patchset to make it work with 4K, with a few caveats
<realmoot> Agreed marcan
<Glanzmann> jannau: I see. Should I try it anyway or we wait until Ry_Darcy has tried it.
<marcan> and we will need to offer 4K kernels in the future for FEX/Android/similar users, running legacy/JITed software that really wants 4K and cannot be recompiled
<marcan> but that should be opt-in, not what everyone jumps to because Chromium doesn't work
<sven> the caveats are "needs a bit cleanup for upstream" and "doesn't support untrusted devices because swiotlb can't align to more than PAGE_SIZE"
<_jannau_> Glanzmann: I thought you already tried and it failed
<sven> i have a branch somewhere that fixes #1, but have been too lazy to look into swiotlb because that part is a bit tricky
<marcan> yeah
<marcan> so in its current state we can't really have 4K + thunderbolt (I will not ship thunderbolt without proper isolation, that's a hard line for me)
nexus7556 has quit [Server closed connection]
<sven> it fails safely ofc
<marcan> yeah
<sven> it'll just not allow any DMA to anything thunderbolt
<Glanzmann> jannau: I see. You mean because I tried it with my screen which offered a 4k resolution but it stayed blank?
<marcan> which basically makes thunderbolt useless other than for display and USB4 tunneling
nexus7556 has joined #asahi
<sven> yeah
<Sobek[m]> Also, at that point will it be possible for that kernel to offer 16k page to processes by default, unless requested otherwise ? (Or is that impossible to do in linux because it would require some special userland interface that cannot be added). That way only processes that require 4k page will opt-in ?
<j`ey> which is already quite usefil?
<j`ey> *ful
<marcan> Sobek[m]: that's a bit of a pipe dream for linux
<sven> i'll probably be motivated to look into what swiotlb does again once thunderbolt is ready ;)
<marcan> would be great if we worked out some magic way of pulling that off that would fly upstream
<marcan> but that's one heck of a discussion
<marcan> and it's unclear if there's a solution that is viable to implement
<marcan> would probably tie into the ongoing memory folios work somehow too
<marcan> but right now that kind of level of discussion is above my pay grade ( ;) )
<marcan> maybe once everyone's running Asahi because these really are the best Linux laptops available then we'll have a good excuse to really bring that up
<Sobek[m]> What's the memory folio thing ? (I can take a pointer to an explanation if that would bring the channel to far off topic for too long)
<marcan> basically building 16K pages on top of 4K pages
<j`ey> it's reorganising how memory is managed in the kernel
<j`ey> (for all arches)
<marcan> yeah
<j`ey> > Managing memory in 4KiB pages is a serious overhead.
<j`ey> first sentence :D
<marcan> if we have a kernel which is already somewhat familiar with the idea of higher-order pages, we'll have a better chance at pulling off hybrid page size userspace
axenntio has joined #asahi
pzcz has joined #asahi
DarkShadow4444 has quit [Server closed connection]
RevHelix has quit [Server closed connection]
DarkShadow44 has joined #asahi
<Sobek[m]> marcan: Is this something that could be brought up to some linux maintainer for advice / opinion on what they think should be done long term. Like Asahi linux alone can't decide on these sort of thing, but perhaps we can get the people in charge thinking a bit and provide guidance on how to make the part fall in place long term ?
<marcan> yes, that's how these things are done :)
pzcz has quit [Remote host closed the connection]
<marcan> but there's no rush to do this right now, and it helps when you already have users asking for it, and hard benchmark numbers, and things like that
<j`ey> lets get normal drivers in first
<sven> marcan: oh... now that I think about this again: If we could somehow manage to sneak subpage protection in without having to hack around the IOMMU API itself too much we could have thunderbolt even without swiotlb :D
<Sobek[m]> j`ey: I love it. Actually it might be a great solution to the problem. The kernel pretendes it deals only in 4k pages, while it in fact tries very hard to use as many 4k pages as possible :D
<marcan> sven: true!
<j`ey> Sobek[m]: yes :)
<marcan> it would be kind of funny if we end up with the opposite model as macOS
<marcan> macOS uses 16K kernel and mixed 4K/16K userspace modes
<marcan> linux might end up with 4K kernel but treating it as 16K or higher management-wise, and then mixed 4K/16K userspace
<marcan> no idea how much performance that'd lose but it might be totally viable
<marcan> 4K kernel might make it a lot easier to support 4K userspace without undue pain
<marcan> since it's always easier to build larger pages out of smaller ones than it is to somehow manage subpages
<Sobek[m]> sven: But is there any way we can get this sort of thing to work ? I could see a very inefficient method which is deny real access to the whole and do the subpage check and if legal modification in teh fault handler (like ZFOD and CoW design tend to do), but it sounds very very inefficient perf wise ?
<sven> DART has hardware subpage protection support with a 4 byte or so granularity
<sven> so you can add a full page to the IOMMU pagetable but only allow access to a subset of it
<ReaganMcFarland[m]> Are there any known issues with enterprise internet issues - like at a University?
<brentr123[m]> Works for me at my college internet
<ReaganMcFarland[m]> We have tunneled TLS but the network manager not giving me the ability to connect
<marcan> oh, like 802.1x WPA auth?
<marcan> or you mean a VPN?
<ReaganMcFarland[m]> WPA2 enterprise
<marcan> yeah, I don't think anyone has tried that, might be broken!
<rowang077[m]> g
<ReaganMcFarland[m]> Where do we file bug reports?
<tpw_rules> matthewayers[m]: did you get this to work? i remember you asking about uni internet?
<marcan> depends on where the problem is
<marcan> would be cool if you can figure it out first
<marcan> it could just be missing packages
<marcan> it could be an actual kernel problem
<ReaganMcFarland[m]> I'll do some digging
<WhyNotHugo> So it seems that `rustc` from the repos does not work, but the one provided by `rustup` DOES work.
<mps> Sobek[m]: I've found that F2FS kernel driver doesn't work with 16K pages and maybe someone will find something else when the kernel is more tested 'in the wild'
axenntio has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<j`ey> WhyNotHugo: yes, it's known that arc build with jemalloc
<j`ey> *arch
<Sobek[m]> 802.1X are a regular pain to configure on various OSes might be troubles only remotely related to Asahi (like just the usual arch + wifi + 802.1X troubles)
<marcan> ReaganMcFarland[m]: if you're just missing a package in the end, you can file it at https://github.com/AsahiLinux/asahi-alarm-builder/ to have it added to the default images
<marcan> if you end up finding it's a userspace issue, file it at https://github.com/AsahiLinux/PKGBUILDs so we can figure out if we can ship a patched version
<marcan> if it's a kernel problem file it at https://github.com/AsahiLinux/linux
<Sobek[m]> mps: Uh, what is that driver for ?
<mps> Sobek[m]: Flash Friendly Filesystem
<marcan> and yeah, there's a good chance this is just "general linux stuff"
<j`ey> ReaganMcFarland[m]: do you have it working on another linux system?
<marcan> I'm happy to preinstall stuff to make it work, since you end up with a chicken and egg network problem otherwise
<mps> f2fs is used on newer android afaik
<j`ey> ReaganMcFarland[m]: also on the m1 with macos?
<marcan> but I can't fix UX issues around it much, other than possibly tweaking default configs
<marcan> not much we can do about pervasive Linux user experience problems this early on, chipping away at things like that is very low on our priority list and best handled upstream if it has nothing to do with the hardware in these devices per se
<marcan> (at least at this point)
<mps> asahi is alpha release and as such I think it is very good (and stable for me)
<marcan> if it ends up requiring some convoluted config that's best handled with a wiki page (or a link to existing docs if any)
<Sobek[m]> Already the fact that Asahi has tackled the keyboard mess will be a huge contribution for many linux users (if hopefully this is a success).
<mps> Sobek[m]: what keyboard mess?
<Glanzmann> jannau: I had to try 4k. I wouldn't say that it doesn't work you can use more than half of the screen. the top half: https://tg.st/u/IMG_20220322_143257143_HDR.jpg
joske has joined #asahi
<maxung[m]> Glanzmann: is this patch still needed for a 16K kernel (https://tg.st/u/0001-4k-iommu-patch-2022-03-11.patch)?
<j`ey> maxung[m]: 4k kernel
<j`ey> Glanzmann: nice clothes rack
<Sobek[m]> mps: "What is the best default keyboard layout for the various languages, and what bug are still hidden". I'll probably find a few stupid thing with the Fr (mac) layout once my backup situation is sorted out and I can tweak stuffs. I remember quite a few headaches with odd characters not being typed by the right keys on my previous mac linux dual boot a couple of years ago.
<maxung[m]> so if i want to compile just for 16k page sizes, i don't need it?
<j`ey> maxung[m]: correct
<maxung[m]> ok thanks
<Glanzmann> maxung[m]: Yes, and you have the most recent version which applies to the current asahi branch. You also need to set CONFIG_ARM64_4K_PAGES=y and CONFIG_ARM64_64K_PAGES=n
<maxung[m]> so I need it or not?
dumistaken[m] has joined #asahi
<Glanzmann> j`ey: I didn't saw the clothes rack in the picture, I thought you complain about the vacuum cleaner.
<maxung[m]> * or not? I'm confused
<dumistaken[m]> hello!
<mps> Sobek[m]: ah. though I fix these things in userspace
<Glanzmann> maxung[m]: No, you only need it for 4k page size kernel.
<Glanzmann> maxung[m]: So if you want to run chrome or webkit stuff.
<WhyNotHugo> j`ey: I'm aware that it's a known jemalloc issue, but then: why does the rustup version work...?
<maxung[m]> I don't, so leaving it out - thanks
<j`ey> WhyNotHugo: doesnt use jemalloc
<j`ey> WhyNotHugo: it's an optional feature
kameks has quit [Ping timeout: 480 seconds]
<_jannau_> Glanzmann: that looks like it's just a m1n1 / bootargs problem. the whole vram is mapped seems to be mapped and is large enough to not crash dcp. I would have expected m1n1 rejects that instead of using just 2/3 of the screen
<WhyNotHugo> j`ey: Oh, so the archlinuxarm package could be patched to do the same, right?
yuyichao_ has quit [Ping timeout: 480 seconds]
<Sobek[m]> mps: I believe you can go around and manually fix stuff, but fixing the original keyboard layout shipped by distros and making sure the default one selected is the sane one is still nice :)
<j`ey> WhyNotHugo: yeah
<j`ey> WhyNotHugo: jannau did a build here https://www.jannau.net/rust-arch/ not sure if he wants people to use it or not
<Glanzmann> jannau: Okay, if you have a patch for me or an idea, let me know and I'll try it.
<Glanzmann> jannau: So you say there should be enough ram for 4k, but m1n1 is not using it?
<mps> Sobek[m]: ok. though ime distro installers usually asks for keyboard layout
axenntio has joined #asahi
<_jannau_> Glanzmann: I'm not sure. I'll look tonight
<j`ey> _jannau_: oh I didnt see it was in the repo now
<_jannau_> I don't think there are builds and those commits are just in branches in the repo
<Glanzmann> jannau: jannau: I see. If I should try something, let me know.
<j`ey> cool, still good to know for pointing people to
<_jannau_> there is still a bootstrap problem, the PKGBUILD rejects to build with rustup so you need to compile binaries on a 4k system first
roxfan has quit [Ping timeout: 480 seconds]
sjg1 has quit [Server closed connection]
sjg1 has joined #asahi
Ry_Darcy has quit [Remote host closed the connection]
Ry_Darcy has joined #asahi
<Glanzmann> Ry_Darcy: I tried the 4k patch with my tv, only half the screen is usable. So maybe you want to try the 2k patch and see if you get a picture with it: curl -sLo /boot/efi/m1n1/boot.bin tg.st/u/2k.bin
yuyichao_ has joined #asahi
rann has quit [Server closed connection]
rann has joined #asahi
<rucadi[m]> marcan @marcan:matrix.org: check the Pull Request on PKGBUILDs, there was a typo on the packaging of the headers and they are missing a folder needed to compile kernel modules
<marcan> rucadi[m]: thanks, let me do a quick revbump release with that and the provides
Gaelan has quit [Server closed connection]
Gaelan has joined #asahi
<marcan> WhyNotHugo: I already have a fix for rust actually, but I pulled it from the repo since it wasn't part of our PKGBUILDs
<marcan> it's pending testing that when *built* on 4K it really works on 16K
<marcan> haven't had a chance to try that yet
<marcan> if someone has a 4K ALARM system lying around (M1 or not, doesn't matter) feel free to try it out and report back if it works on 16K Asahi
kov has quit [Server closed connection]
kov has joined #asahi
oi_wtf has quit [Server closed connection]
oi_wtf has joined #asahi
axenntio has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
XeR has quit [Server closed connection]
XeR has joined #asahi
axenntio has joined #asahi
<marcan> rucadi[m]: new kernel should be up :)
<marcan> (plus headers obviously, which is the only thing that changed)
<rucadi[m]> Thank you :D
<Glanzmann> joske: I often use it to rescue or bootstrap systems from a live system: https://pbot.rmdir.de/3gZQ1rX9_KIzTQV1ky4uuQ
kenzie35 has quit [Server closed connection]
kenzie35 has joined #asahi
<realmoot> On a non technical note, has anyone used a torrent client on Asahi that they can recommend?
FranCEACERO[m] has joined #asahi
<realmoot> Oh just saw KTorrent
<mps> realmoot: that doesn't depend on asahi but on distro
<mps> realmoot: just install one from distro repo and check
<joske> Glanzmann: yes, that usage of bind mount I know, but bind mounting the same folder on itself and then seeing that it has mounted /dev/nvme0n1p5 is very frightening ;-)
<realmoot> I am so grateful for the Asahi team and marcan! Thank you for making all this possible team
<M1bn3mar[m]> hey, how partition a free space using diskutil
<M1bn3mar[m]> *to
fortelling[m] has quit [Server closed connection]
fortelling[m] has joined #asahi
jakebot has quit [Server closed connection]
jakebot has joined #asahi
kode54 has quit [Server closed connection]
kode54 has joined #asahi
bpye has quit [Server closed connection]
bpye has joined #asahi
<M1bn3mar[m]> Glanzmann: i already have a free space available so it to be in the linux file system format so i could manually install gentoo on it, thx tho ; )
<M1bn3mar[m]> *so i need it to be
irth has quit [Server closed connection]
irth has joined #asahi
LiteChen[m] has quit [Server closed connection]
LiteChen[m] has joined #asahi
<Glanzmann> M1bn3mar[m]: See: https://git.zerfleddert.de/cgi-bin/gitweb.cgi/m1-debian/blob_plain/refs/heads/master:/files/quickstart.txt 'use the live system to install' and replace the debian specific steps with the gentoo specific steps.
AlbertoBasaglia[m] has quit [Server closed connection]
AlbertoBasaglia[m] has joined #asahi
eric_engestrom has quit [Server closed connection]
eric_engestrom has joined #asahi
nkaretnikov has quit [Server closed connection]
nkaretnikov has joined #asahi
leon_kernan[m] has quit [Server closed connection]
leon_kernan[m] has joined #asahi
cobre[m] has quit [Server closed connection]
cobre[m] has joined #asahi
zx2[m] has quit [Server closed connection]
zx2[m] has joined #asahi
Emantor has quit [Server closed connection]
Emantor has joined #asahi
beep-boop[m] has quit [Server closed connection]
beep-boop[m] has joined #asahi
altf5[m] has quit [Server closed connection]
altf5[m] has joined #asahi
life-the-user[m] has quit [Server closed connection]
life-the-user[m] has joined #asahi
JohnWells[m] has quit [Server closed connection]
JohnWells[m] has joined #asahi
obrmao[m] has quit [Server closed connection]
obrmao[m] has joined #asahi
US[m] has quit [Server closed connection]
US[m] has joined #asahi
<Glanzmann> 404
trevyn4 has quit [Server closed connection]
trevyn4 has joined #asahi
axboe has quit [Server closed connection]
axboe has joined #asahi
<Lucy[m]> Glanzmann: wait I forgot GitHub defaulted to private... 🤦‍♀️
<Lucy[m]> try now.
tt022[m] has quit [Server closed connection]
tt022[m] has joined #asahi
MZG[m] has quit [Server closed connection]
MZG[m] has joined #asahi
xyaman has quit [Server closed connection]
xyaman has joined #asahi
Quixote[m] has quit [Server closed connection]
Quixote[m] has joined #asahi
raizo[m] has quit [Server closed connection]
raizo[m] has joined #asahi
beacon[m] has quit [Server closed connection]
beacon[m] has joined #asahi
<Glanzmann> Lucy[m]: Works. :-) I need to make a video how I use a Linux desktop in order to show a speedy desktop experience. :-)
nyx_o has quit [Server closed connection]
nyx_o has joined #asahi
damien_vancouver has quit [Server closed connection]
damien_vancouver has joined #asahi
<Glanzmann> Lucy[m]: Don't forget to add it to: https://github.com/AsahiLinux/docs/wiki/SW%3AAlternative-Distros
<brudda[m]> <Glanzmann> "1bn3mar: See: https://git...." <- So I can use this to get gentoo on the M1?
<brudda[m]> Awesome, I was just looking for how to do this
<Glanzmann> brudda[m]: Yes, and if you fail tell me and I'll make you a video or create another installation image.
levadopa[m] has quit [Server closed connection]
levadopa[m] has joined #asahi
<Chainfire> "<realmoot> Only wish sublime-text / jemalloc issue was fixed" ... sublime-text works for me?
<Chainfire> scratch that, brainfart
<brudda[m]> Glanzmann: Thank you! I've not installed gentoo before so I will try not to come to you with any generic issues
<maxung[m]> Glanzmann: wrote "But currently the USB-A Port on the Mac Mini will not work in u-boot and grub". Does that mean there is currently no way to select anything in grub (I also tried USB-C to USB-A dongle)?
<j`ey> maxung[m]: USBC port should work
<Glanzmann> maxung[m]: with a usb-a to usb-c dongle or hub it works. But fancy keyboard sometimes don't work. For example I have keychron keyboard which does not work.
<Glanzmann> But I also have a thinkpad keyboard with a track point which works.
benmkw[m] has quit [Server closed connection]
benmkw[m] has joined #asahi
<maxung[m]> Glanzmann: oh wow I also have a keychron keyboard
Fl1tzi[m] has quit [Server closed connection]
<Glanzmann> maxung[m]: But to be honest, most of the times you don't need to select anything. And the usb-a ports currently have usb 3 speed while the usb-c only have usb 2 ports.
<maxung[m]> thats probably why
Fl1tzi[m] has joined #asahi
ihaveamac has quit [Server closed connection]
ihaveamac has joined #asahi
balrog has quit [Server closed connection]
<maxung[m]> I compiled my own kernel and (luckily) it worked, I was just wondering in the case it doesn't, so I can use the original as fallback
balrog has joined #asahi
<Glanzmann> maxung[m]: A noname keyboard should work out of the box.
<Glanzmann> maxung[m]: I have this keyboard: Keychron K1 Wireless Mechanical Keyboard (Version 4) - 87-key / White Backlight Aluminum Body / Gateron Low Profile Blue
ExeciN[m] has quit [Server closed connection]
ExeciN[m] has joined #asahi
<maxung[m]> I have the K7
matrixmanu[m] has joined #asahi
<maxung[m]> my corsair keyboard works ^^
qiuren[m] has quit [Server closed connection]
qiuren[m] has joined #asahi
eagles13[m] has quit [Server closed connection]
eagles13[m] has joined #asahi
MajorBiscuit has quit [Ping timeout: 480 seconds]
roxfan has joined #asahi
trouter has quit [Server closed connection]
Techcable has quit [Server closed connection]
trouter has joined #asahi
Techcable has joined #asahi
yrlf has quit [Server closed connection]
yrlf has joined #asahi
Tano70 has joined #asahi
Simonx22 has quit [Server closed connection]
nyuhu has quit [Server closed connection]
<dumistaken[m]> Not sure if this is necessarely a asahi issue or me being dumb, but im having some trouble installing vscodium with pacman
Simonx22 has joined #asahi
nyuhu has joined #asahi
<dumistaken[m]> i keep getting the following error
<j`ey> vscode doesnt work under asahi currently due to 16K page sizes issues with chromium
<dumistaken[m]> that explains it?
<dumistaken[m]> kinda?
<dumistaken[m]> i mean im not sure if its related
<dumistaken[m]> im stuck on the package install part
<j`ey> well it should still install, but not run
<dumistaken[m]> doesnt install
<dumistaken[m]> error: missing package metadata in VSCodium-linux-arm64-1.60.0.tar.gz
<dumistaken[m]> error: 'VSCodium-linux-arm64-1.60.0.tar.gz': invalid or corrupted package
<dumistaken[m]> ps
<dumistaken[m]> it does install if i clone the git, and compile it myself, but only if i do makepkg -i. if i do makepkg and then do pacman manually, same issue
guillaume_g has quit [Server closed connection]
winter has quit [Server closed connection]
guillaume_g has joined #asahi
auroramatrixorg[m] has quit [Server closed connection]
winter has joined #asahi
auroramatrixorg[m] has joined #asahi
Ry_Darcy_ has joined #asahi
<dumistaken[m]> im going to try another package, see if i run into the same issue
<Ry_Darcy_> Glanzmann - will not be able to try your fix (4K monitor) this evening. I will do my best tomorrow. My apologies.
Ry_Darcy_ has quit []
noebr[m] has quit [Server closed connection]
noebr[m] has joined #asahi
DominikInfhr[m] has quit [Server closed connection]
DominikInfhr[m] has joined #asahi
<brudda[m]> sorry if this is a stupid question, but which partitions do I gave to format to remove the install?
<brudda[m]> I just don't wanna accidentally nuke macos
<M1bn3mar[m]> brudda use diskutil list
<brudda[m]> I can give the output of diskutil list
artemist has quit [Server closed connection]
<dumistaken[m]> <dumistaken[m]> "im going to try another package,..." <- intresting. i just tried installing 2 other packages from a .tar file and it keeps failing with that same error
mr_macintosh[m] has joined #asahi
<dumistaken[m]> anyone else run into this? its totally possible im a idiot
artemist has joined #asahi
<j`ey> dumistaken[m]: wait, are you running pacman -U on that tarball?
<dumistaken[m]> j`ey: yes, is that wrong? Im pretty new to arch, havent used pacman before
<Glanzmann> brudda[m]: Leave the first, second and last partition.
<Glanzmann> brudda[m]: Past the output of 'diskutil list' to tg.st/p
<WhyNotHugo> realmoot: deluge worked fine for me.
<j`ey> dumistaken[m]: did you just download it from here? https://github.com/VSCodium/vscodium/releases/tag/1.65.2
<j`ey> dumistaken[m]: those are just plain tarballs I guess, not pacman packages
<dumistaken[m]> j`ey: i got the one that said linuxarm64.tar.gz from there, yes
<j`ey> dumistaken[m]: yeah that's not a pacman installable pkg
<j`ey> dumistaken[m]: just a plain tarball
<Glanzmann> brudda[m]: So you want to delete Linux?
<brudda[m]> yes
<brudda[m]> i just want to reinstall though
<Glanzmann> brudda[m]: For you its safe to get root 'sudo -i' and run the wipe script: curl -L https://alx.sh/wipe-linux | sh
<brudda[m]> so i can give gentoo a shot, or just go with the graphics, because for whatever reason my network wont let me install xorg when im connected on the minimal
<Glanzmann> brudda[m]: Actually when you want to gentoo.
<brudda[m]> Glanzmann: thank you
<Glanzmann> You can also leave the partitioning.
<Glanzmann> press return in u-boot and type 'run bootcmd_usb0'
<Glanzmann> To boot from the livestick.
<Glanzmann> Than you can reuse the partitions.
<Glanzmann> Because the esp and stub you need anyway. The root you can reformat with ext4 and put gentoo on it.
<dumistaken[m]> j`ey: so im retarded, makes sense
<dumistaken[m]> ty
THX1984 has joined #asahi
<brudda[m]> okay, so run the wipe script, dont touch the partitions, get a gentoo livestick boot it in uboot, then just install gentoo as normal?
<brudda[m]> also, a dongle should work for USB right?
<Glanzmann> brudda[m]: Don't run the wipe script.
<Glanzmann> brudda[m]: Yes, a dongle works.
<Glanzmann> brudda[m]: Put the debian live on a msdos usb stick and than run debian live. From there install gentoo.
meowcat285[m]1 has quit [Server closed connection]
meowcat285[m]1 has joined #asahi
<Glanzmann> brudda[m]: IIUC you need the stage3 tarball https://www.gentoo.org/downloads/
nzzz_ has quit [Server closed connection]
nzzz has joined #asahi
<Glanzmann> For arm64. but I just downloaded it and had a look at it they don't have a grub.
Ry_Darcy_ has joined #asahi
<Glanzmann> Ry_Darcy_: You read my email?
<AdryzzOLEDEdition[m]> is github up for you? i get a 500 on the wiki
<Ry_Darcy_> @Glanzmann, just did. I will not have the time this evening. Certainly tomorrow. Lot of family gegenwind here.
axenntio has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Glanzmann> brudda[m]: You could use the tg.st/u/grubaa64.efi grub. this needs to be put on ...
ryan_nupp[m] has quit [Server closed connection]
<Glanzmann> AdryzzOLEDEdition[m]: Yes, https://tg.st/u/screenshot-air-2022-03-22-17_15_40.png alx.sh/w
ryan_nupp[m] has joined #asahi
<M1bn3mar[m]> hey, Glanzmann i think brudda hasn't done anything like this before maybe he'll brick his system :/
ReaganMcFarland[m] has quit [Server closed connection]
ReaganMcFarland[m] has joined #asahi
<brudda[m]> the most ive done is install an arch based system on a thinkpad
<Glanzmann> Ry_Darcy_: I see.
<brudda[m]> perhaps i should wait when it comes to gentoo on the m1
<j`ey> brudda[m]: do you have any important info on the m1?
<Glanzmann> brudda[m]: It is already there, I saw two guys running gentoo on m1. :-)
<Glanzmann> brudda[m]: Maybe I should build you an image for gentoo.
<brudda[m]> not particularly anymore, i also have a time machine backup on an external drive
<M1bn3mar[m]> note that: Glanzmann some packages may not compile like webkit-gtk and llvm
<Glanzmann> M1bn3mar[m]: I use a 4k kernel. So I assume they will.
<j`ey> llvm will work
<Glanzmann> M1bn3mar[m]: ALso I don't want to compile anything, I only want to get to the point that the system boots and package it up.
<M1bn3mar[m]> yeah, but will emerge compile??
<Ry_Darcy_> Glanzmann, will do my best tomorrow.
Ry_Darcy_ has quit []
<Glanzmann> M1bn3mar[m]: emerge is pyton and contained in the stage3.
<Glanzmann> So it doesn't have to compile.
<M1bn3mar[m]> Glanzmann: i mean will it be capable of compiling packages like webkit-gtk and llvm
<Glanzmann> M1bn3mar[m]: jey said, it will. So why shouldn't it?
<M1bn3mar[m]> since u know it's about the end user and all
chr has joined #asahi
<j`ey> M1bn3mar[m]: why dont you think llvm will compile?
<Glanzmann> M1bn3mar[m]: That would be my next question, do the gentoo users want the desktop profile openrc or systemd?
<M1bn3mar[m]> i tried it before for aarch64 arch so :/
<j`ey> M1bn3mar[m]: on the m1?
riker77 has quit [Server closed connection]
riker77 has joined #asahi
<brudda[m]> Glanzmann: openrc for sure lol
<M1bn3mar[m]> yeah through qemu i dont know about native tho
<j`ey> M1bn3mar[m]: and how did it fail?
<j`ey> probably OOM?
<kov> M1bn3mar[m], why wouldn't those compile out of curiosity?
eroux has quit [Server closed connection]
<M1bn3mar[m]> openrc ofc lol :p
<kov> oh I just saw that j`ey asked the same question heh
<kov> llvm built fine here along with rust and everything on a 4k kernel (and on a 16k kernel), there's no reason it shouldn't build... I also built webkitgtk on the 16k kernel when I was testing the patch to make it work on 16k pages
<M1bn3mar[m]> i dont have the log's at hand j`ey
<j`ey> M1bn3mar[m]: it was likely the VM didnt have enough memory
phiologe has quit [Server closed connection]
<M1bn3mar[m]> Glanzmann: openrc
eroux has joined #asahi
<M1bn3mar[m]> j`ey: no i gave it 6g and it barely used it so idk
<chr> Glanzmann: Starting from Mac OS Terminal (not in rescue mode), I installed Debian using your "Soon to be released Asahi installer". Thanks for the effort.
<kov> M1bn3mar[m], unless there is something missing from your VM, I'd also bet on not enough memory like j`ey, 6g is likely not enough to build webkitgtk, you'd want a reasonably big swap there
<M1bn3mar[m]> 15g swap
<M1bn3mar[m]> kov:
<kov> k, that should be enough I think, would have to see the error message then
<j`ey> anyway it's kinda pointless talking without the details of what actually went wrong
<kov> agreed
<kov> also, just use a binary distribution, ditch gentoo to help avoid climate change /me runs
<AdryzzOLEDEdition[m]> in all fairness they did buy a power efficient machine
<kov> true
<AdryzzOLEDEdition[m]> also am i the only one who compiles everything just because i like seeing the colorful console output?
axenntio has joined #asahi
Darkmatter66 has left #asahi [#asahi]
<kov> back in my day people just liked the grey and white letters scrolling really fast, kids these days...
<M1bn3mar[m]> Adryzz (OLED Edition): yep on my part
dabaiste^ has joined #asahi
<j`ey> AdryzzOLEDEdition[m]: compile with -n to not actually run any commands :P
<kov> great idea
<AdryzzOLEDEdition[m]> i just install the -git packages
<kov> need patched ninja / make to simulate the actual run time though
Chainfire has quit [Server closed connection]
giskard has quit [Server closed connection]
chr has quit [Remote host closed the connection]
chr has joined #asahi
<Adie[m]> Installer working just fine today - thank you to all you magicians working on Asahi.
chr has quit [Remote host closed the connection]
jbennett has joined #asahi
<brudda[m]> Am I still the only one who is having issues connecting to archarm mirrors
<brudda[m]> Some packages refuse to install, giving 404s, but others install okay. I also can't access it from a browser on macos
<M1bn3mar[m]> brudda: what install did u chose minimal or desk
<brudda[m]> Minimal
<jbennett> Hey, I'm Jonathan Bennett, of Hackaday and FLOSS Weekly. We've been asked to do a FLOSS episode on Asahi, and we think it sounds great. Would one or two of the devs be interested in doing a video interview on Asahi?
<M1bn3mar[m]> k, by def the mirror's are set to japan
Chainfire has joined #asahi
<brudda[m]> Ah
<M1bn3mar[m]> so u need 2 change that
<j`ey> jbennett: marcan ^
<brudda[m]> Okay. Though trying to connect in a browser complains about an expired certificate, that expired today
THX1984 has quit [Ping timeout: 480 seconds]
<brudda[m]> Also, how would I go about changing the mirrors?
<M1bn3mar[m]> i think something like in /etc/pacman.d/mirrorlist or something like that use fm or auto completion to make sure
<clover[m]> <AdryzzOLEDEdition[m]> "is github up for you? i get a 50..." <- Guthub is having issues this morning. https://www.githubstatus.com/
<M1bn3mar[m]> idk bout the cert's but maybe u need 2 the timezone don't quote me on this one
AntoniosPapadakis[m] has joined #asahi
<marcan> j`ey: that's fixed already, it was a snafu in the original images
<marcan> I need to figure out some postinstall script to fix that for existing installs
<marcan> oh sorry
<marcan> wrong ref
<marcan> (was talking about the mirrors)
<marcan> jbennett: sure, I think I can do that :)
<jbennett> marcan: Cool! Shoot an email to ant@twit.tv to figure out the schedule.
<marcan> will do!
<marcan> if anyone needs the non-mangled default /etc/pacman.d/mirrorlist: https://mrcn.st/t/mirrorlist.orig
<marcan> I'll probably push a postinstall script at some point that will replace it with that *if* it hasn't been touched from the original japan state that the images had the first couple of days
c10l2 has joined #asahi
<Glanzmann> cbr: You're welcome.
<brudda[m]> What da ya know
<realmoot> Glanzmann: You run a 4k kernel do you mean this fixes vscodium and sublime-test/jemalloc issues?
<Glanzmann> realmoot: Yes.
<brudda[m]> pacman -Syyu fixed all my problems
<realmoot> Glanzmann: How do you get that kernel?
<brudda[m]> Probably should have tried that earlier, I guess I'm used to LTS lol
<marcan> jbennett: mail sent :)
<realmoot> marcan: About mirror issues, I have been facing constant update issues due to mirrors. WHat should I do to fix it - replace /etc/pacman.d/mirrorlist with the one you posted?
<realmoot> Just cp it?
<realmoot> Or wait for postinstall script
alyssa has joined #asahi
c10l has quit [Ping timeout: 480 seconds]
<realmoot> There's a 'mirrorlist.asahi' file with the following:
<alyssa> Silly question but is the power button the Mini supposed to do anything?
<realmoot> #
<realmoot> # Asahi Linux repository mirrorlist
<realmoot> #
<realmoot> # Worldwide CDN (bunny.net)
<realmoot> Server = https://cdn.asahilinux.org/$arch/$repo
<realmoot> # Germany
<realmoot> Server = https://de.mirror.asahilinux.org/$arch/$repo
<realmoot> # Japan
<realmoot> Server = https://jp.mirror.asahilinux.org/$arch/$repo
<alyssa> dmesg'd as an unknown SMC button event on random downstream
bstwmnt[m] has joined #asahi
<alyssa> Oh, interesting...
<tpw_rules> alyssa: it seems to shut down if you hold it for a second?
<tpw_rules> IME
<alyssa> Holding it down triggers "Power button held down", button 0xfe (= BTN_POWER_HELD1)
<alyssa> But just tapping it is unknown, button 0x01 (not 0x6 = BTN_POWER)
<alyssa> Kinda curious if I patch BTN_POWER to be 0x01 instead if I get useful behaviour on the mini
<realmoot> Sorry guys for the dumb questions... How do I fix the mirrorlist to account for unreachable mirrors?
<alyssa> right, if I patch BTN_POWER to be 0x1, tapping the power button triggers an immediate shutdown
<alyssa> it's not clear that's better :-p
<alyssa> Ooh, but in GNOME tapping it causes the power off dialog to show up which is what I was hoping for
<realmoot> That's great alyssa thanks
<alyssa> so yes, that is better (and makes the power button the mini act the way I would expect from a "normal" linux machine)
<bstwmnt[m]> I just noticed that ALARM is Arch Linux ARM. I've been wondering for 3 days what the hell ALARM stood for when I first installed Asahi Linux Minimal.
<ReaganMcFarland[m]> Some update on the WPA2 Enterprise TTLS stuff - looks like it's right when I put my username and password into the KDE connection manager is when it drops it from the drop down list
<ReaganMcFarland[m]> * Some update on the WPA2 Enterprise TTLS stuff - looks like it's right when I put my username and password into the KDE connection manager is when it no longer will show in the drop down list of available connections
Taro[m] has joined #asahi
<ReaganMcFarland[m]> And it will lose the "Connect" option when right clicking
<brudda[m]> ReaganMcFarland[m]: Is that the same as 802.1x?
<ReaganMcFarland[m]> Yes
r0ni has quit [Quit: leaving]
<brudda[m]> Okay, I'm also struggling with that atm, but I can just connect to an alternative network or mobile hotspot for now
alyssa has left #asahi [#asahi]
<ReaganMcFarland[m]> Yea my uni has guest wifi which I've been using atm
<ReaganMcFarland[m]> Trying to debug 802.1x though
<ReaganMcFarland[m]> I think network manager uses wpa_supplicant for this stuff IIRC
<ReaganMcFarland[m]> Maybe that is where it's failing
<brudda[m]> I know that iwd doesn't support it
<brudda[m]> I just checked and ours provides a python script to set the profile.
<brudda[m]> Uses network manager from the looks of it
<ReaganMcFarland[m]> I'm getting this from running "wpa_cli status"
<ReaganMcFarland[m]> wpa_cli status
<ReaganMcFarland[m]> Failed to connect to non-global ctrl_ifname: (nil) error: No such file or directory
r0ni has joined #asahi
<SamHep0803[m]> i've never gotten 802.1x to work on an arch install period
<SamHep0803[m]> whether it was asahi or just a normal intel x64 install
<SamHep0803[m]> idk maybe my school is doing some weird stuff with certificates and that
<povik> Glanzmann: any clues to the hang? can you get the kernel console on screen?
<SamHep0803[m]> but thats just my 2 cents
<ReaganMcFarland[m]> I'm not very adept in networking - i'll see if theres an open bug report and add my 2 cents
<ReaganMcFarland[m]> I'm unable to initialize the driver interface when using wpa_supplicant on the deivce
<ReaganMcFarland[m]> s/deivce/device/
nicolas17 has joined #asahi
<piroko> I've already installed asahi linux on my mba. I now want to install openbsd. I have a bootable usb openbsd installer, but I'm not sure the magic incantation required to make u-boot boot off of it
<alden4[m]> does this mean my disk is corrupted?
<piroko> yeah haha
<alden4[m]> oh no lol
<AdryzzOLEDEdition[m]> piroko: if it is for aarch54 and there are M1 drivers then yes
<piroko> Adryzz (OLED Edition): yeah it should work according to the mailinglist
<alden4[m]> does that mean i should backup my computer and reformat my drive?
<maxung[m]> alden4[m]: I had the exact same error message yesterday. Boot into recovery and in disk utility show all devices, then click on the container (important!) of your main hard drive and run first aid. afterwards reboot and it should work
<piroko> but they want to boot from the "UEFI environment only (m1n1 + U-Boot + ESP)" install option
<alden4[m]> maxung[m]: sounds good, i'll let you know if that works
<j`ey> piroko: probably easiest just to delete the current partitions and reuse the installer
<meowcat285[m]1> alden4: Did you try doing disk first aid from 1TR? That worked for me
<realmoot> Even nslookup does not work on Asahi
<realmoot> :(
<realmoot> j`ey: The issue is I can't see Asahi partitions in macos
<realmoot> diskutil list shows it
<piroko> <j`ey> "jeremy: probably easiest just to..." <- yeah but then I hvae to resize again :(
<realmoot> but then erasevolume just leaves free space which is unusable
<realmoot> Perhaps some few lines of commands to help newcomers out in uninstalling Asahi would help
<Glanzmann> povik: I just tried to reproduce it. And I noticed something. It is not alsamixer but 'f6' is sleep. So I tried to force my system into sleep. Nothing to do with the sound. Sorry for the false alert.
<realmoot> Can anyone read my messages?
<realmoot> I mean are they visible
<j`ey> yes
<realmoot> Ok
<realmoot> I had used ipsw file and Apple Configurator 2 to restore mac and nuke the drive to reinstall Asahi
<realmoot> Because the partition scheme is weird
<Glanzmann> realmoot: Yes.
<realmoot> Will keeping 16K pagesize really put pressure on app developers to build for M1?
<realmoot> I feel that M1 is a small segment while arm SBCs are larger
<AdryzzOLEDEdition[m]> the more pressure the better
<povik> Glanzmann: :D
<povik> well good then
axenntio has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<j`ey> realmoot: better now than later
m2231puppy has joined #asahi
<piroko> hooray I managed to delete the volume (creating free space) that the installer can now use to re-install
frog_ has joined #asahi
<bstwmnt[m]> now you can do a tiling wm + astolfo wallpaper for the real arch linux experience
<gsora[m]> Don’t forget neofetch!
<bstwmnt[m]> yeah that too
<hello9999901[m]> lol
<frog_> true
<bstwmnt[m]> i was sad that i3wm wasnt built for arm and wayland but i found another friend
<bstwmnt[m]> sway
<gsora[m]> Best wayland experience ever with wayland
<frog_> have any of you tested asahi yet?
<gsora[m]> Sway can even enable freesync if your monitor supports it!
<hello9999901[m]> what are the technical limitations for bringing windows to m1? apart from drivers and proprietary code (which are huge issues in and of itself) :)
<gsora[m]> Will :v
<bstwmnt[m]> frog_: yeah, tried to use kde first with minimal install and forgot to install a terminal
<frog_> how is it
<bstwmnt[m]> not fun. do recommended install
<frog_> alright
<hello9999901[m]> gsora[m]: if that's replying to me then LOL ahaha
<hello9999901[m]> i must agree
<gsora[m]> It is :)
<frog_> have been keeping my system on 12.0.1 so will have to leave for 12.3 im assuming?
<hello9999901[m]> so true so true
<tpw_rules> frog_: yes
<frog_> alright
<frog_> im about to restart to update
<frog_> better be worth linux
<piroko> frog_: yep!
<piroko> runs supremely well
<bstwmnt[m]> yeah other than the 3 hour issue of trying to fix my internet because i thought i was too good for the minimal installer, it was surprisingly good.
<frog_> nice, alright
<bstwmnt[m]> general performance is nice even without gpu acceleration
<frog_> ill see ya guys later ig once i update
<bstwmnt[m]> parallel downloads are wack
<frog_> is it worth taking a backup of macos before installing asahi
<j`ey> yes if you have important data
<brudda[m]> I did
<frog_> just home folder?
<brudda[m]> I just did a time machine backup
<tpw_rules> run the first aid in disk utility before upgrading too
<frog_> ok, why though
<brudda[m]> But it will automatically make an invisible local one even if you use an external drivr
<brudda[m]> If you use time machine
<tpw_rules> a few people have reported apfs corruption. the resize the installer needs to do will fail if it's detected
<brudda[m]> So you need delete that before partitioning, otherwise it will just fail
<tpw_rules> and in bad cases you'll need to reformat your disk to get rid of it. easier to do that before the update
<frog_> ah, ok
<realmoot> In the default asahi desktop install, is .profile missing for anyone else?
thasti has quit [Remote host closed the connection]
the_lanetly_052 has quit [Ping timeout: 480 seconds]
thasti has joined #asahi
m2231puppy has quit [Quit: m2231puppy]
Spireshield[m] has joined #asahi
<frog_> i verified my container
<frog_> am i safe to update
<tpw_rules> if it said everything is OK, then yeah
<frog_> ok, thanks
<frog_> gooodbye
frog_ has quit [Quit: Page closed]
m6wiq has joined #asahi
<lkvrsfld[m]> lel
<lkvrsfld[m]> archlinuxarm.com certificate isn't valid
<lkvrsfld[m]> s/com/org/
<bstwmnt[m]> yeah i just noticed that earlier
<brudda[m]> Okay glad I wasn't the only one with that issue
<lkvrsfld[m]> I'll try generating one, including marcans repository, and the needed packages, just to use archinstall script
<lkvrsfld[m]> is there any interest having an arch iso (boot drive) including the archinstall script?
realmoot has quit [Quit: Konversation terminated!]
m2231puppy has joined #asahi
m2231puppy has left #asahi [#asahi]
m2231puppy has joined #asahi
<mps> this gives me idea, I could (maybe) create alpine iso to simplify installation for those who want to install alpine
<mps> enchiladasconpinguino[m]: what is you opinion about this
<dan[m]123> > <@prohackerbro:matrix.org> is there any interest having an arch iso (boot drive) including the archinstall script?
<dan[m]123> Yes
<dan[m]123> > I'll try generating one, including marcans repository, and the needed packages, just to use archinstall script
<alden4[m]> <maxung[m]> "I had the exact same error..." <- ty so much, that worked, for anyone following along at home the container is hidden, you need to enable it
<lkvrsfld[m]> <mps> "this gives me idea, I could (..." <- regarding such kind of things, maybe we should create them on github/gitlab and build a github action (or corresponding) to keep the image up to date
LuigyLeon[m] is now known as luigy[m]
axenntio has joined #asahi
<mps> lkvrsfld[m]: I doubt this is possible for alpine because it has specific scripts to create iso images
axyliah[m] has joined #asahi
<mps> j`ey: here is result of min1 build https://tpaste.us/gZMr
<lkvrsfld[m]> mps: because of licence issuues or why?
<j`ey> mps: it built!
<mps> j`ey: so looks like it works but have to find how to add submodules in APKBUILD
<mps> lkvrsfld[m]: no license but how to have environment for scripts to work
<maxung[m]> <alden4[m]> "ty so much, that worked, for..." <- np ^^
<lkvrsfld[m]> mps: docker
<lkvrsfld[m]> docker has qemu-user-static support
<mps> lkvrsfld[m]: I'm not expert in this but afaik it will not work in docker
Tano70 has quit [Quit: Leaving]
learning has joined #asahi
<mps> lkvrsfld[m]: I had to make serious hacks just to make iso with specific kernel for armv7
<lkvrsfld[m]> mps: hmm okay i understand
<mps> and alpine is mostly used in dockers, fun part :)
<mps> actually docker company long time sponsors alpine
<lkvrsfld[m]> alpine is great as a base
<mps> sorry, my opinion is not 'great' but better than other distros I tested
<lkvrsfld[m]> most images work off alpine, or have alpine-versions (ofc only if musl isn't any problem for the software though)
<axyliah[m]> hi, I apologize in advance if this question is upsetting anyone - I am really excited over the hardware of M1, and the idea of being able to install linux on such a great hardware makes me really interested in getting a macbook. I have never personally owned a laptop or even a mac at all, and I cannot decide whether to get MacBook Air or Macbook Pro. How did you decide on which machine to get?
<mps> lkvrsfld[m]: I know (I'm alpine developer) but nowadays alpine also becoming not good as it was earlier
<lkvrsfld[m]> axyliah[m]: talking of the Air and Pro 13, i would personally choose the Air. it comes with almost the same performance (for me even identical, since having modded the internals), has no touchbar (which isn't even supported on linux yet) and just looks + feels better (and is way cheaper)
<j`ey> the macbook pro 14 is more expensive, doesnt have a touchbar and is m1pro, so better than m1
<mps> mb pro 13 has fans in case temp goes high
<axyliah[m]> thinking between M1 (Air) and M1 Pro 14", planning to keep it for as long as possible. Air feels like the best bang for the money, but the base configuration only has 8GB RAM, which worries me in terms of longevity
<mps> I thought that toucbar missing will be big problem but after I made hack touchbar driver is not needed for me
<lkvrsfld[m]> mps: i have a co-worker which runs alpine bare-metal and solves software problems (as an example, MS Teams) with https://github.com/mviereck/x11docker. He is having a NUC though, don't know how it would work on a mobile platform. arch is pretty near to alpine regarding size and flexibility (and their repos), so why don't use just that
<j`ey> upgrading the air to 16G is still quite a bit cheaper than the base mbp 14
<lkvrsfld[m]> mps: what would be the reason of it getting not goot?
<lkvrsfld[m]> s/goot/good/
<mps> lkvrsfld[m]: nowadays people started to add bloat and not needed dependencies to packages
<lkvrsfld[m]> mps: hmm okay that's always bad
<mps> lkvrsfld[m]: on mobile platform, there is postmarketOS which is alpine based
<lkvrsfld[m]> mps: my bad, tanking about mobile "laptop" platform
<lkvrsfld[m]> s/tanking/talking/
<mps> lkvrsfld[m]: I run alpine on old and small SBCs (n900 256MB ram) and big servers and everything in between
learning_ has joined #asahi
learning_ has quit []
<lkvrsfld[m]> mps: loved the fact i had to "learn" openrc wanting to run my personal hostings / services on alpine
<lkvrsfld[m]> it made me understand that systemd is really just less advantages than disadvantages
<mps> lkvrsfld[m]: heh, I advocated on debian-devel ML switching to systemd, shame on me :)
<mps> I was early systemd adopter, first versions, but after time I started to hate^Wdislike it
<lkvrsfld[m]> mps: normal debian without systemd... what a dream
<lkvrsfld[m]> well for a common distro like debian, systemd is pretty cool
<mps> lkvrsfld[m]: debian could be set to use openrc
<mps> it is not complicated
<lkvrsfld[m]> mps: isnt there devuan having openrc?
<lkvrsfld[m]> does it play well with general desktop packages?=
<lkvrsfld[m]> mps: very nice
<mps> lkvrsfld[m]: yes, devuan has it and also systemV rc
<mps> about two (or 3) years ago I used openrc on few servers and that was easy to set, but after that time I converted these servers to alpine
<mps> so don't know current state
mnc7[m] has joined #asahi
<lkvrsfld[m]> i love the fact you can add any func to the services - rc-service "service" goat
<mps> talking about all good and bad points of distros all they are good at the end, we are happy to have them and not be locked to proprietary systMESS
<mps> s/systМЕSS/systems/ :)
<lkvrsfld[m]> MESS just nails it xD
learning has quit [Quit: Page closed]
m6wiq has quit []
pbrobinson has quit [Ping timeout: 480 seconds]
pbrobinson has joined #asahi
kita99 has joined #asahi
m6wiq has joined #asahi
PaterTemporalis has quit [Ping timeout: 480 seconds]
joske has quit [Remote host closed the connection]
<piroko> successfully got OpenBSD installed and running on the m1
<piroko> it was.... super tiny (max resolution) and graphics performance was quite a bit slower than linux
<piroko> and the default trackpad acceleration curve on OpenBSD never feels "right" to me
<piroko> even after tweaking
<piroko> BUT other than that, it runs!
<kettenis> \8/
<kettenis> piroko: what type of screen are you using?
<kettenis> console font on the 14" macbook pro is quite ok for me
<hellosway[m]> <j`ey> "upgrading the air to 16G is..." <- the ram is not the only difference
axenntio has quit [Quit: Axenntio]
Darktux has joined #asahi
<j`ey> hellosway[m]: I know, they just mentioned the base air only having 8G
<Darktux> Hi! just testing Asahi on my MbP M1. How fast it is!
<aem64[m]> video codec and 3d acceleration are my top wanted feature
m6wiq has quit []
<Darktux> I can't imagine how fast these machine will be once everything work flawlessly.
<Tramtrist> Of course I'm wrong about this but.. if Apple actively supported Linux.. they'd gain a lot of good will and sales
<Tramtrist> At least... from me 🤣
Darktux has quit [Quit: Konversation terminated!]
<Tramtrist> Would be nice if they threw alyssa a bone on the GPU documentation
Darktux has joined #asahi
<aem64[m]> it's be nice to have a guide on setting up box64 on here (since the pagesize is 16k) at least i think thats the problem
<Darktux> Indeed, same from me.
Darktux has quit []
kelito has joined #asahi
<kelito> s/Dartkux/kelito/g
<aem64[m]> do printer drivers work?
<nicolas17> I don't see why not, if it's just USB2
<Glanzmann> j`ey: Are their any news about the new macbook air?
<nicolas17> I have a raspberry pi 3 as print server, with cups+gutenprint, it works better than Windows (because EPSON never released 64-bit drivers for this old printer so Windows just gets a dumb generic driver now)
<j`ey> Glanzmann: dont think so
<Glanzmann> I see, thanks.
<piroko> kettenis: 14" air. console was great. X was... max resolution haha
<piroko> oh, also, the installer assumes sd0l for the EFI partition (to snag the wifi firmware from), but my partition was sd0m. most likely a result of me not completely wiping all the old archlinux volumes before doing a re-install
<piroko> point being, I'm not sure if sd0l can always be relied on. but not sure of a reliable way to script it either. I guess if you're savvy enough to mess with the partitions, you're savvy enough to find it and manually do the firmware install
m2231puppy has quit [Quit: m2231puppy]
m2231puppy has joined #asahi
<kettenis> piroko: yes, a classic X desktop doesn't work very well on a retina display ;)
<kettenis> and yes I realize the hardcoded sd0l isn't optimal
<hellosway[m]> wayland works, I have gnome running on a normal intel mac, and it looks fine
<hellosway[m]> m1 macs have the same display
<tpw_rules> piroko: can bsd easily read the device tree? under linux the GPT ESP UUID is in /proc/device-tree/chosen/asahi,efi-system-partition
m2231puppy has quit [Quit: m2231puppy]
<aem64[m]> will there be a way for transfering files between macOS and linux?
<aem64[m]> also can i make caps lock function as lcontrol
<ChaosPrincess> make a new partition and format it as fat32
kelito has quit [Quit: Konversation terminated!]
<kettenis> tpw_rules: there is no easy way to translate a GPT ESP UUID into a BSD disklabel partition
<kettenis> and at this point I have no intention to write code to support that
<kettenis> one of the reason I'm not fond of marcan's "multiple ESP" approach
<piroko> anyway, my goal was not to point out the only the negative things. I'm floored by the fact that this thing can run openbsd
<piroko> seriously impressive work :)
<kettenis> no worries
<tpw_rules> kettenis: i have to be honest i have no idea what you mean. is there no /dev/disk/by-partuuid equivalent?
<kettenis> tpw_rules: no there isn't
<tpw_rules> ok
<gsora[m]> jeremy: same! Fastest OpenBSD laptop in the west
<gsora[m]> Weirdly enough jcs didn’t post an article yet
<meowcat285[m]1> Shame I will have to uninstall asahi later today
<meowcat285[m]1> I need to space for movies for a trip
<meowcat285[m]1> I am sad :(
<gsora[m]> Oh no! Watch movies from Linux!
<meowcat285[m]1> gsora[m]: All of my movies are are on itunes sadly
<Tramtrist> Tis a tale of woe
Tano70 has joined #asahi
yrz[m] has joined #asahi
<Sobek[m]> By the way, speaking of the GPU side of things, I’ve seen stuff about OpenGL, but is Vulkan also on the roadmap ? High performance nowadays tend to also involve vulkan (and then stuff like DxVk are built on top of it). (@alyssa ?)
<aem64[m]> ooh vulkan support would be amazing
<aem64[m]> considering macos doesn't support it
<gsora[m]> I’ll be happy with a functional compositor, I’ll be honest
<Sobek[m]> aem64[m]: (Hmm MoltenVk is a thing on macOS, I think the Vulcan tutorial more or less targets macOS using it as part if the SDK, and there are perhaps 1-2 obscure corners that don’t work but otherwise it works well)
<aem64[m]> Sobek[m]: afaik moltenvk doesn't work as well as native vulkan
<aem64[m]> because isn't it translating it to metal?
<clover[m]> is this manjaro user reporting the same audio playback bug that is in asahi (where you have to switch profiles before sound plays when streaming music, through headphones) https://forum.manjaro.org/t/pulseaudio-restart-only-works-with-music-inside-browser/96608
JamesReilly[m] has left #asahi [#asahi]
AndrewGlaze[m] has joined #asahi
AndrewGlaze[m] is now known as Candygoblen123[m]
Tano70 has quit [Quit: Leaving]
<aem64[m]> how often will asahi get updates?
<j`ey> when new features are ready
<j`ey> (or theirs a rebase to newer linux versions, but I suspect that wont be that often)
<Candygoblen123[m]> I saw an update today so it seems to be pretty frequent
<j`ey> wasnt a new kernel though, just an issue with the kernel headers package
justashadow[m] has joined #asahi
crw has joined #asahi
<crw> Quick question, If I install the Arch minimal install, what's the wifi setup situation look like?
<crw> Just normal arch wifi setup? or will there be complications
<mps> crw: look if iwd is installed is installed by default
MaxNurzia[m] has joined #asahi
kelito has joined #asahi
<kelito> Back!
crw has quit [Remote host closed the connection]
kelito has quit [Remote host closed the connection]
<M1bn3mar[m]> hey Glanzmann u sead that u r gonna make a gentoo installer no??
eroux has quit [Ping timeout: 480 seconds]
<enchiladasconpinguino[m]> <mps> "this gives me idea, I could (..." <- great idea, new install I did with official installer is working great
<enchiladasconpinguino[m]> only needs your linux-asahi.apk and small tweak to grub
<mps> enchiladasconpinguino[m]: what is tweak to grub?
<mps> and I look if I could clean linux-asahi APKBUILD and merge it tomorrow
<mps> also I got m1n1 APKBUILD with rust to work, hackish though but works
<enchiladasconpinguino[m]> explicit "set root", add "rootwait", and use PARTUUID to find root
<enchiladasconpinguino[m]> not too many changes for grub
<matthewayers[m]> j`ey: Do you know how I'm supposed to update the Debian build that Glanzmann created? Or should I start fresh with a new partition with the alpha release? I am using the kernel build from 2022-03-10
eroux has joined #asahi
<matthewayers[m]> * j`ey: Do you know how I'm supposed to update the Debian build that Glanzmann created? Or should I start fresh with a new partition with the alpha release? I am using the kernel build from 2022-03-10
branon has left #asahi [WeeChat 3.0]
<mps> enchiladasconpinguino[m]: `set root='hd0,gpt2'` is not needed in my tests but ok
<j`ey> matthewayers[m]: he has a script you can run to update the kernel, but otherwise I assume apt update as per usual
<mps> enchiladasconpinguino[m]: also your tweaks are for btrfs, and I think default fs should be ext4
<enchiladasconpinguino[m]> mps: hrmm maybe just the other two fixes is enough, not sure yeah
<matthewayers[m]> Gotcha. I ran it and I'm still on the same kernel version even after restart... hmm
<enchiladasconpinguino[m]> no tweaks made for btrfs support, just standard installer
<j`ey> matthewayers[m]: then no new kernel!
<enchiladasconpinguino[m]> didn't do a manual installation at all
<matthewayers[m]> Lol
<j`ey> matthewayers[m]: latests is 20220310
<mps> enchiladasconpinguino[m]: and please do not use 'quiet' in kernel cmdline, alpine is not ubuntu ;)
<enchiladasconpinguino[m]> lol its just what the installer does by default, didn't add that
<mps> enchiladasconpinguino[m]: also, modules are just for examples if someone wants to change/add some
<enchiladasconpinguino[m]> usual `# setup-alpine` from .iso is all
efegurkan[m] has joined #asahi
sheepgoose_ has joined #asahi
<mps> enchiladasconpinguino[m]: yes, I know this about 'quiet' even raised this to alpine TSC but they voted to keep it. now you know why I'm loosing good thinking about alpine :)
<enchiladasconpinguino[m]> this grub config might be "wrong" in multiple ways, no argument from me..
<enchiladasconpinguino[m]> I simply was trying to make the least amount of required changes from standard installer for it to still boot
<mps> I prefer to see kernel messages on console early
<mps> enchiladasconpinguino[m]: anyway, thank you for testing and fixes
<enchiladasconpinguino[m]> of course, thanks for kernel work ;)
<enchiladasconpinguino[m]> nothing actually wrong with your config for dev mapper or encryption
<enchiladasconpinguino[m]> and I went through the kernel config and compared to stock lts
<enchiladasconpinguino[m]> initramfs mounting of FDE root is complex, will keep investigating this
realmoot has joined #asahi
<brudda[m]> has anyone made an ASCII for the asahi Linux logo for that beautiful neofetch output? I know technically its just alarm by default, but I think the fact its on an M1 deserves that much at least
<enchiladasconpinguino[m]> I don't think so yet, but def need that lol
<realmoot> ^
<enchiladasconpinguino[m]> other arch flavors are detected
<realmoot> Asahi boots so damn quick im impressed
<brudda[m]> it does
<brudda[m]> Way faster than macos lol
<realmoot> yeah
<realmoot> I ran geekbench and I get 1700/7000ish which is amazing tbh
<j`ey> iBoot still feels quite slow
<brudda[m]> I might play around with an ASCII logo lol
pzcz has joined #asahi
<mps> j`ey: https://tpaste.us/Oq1m this is m1n1 APKBUILD which works
pzcz has quit [Remote host closed the connection]
<realmoot> BTW I never realized the pre-bundled text editor KATE exists so I'm using that and not missing sublime so far...
badlydrawnface[m] has joined #asahi
<gsora[m]> A kernel dev friend swears by it
<nicolas17> p sure _marcan uses Kate for everything, I've seen it on his asahi coding streams
<j`ey> mps: oh you retarballed it?
<mps> j`ey: yes, have no better idea
THX1984 has joined #asahi
<badlydrawnface[m]> the website faq for uninstallation says to simply remove the modified partitions
<badlydrawnface[m]> however you need to refer to startup disk to change the boot order, otherwise you will brick
<j`ey> mps: listing them all out :/
<badlydrawnface[m]> * boot order before deletion, otherwise
<j`ey> nicolas17: yep he does, and people always ask why he isnt using vscode or something
<mps> j`ey: this is just POC, not final solution
<gsora[m]> If Kate had LSP support, I would try it
<badlydrawnface[m]> j`ey: bc vscode isn't working bc electron isn't working bc chromium doesn't support 16k page sizes
<badlydrawnface[m]> i think
<j`ey> badlydrawnface[m]: this is on his current main computer, which is x86
<gsora[m]> Last time I used it was in… kde 3.5 probably
<badlydrawnface[m]> oh pepega
<enchiladasconpinguino[m]> using vscode right with 4k pages
<badlydrawnface[m]> sorry
<enchiladasconpinguino[m]> could also make a vm with 4k pages from stock asashi install, in the short term
<enchiladasconpinguino[m]> * vscode right now with 4k
<marcan> brudda[m]: trick_?
<marcan> er
<marcan> badlydrawnface[m]: brick?
<badlydrawnface[m]> what?
<marcan> when you delete the default boot OS it tries one boot, fails, falls back to boot recovery, and asks you to pick another startup disk
<marcan> I don't see it bricking
<badlydrawnface[m]> odd
<badlydrawnface[m]> i might have deleted too many partitions
<badlydrawnface[m]> macos's partition style is very foreign to me
<badlydrawnface[m]> i unbricked it btw
<marcan> if you deleted the recovery partition that would be why
<marcan> the last one
<marcan> that one is system recovery
<marcan> never touch that one :)
<badlydrawnface[m]> note to self: NEVER, EVER touch disk0s1 through disk0s3
<badlydrawnface[m]> disk0 is the equivalent to sda/nvme0n1 on mac it seems
<marcan> yes
<badlydrawnface[m]> only delete disk0s4 and disk0s5
<badlydrawnface[m]> what about disk2?
<realmoot> But last time I bricked my install
<marcan> those are virtual disks because APFS is weird
<badlydrawnface[m]> marcan: yeah 🥴
eroux_ has joined #asahi
<marcan> you also shouldn't be thinking in terms of numbers
<realmoot> Not really sure which volumes to delete because there's one called 'EFI-EFI Asahi' so I left that
<badlydrawnface[m]> will the installer deal with it, or create duplicates?
<marcan> because macOS partition numbers do *not* correspond to partition table indexes
<marcan> they are allocated on demand
<marcan> and can be out of order following disk ops
<badlydrawnface[m]> which is why i see something about a "cleanup script" in the blog
<badlydrawnface[m]> does it fix that?
<marcan> if you delete the disk0sX container the corresponding diskY virtual disk will go away with it
<badlydrawnface[m]> ohhhhh
<badlydrawnface[m]> got it
<marcan> it's like apfs pretends each container is another disk with more partitions that are actually volumes, but it's not even real partitions, the whole system is weird
eroux has quit [Ping timeout: 480 seconds]
<marcan> you want to delete disk0s3, 4, 5
<marcan> disk0s3 == disk2
<marcan> you can also just delete disk2, it is equivalent to disk0s3 for the purposes of `diskutil apfs deleteContainer`
<marcan> see how it says "Physical Store disk0s3"
<marcan> either way the cleanup script is https://alx.sh/wipe-linux but it's pretty extreme, should work fine for simple cases like yours
<marcan> I need to get some sleep, but I hope that made some sense :)
<j`ey> marcan: i naively assumed you had just woken up :P
<nicolas17> yeah I haven't figured out his timezone yet
<badlydrawnface[m]> 👍️
<badlydrawnface[m]> th
<badlydrawnface[m]> * thx
<badlydrawnface[m]> disk2 is a container
<badlydrawnface[m]> needed to use `deleteContainer` instead of `deleteVolume`