<kallisti>
I seem to have `LIBGL_ALWAYS_SOFTWARE=1` set across all my shells / terminals. I have been grepping around the filesystem but I am not sure where this is being set. Is there something I'm missing? I think I checked the "usual places" `~/.zshrc` `~/.profile` `~/.bashrc` `~/.bash_profile` `~/.bash_rc` `/etc/profile.d/*` `/etc/profile` , etc.
SoniEx2 has quit [Ping timeout: 480 seconds]
<psykose>
grep -r LIBGL_ALWAYS_SOFTWARE /
<kallisti>
I was trying something like that with grep -rni but let me try again...
<kallisti>
Yeah not sure regarding the segfault when greping via /. But ended up finding it in /etc/profile.d/ . Not sure why I didn't find it there before. Big shame on family...
<kallisti>
brb
kallisti has quit []
kallisti has joined #asahi
kallisti has quit []
ten1572371 has joined #asahi
kallisti has joined #asahi
Soni has joined #asahi
<kallisti>
Another Asahi question. Does Asahi with the plasma desktop use SDDM as the login manager? I can't see how different desktop envs are configured via SDDM. I am trying to figure out how to set environment vars for each respective desktop environment
Zopolis4 has joined #asahi
ten157237 has quit [Ping timeout: 480 seconds]
ten1572371 is now known as ten157237
kallisti has quit [Remote host closed the connection]
jacksonchen666 has quit [Remote host closed the connection]
jacksonchen666 has joined #asahi
kallisti has joined #asahi
jacksonchen666 has quit [Remote host closed the connection]
jacksonchen666 has joined #asahi
kallisti has quit [Remote host closed the connection]
chadmed_ has quit [Remote host closed the connection]
thansen has joined #asahi
chadmed_ has joined #asahi
<chadmed_>
with a slightly lower capacity factor on the ecores (based on coremarks/mhz) reniceing pipewire to -11 fixes things mostly
<chadmed_>
now to make pipewire stop trying to rely on dbus rtkit
cylm_ has joined #asahi
<psykose>
in relatively sure the nice value doesn't do anything at all there
<chadmed_>
well the problem at the moment is more that threads are stalling waiting for time on the ecores
<chadmed_>
thats not really noticeable for background tasks/whatever kernel workers are doing, but with pipewire it causes audible stalling
<chadmed_>
thats far more noticeable than baloo taking a bit longer to index files or whatever, so we tell the scheduler to make that stuff wait for pipewire
<chadmed_>
ultimately the scheduler is still broken, if the ecores are running into stalls that frequently it should be moving stuff like pipewire off them
<chadmed_>
i guess it just doesnt really know how to deal with a soc thats asymmetrical in the opposite direction to whats usually found on these types of things. having so many more pcores than ecores is kind of unusual. i suspect it will work far better on a base m1
as400 has quit [Remote host closed the connection]
<chadmed_>
pipewire tries to set itself to -11 niceness by default anyway, its just failing on my machine for some reason
as400 has joined #asahi
<bluetail07>
is 'nvram' available on asahi? Wanna try to pass something that is on mac on linux. "sudo nvram boot-args="enableDither=0 PE_i_can_has_debugger=1 iomfb_RuntimeProperty_enableDither=0""
<j`ey>
bluetail07: not currently, but it should be from 1TR
<sven>
what are you trying to do? those arguments sound like something that XNU (or one of its kexts) parse
<_jannau_>
yes, linux/the dcp driver dowesn't care about those
<bluetail07>
I am almost certain its the cause and I know its discussed to death already
<sven>
i'm not sure what that is but it sounds like you'd need to change the linux dcp driver to set "enableDither"
<bluetail07>
shit
<sven>
linux doesn't read nvram at all
<bluetail07>
dcp is not gpu right? but display co processor
<j`ey>
yes
<_jannau_>
it will be useful to identify runtime propertiees (they are numerical on the iomfb FW interface)
<sven>
there's a small chance iBoot would read that but I think that's unlikely because this sounds like a property of the iomfb endpoint which iboot doesn't use
<bluetail07>
yea I have this very annoying issue where I cannot use my external monitors cause it dithers that bad and its seizure causing
<sven>
and PE_i_can_has_debugger is entirely unrelated to dcp
<bluetail07>
Can I temporarily save into nvram with it having persistency from mac?
<bluetail07>
(I ask cause one can clear nvram)
<sven>
that kinda defeats the purpose of non-volatile ram
chadmed_ has quit [Remote host closed the connection]
<sven>
you can set it and then clear it again I guess
<_jannau_>
supporting this in the dcp driver as such is not an issue but figuring out how to expose that to user space is. I don't want to add module parameters for that
<sven>
but you'd need to reboot after changing the value to see its effect anyway
<bluetail07>
_jannau_ is there hope in the future to get it or not? I cannot figure. And I hope I dont sound demanding. It's just the one thing that I miss at the moment
<j`ey>
bluetail07: you should probably see with macos if it actually fixes your isue
<JTL>
^
<bluetail07>
I will try
<JTL>
If certain parameters work to fix issues under macos, the thing to do is investigate what it would take to implement them under asahi drivers
<bluetail07>
I will try now, meaning I will go offline in IRC till I am done
bluetail2 has quit [Remote host closed the connection]
bluetail2 has joined #asahi
MajorBiscuit has quit [Quit: WeeChat 3.6]
bluetail2 has quit []
bluetail has joined #asahi
MajorBiscuit has joined #asahi
<bluetail>
The parameter doesnt fix the issue on Linux after applying in macOS. Under macOS, the issue is almost not noticeable too, no matter if dithering is on or off ! I now think something else is different and I do not know how to compare it. I even made sure both use the same color profile
flying_sausages has quit []
MajorBiscuit has quit [Quit: WeeChat 3.6]
hertz has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
MajorBiscuit has joined #asahi
<bluetail>
I kinda give up at this point. Many people have this issue, and its temporary in nature on macOS, but seems permanent on linux. Perhaps this is not fixable, especially since its so long persisting and other people have the same issue. But other people do not have it at all. I gotta do something else now.
<bluetail>
might as well be a widespread defect
flying_sausages has joined #asahi
possiblemeatball has joined #asahi
jacksonchen666 has quit [Remote host closed the connection]
Guest815 has quit [Quit: G-line: User has been permanently banned from this network.]
vx has joined #asahi
vx is now known as Guest846
roxfan2 is now known as roxfan
ktz has quit [Quit: WeeChat 3.7.1]
jamespmorgan has quit [Remote host closed the connection]
jamespmorgan has joined #asahi
<thansen>
just trying this out on a 2021 mbp 14 and installer seems to have gone fine but the keyboard/touchpad don't work, is that expected?
<opticron>
I have a mbp 14" and my touchpad and keyboard work fine. Do you have a USB keyboard you can use to get to a console and look at the kernel log? Someone else may have better suggestions for things to look at
<thansen>
yeah no problem with that, I can console etc
<thansen>
fully updated already with pacman
<j`ey>
so they keyboard works in the console but not X/wayland or?
<thansen>
no, I have external kb/mouse that work, the built-in does not work anywhere
<j`ey>
then I guess if you can post your dmesg somewhere
<thansen>
the macos install is still on 12.6...would that possibly cause the issue?
<j`ey>
no it shouldn't
<j`ey>
you can also run the asahi-diagnose script and upload the results
<thansen>
which do you prefer?
<j`ey>
both
Zopolis4 has quit [Quit: Connection closed for inactivity]
<jannau>
thansen: how often have you rebooted/powered off the macbook? there is an issue there the keyboard/touchpad driver occasionally fails to probe
<thansen>
the laptop hasn't been used much honestly, only powered off/rebooted the last few minutes from doing the install
<jannau>
that might be fixed but is not yet in the shipped kernel
<j`ey>
spi-hid-apple-of spi0.0: waiting for device info failed: -19
<thansen>
to try and get around the probe issue do I just 'turn if off and on again'?
<j`ey>
yeah
<jannau>
yes, I hit that approximately 1 in 25 boots but it is sometimes sticky over reboots
<thansen>
so do a full power off?
<jannau>
yes
<thansen>
yeah, seems to be failing every time in my case :(
<jannau>
thansen: if you're comfortable with it please try changing the repository from "asahi" to asahi-dev" in /etc/pacman.conf and update
<jannau>
the kernel in that repository already has the fix I was speaking about
<thansen>
easier to do that or just install linux-asahi-edge package?
<j`ey>
edge wont fix it
<j`ey>
edge is the same source code as the non-edge version, with diffrent kernel options
<thansen>
ah ok
<thansen>
I don't see asahi-dev in pacman.conf
julio7359 has joined #asahi
<j`ey>
you change asahi to asahi-dev
<jannau>
you have to change '[asahi]' to '[asahi-dev]'
<thansen>
ah ok
<thansen>
moment then...
<thansen>
just the kernel package is the only difference?
<jannau>
I think so, maybe a never mesa-asahi-edge as well
<thansen>
ok cool, works!
<thansen>
should I attempt the accelerated gpu driver/wayland or not yet?
WindowPa- has joined #asahi
WindowPain has quit [Ping timeout: 480 seconds]
___nick___ has joined #asahi
JoshuaAshton has quit []
___nick___ has quit []
JoshuaAshton has joined #asahi
___nick___ has joined #asahi
JoshuaAshton has quit []
JoshuaAshton has joined #asahi
jess has quit [Quit: brb]
paolo_kuk__ has joined #asahi
landscape15 has joined #asahi
<landscape15>
Hi. How can I tell which Recovery OS I booted into from the terminal?
<landscape15>
I mounted all the APFS Containers but it seems like the file system is a virtual disk5 and I don't know how to find it's physical apfs container
hertz has joined #asahi
hertz has quit []
landscape15 has quit [Remote host closed the connection]
hertz has joined #asahi
ten157237 has quit [Remote host closed the connection]
ten157237 has joined #asahi
tom has joined #asahi
StupidYui has joined #asahi
<tom>
Hello!
jamespmorgan has quit [Remote host closed the connection]
creechy has quit [Quit: ZNC 1.7.5+deb4 - https://znc.in]
creechy has joined #asahi
___nick___ has quit [Ping timeout: 480 seconds]
chadmed_ has quit [Read error: No route to host]
jess has joined #asahi
lewurm has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
lewurm has joined #asahi
vmeson has quit [Quit: Konversation terminated!]
vmeson has joined #asahi
lewurm has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
lewurm has joined #asahi
Treibholz has quit [Quit: WeeChat 3.7.1]
Treibholz has joined #asahi
tom_ has joined #asahi
tom has quit [Ping timeout: 480 seconds]
Guest846 has quit [Quit: G-line: User has been permanently banned from this network.]
mini0n has quit [Ping timeout: 480 seconds]
vx has joined #asahi
vx is now known as Guest887
hightower3 has joined #asahi
hightower2 has quit [Ping timeout: 480 seconds]
SSJ_GZ has quit [Ping timeout: 480 seconds]
tom_ has quit [Read error: Connection reset by peer]
Ariadne_ is now known as Ariadne
kallisti has joined #asahi
<kallisti>
Anyone know what boot params to set to boot into a console and disable SDDM / etc? I broke my install setting the mesa package on edge as someone mentioned on reddit that would speed up the jetbrains ides
<kallisti>
Otherwise not having luck using key sequences to get a console but def can't see SDDM.
<thansen>
just `s` will make systemd boot to single user mode
<kallisti>
thansen: so edit the kernel boot params and put an s at the end?
<thansen>
sure, or the front
<kallisti>
thansen: thanks!
<thansen>
yup
<kallisti>
will give it a shot.
kallisti has quit [Quit: brb]
<thansen>
you'll need the root password though
<thansen>
I think..
kallisti has joined #asahi
<kallisti>
thansen: I get a message "Can't open access to console, the root account is locked" . Then it looks like I need to try this some other way
<thansen>
ah, not sure what to tell you then :(
<kallisti>
mps previously shared a way for me to mount the disk using qemu and alpine. I can mount the disk but I don't have the network setup. Man this really put me back...
<kallisti>
not sure we actually tried using the chroot command...
<kallisti>
looks like i got the network setup in alpine. Is it possible to chroot asahi and use pacman from the chroot to fix the broken mesa? Any other tricks to get a console login shell and skip SDDM?