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-alt #asahi-stream #asahi-offtopic | Keep things on topic | Logs: https://alx.sh/l/asahi
gladiac has quit [Quit: Ping timeout (120 seconds)]
gladiac has joined #asahi
Catyre has joined #asahi
yrlf has quit [Quit: Ping timeout (120 seconds)]
yrlf has joined #asahi
doggkruse has joined #asahi
Catyre has quit [Ping timeout: 480 seconds]
doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
hays has quit []
hays has joined #asahi
Catyre has joined #asahi
Catyre has quit [Remote host closed the connection]
kaseyreed has quit [Quit: Konversation terminated!]
Catyre has joined #asahi
<Glanzmann> Debian users: I updated the kernel with the most recent asahi branch. Apparently it includes a bluetooth hang on boot from upstream patch. curl -sL https://tg.st/u/ksh | bash
marvin24_ has joined #asahi
marvin24 has quit [Ping timeout: 480 seconds]
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi
PhilippvK has joined #asahi
phiologe has quit [Ping timeout: 480 seconds]
Catyre has quit [Remote host closed the connection]
pyropeter3 has joined #asahi
pyropeter2 has quit [Ping timeout: 480 seconds]
marvin24 has joined #asahi
marvin24_ has quit [Ping timeout: 480 seconds]
kov has quit [Quit: Coyote finally caught me]
Catyre has joined #asahi
Catyre has quit [Ping timeout: 480 seconds]
dethklok_x has joined #asahi
dethklok_x has quit []
skipwich has quit [Ping timeout: 480 seconds]
<MichaelMesser[m]> https://twitter.com/never_released/status/1550185220317528064 Does this mean Linux on M1 can support a Nvidia eGPU?
<tpw_rules> i mean that provides no evidence except "nuh uh"
<tpw_rules> the poster claims to be working on gpu virtualization but that probably fails the "practical" test
skipwich has joined #asahi
<ElvishJerricco[m]> Yea I don't really see anything in the poster's tweets that indicates they have any reason to believe it'll be better than marcan says.
mx13606[m] has joined #asahi
<sven> i guess they know sone magic register (which I think is unlikely) and/or are talking about horrible hacks
<tpw_rules> what would a GPU's BARs typically be mapped as? GRnE?
<tpw_rules> or GRE rather
<sven> iirc the drivers map it as normal-nc
<sven> maybe even as cacheable
<sven> essentially for framebuffers, textures, etc. inside gpu ram
Catyre has joined #asahi
Catyre has quit [Ping timeout: 480 seconds]
ciggi_ has joined #asahi
ciggi has quit [Read error: Connection reset by peer]
<kode54> is there an updated install script?
<jannau> updated in which way?
<kode54> will work with a machine that has had Ventura installed on it
<jannau> yes. It has nothing to do with the installer script but with the u-boot and kernel binaries it installs. The binaries were refreshed with the update
<kode54> gotcha
MajorBiscuit has joined #asahi
linearcannon has joined #asahi
Major_Biscuit has joined #asahi
MajorBiscuit has quit [Ping timeout: 480 seconds]
fossdd has joined #asahi
thevar1able1 has quit [Ping timeout: 480 seconds]
systwi has joined #asahi
thevar1able has joined #asahi
the_lanetly_052 has joined #asahi
<marcan> they're talking about HMM style management with page faults flipping pages between GPU and CPU
<marcan> apparently that's a thing nvidia does on some systems
<marcan> I'm not convinced it'll work with their blob driver though, without a bunch of work on their side which I doubt anyone will care to do
<marcan> and it wouldn't work for radeon/nouveau AIUI
the_lanetly_052__ has joined #asahi
the_lanetly_052 has quit [Ping timeout: 480 seconds]
kov has joined #asahi
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<kettenis> marcan: not sure what you mean with HMM, but NVIDIAs "open source" kernel driver has some interesting hints about how they implement write combining on arm64:
<kettenis> and
<kettenis> which suggests that they do map video RAM as DEVICE_GRE (or DEVICE_nGnRE if the kernel doesn't support DEVICE_GRE)
<maz> odd. I've had robust discussions with some of the NV folks who were adamant they wanted to implement WC as NC. different groups, I suppose.
<maz> they certainly were advocating for the S2 mappings to be NC instead of nGnRE when mapping PCIe devices.
<Sobek[m]> Can anyone remind me what does (n)G(n)R(n)E stands for?
<maz> Gathering, Reordering, Early response
<j`ey> and the (n) is the non- versions
<kettenis> maz: the left hand doesn't know what the right hand is doing ;)
<Sobek[m]> Is this specifically for memory space for PCIe devices? Does it imply cacheable or is it orthogonal with NC (not cacheable) and WC (write cacheable am i right?) ?
<kettenis> anyay, (sadly) this means that the nnvidia blob drivers have a fair chance on working as an eGPU on M1
<Sobek[m]> Well, we'll see. Given raspberry pi seem to be affected by similar issue, would the same of change on nvidia part make it work on M1 and raspberry pi, if small tweaks are needed ?
<kettenis> I believe rpi is whole different level of pain as the PCIe controller has issues with 64-bit transactions on the bus it is connected to
<Sobek[m]> (If anyone else is trying to make sense of the GRE things : https://developer.arm.com/documentation/100941/0101 appears to be the relevant doc)
<mini> hey - for a m1 mac mini, am I likely to have issues with external 4K HDMI displays for a console for initial setup? I don't have anything 1920x1080 or similar
<mini> I haven't had this running linux since pre-installer existing times, and that was all done over USB console access & SSH
dabaum|2er0 has quit [Ping timeout: 480 seconds]
<Glanzmann> mini: You need to append the resolution in the /boot/efi/m1n1/boot.bin: echo 'display=3840x2160' >> /boot/efi/m1n1/boot.bin
<Glanzmann> mini: I do the same on my mini but for a 2k monitor. At some point we will have a dcp driver that automatically handles that.
<mini> Glanzmann: thanks. I'm aware we're waiting for a proper DCP driver, that's why I wasn't sure :)
<Glanzmann> mini: But it works today. RayDarcy is using it. I tested it on a 4k screen and I'm using it on a 2k screen.
dabaum|2er0 has joined #asahi
alarm has joined #asahi
<alarm> hi guys, may i ask a question, does the aux for the audio works?
<j`ey> it should do
Sellerie has joined #asahi
<alarm> i am pluggin them but they doesn t work?
<alarm> is there any utility to manage the audio ?
ChaosPrincess has quit [Quit: WeeChat 3.5]
<Glanzmann> alarm: alsamixer
ChaosPrincess has joined #asahi
Zaprit has joined #asahi
<as400[m]> audio works only on m1, right ? m1 pro/max are not supported yet ?
dabaum|2er0 has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
dabaum|2er0 has joined #asahi
alarm has quit [Quit: Lost terminal]
alarm has joined #asahi
alarm has quit [Quit: leaving]
<Manouchehri> are there any cloud providers that are known to let you install asahi?
qyousef_ has quit [Ping timeout: 480 seconds]
<Manouchehri> Scaleway is a no-go from my experience, since they don't let you access a physical KVM on it (you just have SSH and VNC with the native macOS stuff).
<ChaosPrincess> unlikely, the install requires you to press and hold the power button to 'assert physical presence', and well, you are not physically present
<sven> aws at least presses the button with a small solenoid inside the rack because they kinda need that to restore them to a sane state
<Glanzmann> Manouchehri: Probably not, but i received an email this week from aws that m1 instances are GA.
<sven> wouldn't surprise me if the rest does something similar
<sven> but why do you want to run asahi on a cloud machine? isn't it cheaper to just get "regular" VM?
<Glanzmann> sven: Yes. Break even is around 3 months.
<Manouchehri> Glanzmann: oh right, aws offers them. I forgot since they charge an insane amount.
<Manouchehri> sven: the M1 is one of the best aarch64 boxes for dev work IMO.
<Manouchehri> I already have one, but it'd be nice to be able to pay for a few more short-term when needed.
<j`ey> Manouchehri: run a linux VM inside macos?
<sven> huh, i always assumed you could just get more powerful vms for general dev work if required
<Manouchehri> sven: aarch64 options with KVM and rr support are veeerrrrrry limited.
<j`ey> graviton at aws?
<sven> iirc oracle has arm64 vms as well
<Manouchehri> j`ey: iirc c6g*.metal are the only ones that work with rr.
<j`ey> what changes between the VMs that makes rr not support it?
<Manouchehri> no hardware perf counters accessible
<_jannau_> VMs don't help when KVM is a requirement. hetzner announced 80 Core Ampere Altra servers but availability is limited and only useful for long term use
<Manouchehri> I may be wrong (need to double check), but I thought metal instances were the only arm instances at AWS that supported kvm
<Manouchehri> and c6g.metal is over $2/hour. After a month, might as well get a Mac Mini.
<Manouchehri> $0.64 to $1ish an hour for spot instances.
<_jannau_> hetzner wants €169 per month + the same amount for setup
<maz> Manouchehri: indeed, as N1 doesn't support NV (nor does KVM for now)
<Manouchehri> maz: NV?
<j`ey> nested virt
<Manouchehri> _jannau_: interesting, lemme check them again
<maz> nested virt,
<_jannau_> https://www.hetzner.com/dedicated-rootserver/matrix-rx hmm, the announcement had apparently prices without VAT
yuyichao_ has joined #asahi
<Manouchehri> maz: isn't nested virt not a thing on aarch64 because of no linux kernel support?
<_jannau_> Manouchehri: "nor does KVM for now"
<Manouchehri> oh lol, I read that earlier this week, didn't make the maz == maz connection
<j`ey> I hope markan or j4nnau will try that patch set at some point.. that would be neat
<Manouchehri> _jannau_: try it on which platfom?
<Manouchehri> *platform
<_jannau_> m2
Zaprit has quit [Ping timeout: 480 seconds]
<Manouchehri> what's the difference in armv8.X between the M1 and M2?
<_jannau_> I think they are both 8.5 but m2 implements at least NV as optional feature
<Manouchehri> interesting, maybe I should wait until the Mac Mini M2 comes out before buying another one.
<j`ey> also: ECV, FGT, BTI, TWED
<maz> both really a mix of various arch revisions, with stuff ranging from 8.3 to 8.6... ARMv8.x is only a vague indicator these days.
<Manouchehri> thanks! I'm really starting to dislike the ARMv8.x naming structure :p
<maz> there is no structure, that's the very point! ;-)
<j`ey> Manouchehri: as I said before, just look at the FEAT_* and ignore the .x version
<Manouchehri> j`ey: is there a good online source for finding those feature flags on systems I don't own? :P
<j`ey> dunno, but that would be cool
<j`ey> ID reg database
<maz> don't think there is one. maybe we should poll information!
<Manouchehri> https://openbenchmarking.org/result/2203239-NE-APPLEM1AS32 <- maybe we could add a test script to openbenchmark / phoronix-test-suite that uploads stuff like /proc/cpuinfo?
mindfreeze_ has joined #asahi
Zaprit has joined #asahi
Zaprit has quit [Ping timeout: 480 seconds]
Zaprit has joined #asahi
the_lanetly_052__ has quit [Ping timeout: 480 seconds]
bisko has joined #asahi
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
alarm has joined #asahi
bisko has joined #asahi
<alarm> a
petterroea has joined #asahi
<alarm> sorry but i cant manage to make the sound work not even with alsamixer, it doesn t seem to get my headphones on the aux
<alarm> am i the only one?
<opticron> just to verify, which device?
<alarm> device?
<alarm> i just have the integrated audio on alsamixer
<alarm> not my headphones
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<j`ey> alarm: what m1 device?
<alarm> macbook pro 13 m1
<alarm> sorry but the question how do you answerd with the name of the person
<opticron> yeah, the headphone jack doesn't work on that yet
<opticron> but bluetooth does as of the last few days, so that's a viable way to get audio out
<alarm> how can i get the audio out?
<alarm> how do i connect the headphones bluetooth
<alarm> like i have the beats solo pro
<opticron> I'm not sure what you're running on it, but you'd open the bluetooth settings panel and pair it once the headphones are in pairing mode
<alarm> oh man
<opticron> err, wait, I might be wrong about that model, I thought I read 14"
<alarm> one time they worked the headphones with the jack
<alarm> but then they stopped
doggkruse has joined #asahi
<opticron> 3.5mm jack say ssee notes
<alarm> where the notes at
<alarm> sorry my bad
<opticron> aaand the notes on that page don't say anything about the 3.5mm jack
doggkruse_ has joined #asahi
<alarm> ahahha
<alarm> rip
<alarm> so back to the bluetooth
<alarm> i do not have the desktop installation
<alarm> i have a tailing window manager
<alarm> do yuo know any sw to make the bluetoot pairing
<opticron> I...haven't done bluetooth management without a DE in probably a decade, so I'm not going to be very helpful
<opticron> before the page reorg to get things in tables, the M1 MBP 13" table claimed that 3.5mm jack support was in linux-asahi
<alarm> sorry didn t understood it
<alarm> what?
<opticron> I think the reorg on the wiki lost some information
<opticron> *reorganization
<alarm> oh damn
<alarm> let me ask u this
<alarm> which is the services to enable for bluetooth
<alarm> i ve found a sw to pair the headphones
<alarm> it s called bluex
<alarm> it s called bluex
<alarm> like i ve tried to do systemctl enable bluetooth
<alarm> but none
doggkruse has quit [Ping timeout: 480 seconds]
<opticron> alarm, hcitool?
kaseyreed has joined #asahi
<alarm> opticorn: what is it?
<_jannau_> deprecated and not available on asahilinux, bluetoothctl is the program to connect bt devices on the command line
<_jannau_> the headphone jack is supposed to work on the original m1 devices
<opticron> yes, I'm showing the age of my knowledge
<opticron> and I fixed the wiki feature tables to indicate support of 3.5mm jack on M1 devices
<alarm> should i install hcitool?
Gaspare has joined #asahi
<sven> why?
<sven> just use bluetoothctl
<_jannau_> I don't think hcitool is available in asahi linux / arch linux arm
<alarm> yeah i am using bluetoothctl but it says wiating to connect to bluetoothd
<alarm> but i do have started the bluetoot serice
<alarm> but i can t start it i get inactive(dead)
<alarm> and this log message : Jul 22 15:13:31 alarm systemd[1]: Bluetooth service was skipped because of a failed condition check (ConditionPathIsDirectory=/sys/class/bluetooth)
<j`ey> alarm: what does uname -a show?
<j`ey> wondering if you have a new enough kernel
<alarm> Linux alarm 5.17.0-rc7-asahi-next-20220310-5-2-ARCH #2 SMP PREEMPT Tue, 22 Mar 2022 14:28:10 +0000 aarch64 GNU/Linux
<alarm> that s the output of the uname -a
<j`ey> ok yeah, that doesnt have bt support
<alarm> how can i update it?
<j`ey> pacman -Syu
<alarm> it says to me there is nothign to do
<alarm> already up to date
<_jannau_> what does `pacman -Qs linux-asahi` report?
eroc19905 has joined #asahi
<alarm> _jannau_: local/linux-asahi 5.19rc7.asahi1-1 The kernel and modules
<alarm> that what i got with that command
qyousef has joined #asahi
kefu__ has joined #asahi
mindfreeze__ has joined #asahi
vx^ has joined #asahi
maennich_ has joined #asahi
<j`ey> alarm: reboot?
<_jannau_> have you rebooted since you ran pacman -Syu the last time?
<alarm> let me try, i will be back in a minute, thanks
alarm has quit [Remote host closed the connection]
Chainsaw_ has joined #asahi
h_ has joined #asahi
mindfreeze_ has quit [synthon.oftc.net graviton.oftc.net]
hays has quit [synthon.oftc.net graviton.oftc.net]
Erus_Iluvatar has quit [synthon.oftc.net graviton.oftc.net]
dimbulb has quit [synthon.oftc.net graviton.oftc.net]
eroc1990 has quit [synthon.oftc.net graviton.oftc.net]
vx has quit [synthon.oftc.net graviton.oftc.net]
blevins has quit [synthon.oftc.net graviton.oftc.net]
JTL has quit [synthon.oftc.net graviton.oftc.net]
kefu_ has quit [synthon.oftc.net graviton.oftc.net]
Chainsaw has quit [synthon.oftc.net graviton.oftc.net]
djorz has quit [synthon.oftc.net graviton.oftc.net]
maennich has quit [synthon.oftc.net graviton.oftc.net]
Chainsaw_ is now known as Chainsaw
qyousef has quit [Quit: WeeChat 2.8]
blevins has joined #asahi
JTL has joined #asahi
Erus_Iluvatar has joined #asahi
dimbulb has joined #asahi
mindfreeze__ is now known as mindfreeze_
qyousef has joined #asahi
alarm has joined #asahi
<alarm> guys i love you!!!
<alarm> headphones working!!!
<Manouchehri> real stupid question, how can I tell which dtb my system is using?
<alarm> you made my day guys, thanks
<sven> :)
<j`ey> Manouchehri: try cat /proc/device-tree/compatible
<Manouchehri> ah /boot/efi/asahi/stub_info.json
<Manouchehri> j`ey: ty, that's a better option
<Manouchehri> kexec_load failed: Function not implemented
<Manouchehri> AWW
<Manouchehri> sorry caps lock
<j`ey> no PSCI yet
<j`ey> which kexec needs
<Manouchehri> ahh, thanks!
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi
Gaspare has quit [Quit: Gaspare]
<Manouchehri> tpw_rules: thanks for the help with 4k builds, I made a PKGBUILD for others. :) https://github.com/AsahiLinux/PKGBUILDs/pull/18
<j`ey> https://github.com/AsahiLinux/PKGBUILDs/pull/6 a pr for it was closed previously
<Manouchehri> oh lol
<j`ey> check how it sets pagesize, to avoid duplicating the config
<sven> yeah, there's a reason that patch isn't part of the asahi branch :D
<Manouchehri> j`ey: yeah I like how it used source
Zaprit has quit [Ping timeout: 480 seconds]
Zaprit has joined #asahi
<Manouchehri> sven: do you have a git repo of the patch somewhere?
<sven> i think it's still in some branch in asahilinux somewhere
<sven> but it's kinda outdated at this point anyway
<Manouchehri> ah right, tpw_ru1es' one was updated
<sven> https://github.com/AsahiLinux/linux/tree/iommu-4k/dev but fair warning: no support whatsoever from my side and not recommended to ship to users
<Manouchehri> sven: this is unacceptable, I demand a 100% refund of all support fees I've paid you.
<sven> :D
<sven> on a more serious note, i'll eventually upstream that patch once robin (major iommu contributor) lands his iommu api cleanup series. once that thing lands I can remove the ugliest hack in there :)
<Manouchehri> I don't see anything majority wrong with my system, it seems to be working fine
<Manouchehri> *majorly
Zaprit has quit [Ping timeout: 480 seconds]
Catyre has joined #asahi
Major_Biscuit has quit [Ping timeout: 480 seconds]
Gaspare has joined #asahi
Major_Biscuit has joined #asahi
alarm has quit [Remote host closed the connection]
Catyre has quit [Remote host closed the connection]
<Sobek[m]> <sven> "on a more serious note, i'll..." <- and this will result in a kernel config option to build 4K vs 16k ?
<Manouchehri> Sobek[m]: that already exists
<Sobek[m]> Could the selection be dynamic using a boot argument ?
<j`ey> Sobek[m]: no
<Sobek[m]> Manouchehri: Why is the patch needed then ?
<Manouchehri> Sobek[m]: because it can't be configured after compiling?
<sven> because the iommu subsystem currently doesn’t work when iommu pagesize > kernel pagesize
<sven> that patch fixes that
<Sobek[m]> So the patch is needed for the config option to work but the config option already exist , got it.
<sven> more or less. the config options already works, the kernel will boot just fine
<sven> you just won’t be able to use anything attached to the PCI bus e.g.
<Sobek[m]> Anyway currently updating my Asahi partition….
Catyre has joined #asahi
Luciee has joined #asahi
Luciee has quit []
qeeg__ has quit [Remote host closed the connection]
qeeg has joined #asahi
doggkruse has joined #asahi
doggkruse_ has quit [Ping timeout: 480 seconds]
<Glanzmann> sven: Are you looking for testers, than I would try it out and report back.
<Glanzmann> context: 4k-iommu
<sven> no
<sven> that patch will change quite a bit once the iommu api cleanup lands
Major_Biscuit has quit [Ping timeout: 480 seconds]
<Glanzmann> sven: I see. Than I'll not try it until you tell me or it is merged into the asahi tree.
<Manouchehri> Glanzmann: what do you want to test?
<Manouchehri> it works fine here :)
<Glanzmann> Or rather master branch.
<Glanzmann> Manouchehri: The 4k-iommu patch from sven.
<Manouchehri> Glanzmann: do you need 4k pages?
<Glanzmann> Manouchehri: Nope, but to many Debian users complained that software is not working, I decided to ship the 4k patch after consulting with jannau.
<Glanzmann> Okay it was not a consulting, it was more of asking for his opinion long before I shipped it to users. He probably forget about it but j e y can find it in the backlog. :-)
<Glanzmann> Sobek[m]: Well, you could install two kernels. One 4k and one 16k and choose the right one. Dt is the same, isn't it?
<Glanzmann> Manouchehri: I know what m a r c a n decided. And I thought it was a bold move. But on the other hand this is how you get things done or software fixed.
<Glanzmann> I just wanted to deliver a Debian which just works ...
<Manouchehri> Glanzmann: ya that's what I'm already doing on arch :)
<marcan> it's possible that kexec is ~never going to work on these systems, if we don't find a reliable way of resetting *all* firmwares
<marcan> AGX and MTP are still a question mark there
doggkruse_ has joined #asahi
<marcan> MTP we can probably find a way to hand off if it comes to that
<marcan> AGX... that's a lot less likely
doggkruse has quit [Ping timeout: 480 seconds]
<Glanzmann> marcan: Are you using kexec on servers or workstations?
<Glanzmann> Or is there a particular usecase you have in mind like using the linux kernel as bootloader or something like that?
<marcan> I don't personally use it
* Glanzmann neither. I see.
<ar> non-IBM power machines tend to use linux as a bootloader (petitboot)
<ar> but i'm not sure that's a very good idea
___nick___ has joined #asahi
___nick___ has quit []
___nick___ has joined #asahi
<Glanzmann> marcan: What did you do with the windows on the macos? Is this running in a simulator or ist a rdp session?
<Glanzmann> s/macos/mac pro/
<j`ey> qemu
VasilisTheChu has quit [Read error: Connection reset by peer]
<Glanzmann> I see. Wow with 4k pages?
<Glanzmann> and x86?
doggkruse_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<j`ey> well the page size doesnt matter for qemu, when emulating
<Glanzmann> Oh, I did not know that.
<Glanzmann> Was it an arm or x86 guest?
<j`ey> x86 guest, dont think windows on arm worked back then :P
<Glanzmann> I see.
<Glanzmann> j`ey: Was this run using a jit or were the instructions simulated?
<j`ey> well theyre translated from x86 into arm, so yes jitted.. but also simulated? :P
<Glanzmann> j`ey: When I said simulated, I thought bochs interpreter style. :-)
doggkruse has joined #asahi
<j`ey> ah, nah
<Glanzmann> Wow, this is a big step. Didn't know this was already done. Now I want to know how much overhead it has.
<Manouchehri> marcan: kexec would be fun, but it seems totally unnecessary given how fast everything boots
<Manouchehri> Glanzmann: I have android running in VMs on mine :)
<Glanzmann> I see.
doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Cy8aer[m]> For those who want to have BT music quit: I did it with pipewire and blueman
<Cy8aer[m]> (saw some questions about this). Of course you need the firmware updated as well. For those debians: It is a bit difficult to explain what I did. I will write some docu
<Cy8aer[m]> s/docu/documentation/
<marcan> Glanzmann: qemu is 11 years old, this isn't new
<j`ey> 11 years old.. that seems newer than expected
<marcan> well
<marcan> 1.0 was 11 years ago
<marcan> apparently it started 8 years before that
<j`ey> ah
<j`ey> you can redo that video with the m1 at 600MHz :P
ptudor_ has joined #asahi
<marcan> m1 actually performs worse than it should, I think because the qemu video output isn't framerate capped and it ends up thrashing the software compositor
<marcan> it works well at 1:1 pixel scale, but zoomed you can see it get sluggish
ptudor has quit [Ping timeout: 480 seconds]
AoV has joined #asahi
<Glanzmann> marcan: I know qemu, I did not know that It can x86 on arm64. :-)
ptudor_ has quit [Quit: Strict-Transport-Security: max-age=48211200; preload]
<jannau> without kvm it can any supported guest architecture on any supported host architecture, it started as binary translation emulator
<jannau> Glanzmann: I have indeed no recollection of recomending to ship the 4k iommu support patch to end users
doggkruse has joined #asahi
<jannau> I certainly said you need it for a 4k kernel with PCI support
<jannau> but that is not a recomendation to ship the patch and a 4k kernel
<Cy8aer[m]> Ok, here my debian bluetooth update doc: https://g3la.de/hedgedoc/#Bluetooth
doggkruse_ has joined #asahi
<Glanzmann> jannau: Thank you for reminding me, I forgot about it. And I'm actually using qemu to package Debian for m1 on amd64.
<j`ey> Cy8aer[m]: cool thanks, I haven't gotten bluetooth working yet, should try soon
doggkruse has quit [Ping timeout: 480 seconds]
<Glanzmann> jannau: No, I asked you long before the release, if you would go with 4k and don't break so much or 16k for performance and you said something along the lines, that you prefer 16k yourself but would probably ship 4k to make adiotion easier or something along the lines. But since I'm not keeping any logs, I only have my memory. :-)
<Glanzmann> Cy8aer[m]: Nice write-up.
djorz has joined #asahi
<Cy8aer[m]> Hope it not just works for me
<Glanzmann> Cy8aer[m]: I went the lazy route and added an additional stub to grab the firmware. And by not paying attention deleted my active stub and than I had to test my asahi boot skills. Installed a new stub and used the live to reinstall grub and fix fstab ...
<Cy8aer[m]> Ok - till now I did not have a crash. So I miss these skills ☹️ 😉
<Cy8aer[m]> Glanzmann: but you yourself told me to do it my way...
<Glanzmann> Cy8aer[m]: This was also the first time I acceidently deleted the wrong partition. But doesn't matter. I also need to a video soon when marcan delivers the gpu drivers in order to get the early adopters to an up2date stub.
<Cy8aer[m]> yeah then I need to re-stub too (and hope for your documentation)
<Glanzmann> Cy8aer[m]: I know, I thought about going that way but than go the lazy but more time consuming way. I also tried to call the firmware extraction python code from the installer, but failed.
<Glanzmann> Cy8aer[m]: Who know maybe we stay lucky and can stay on the stub and just need some firmware. :-)
<Cy8aer[m]> let's see...
<Cy8aer[m]> I need to work on my private invidious server now because it does not show live streams (no live lina). But it should work...
kaseyreed has quit [Quit: Konversation terminated!]
<Glanzmann> jannau: Probably yes, okay I interpreted a lot into 'but I'm not your typical user'. :-)
<jannau> the "but I'm not the typical user" ment I will compile my own kernel anyway and don't care what distros ship
<Glanzmann> At the moment I'm also compiling a kernel a /'''nv
<Glanzmann> /'vg.jv'bjv
<Glanzmann> .
<Glanzmann> jannau: I see. ^ this was my daughert who wanted to 'work on the laptop'.
ptudor has joined #asahi
Major_Biscuit has joined #asahi
___nick___ has quit [Ping timeout: 480 seconds]
<Manouchehri> going a bit crazy, getting a new build failure for a package I'm pretty sure I could build yesterday. https://gist.github.com/Manouchehri/05dc1aafb115e2e15500f49e2b377223
<Glanzmann> Manouchehri: I would file a priority one case with sven and ask him while he corrupts your buffer cache and file system with his fancy new 4k mmu patch.
<Glanzmann> Or it is just a missing package.
<Glanzmann> wh*
<Manouchehri> Glanzmann: I already fired sv3n in this chat earlier, remember?
<Glanzmann> Rehire him.
<Glanzmann> Also now is the time to check your backups. Hopefully not. I wish that to noone. *knock on wood*.
Major_Biscuit has quit [Ping timeout: 480 seconds]
<Manouchehri> sanity check, can someone print their /etc/pacman.conf?
<Manouchehri> never mind, got it working, it turns out something broke upstream a few commits ago
<Manouchehri> god bless /var/log/pacman.log
<Glanzmann> Good night sleep everyone.
yamii has quit [Quit: WeeChat 3.6]
yamii has joined #asahi
<clover[m]> what happened to the Matrix room icon? I feel like it used to be the asahi logo but now its a blue 'A'
<Astra[m]> Still the asahi logo for me
bluetail has joined #asahi
<bluetail> Hello. I found that the HDMI issue was resolved according to the news. I want to switch my m1 mac to use Linux because of problems with special characters in paths on macOS. I just hope Docker runs properly on there.
<bluetail> Daily it serves also as a NAS as weird as it may sound
<bluetail> Is the migration recommendable? I am good with Linux but I dont know if the software runs properly yet
<Lucy[m]> Only dual booting is supported right now anyway, so you won't really risk anything
<Lucy[m]> Give it a try and see if it works well enough for you
<bluetail> Lucy[m], theres no risk as long as I follow the guide yes. But what about time loss. Is it worth doing at the moment? I want to fix the occasional issues
<bluetail> Can you recommend it at the time?
<bluetail> would you use it instead of macOS for 'serving' purposes?
<bluetail> I remember one dealbreaker was that I couldnt manage to setup the VPN... Is openvpn supported now?
<Lucy[m]> Probably yeah, because I'm way more comfortable with Linux/BSD. Docker should be fine, you just won't have any have any GPU acceleration, but you likely knew that. Not sure about openvpn, someone else might be able to comment on it.
<bluetail> Good. I will set it up.
doggkruse_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
doggkruse has joined #asahi
Gaspare has quit [Quit: Gaspare]
<bluetail> Why do I fail to use my monitor in portrait mode?
<bluetail> Could that be one of the current limitations?
<tpw_rules> yes, you can't change monitor resolution
<bluetail> Ah I see. I guess SSH/VNC is then a must until it is all ready
<bluetail> Is there a reason I had to run pacman -Syu 3 times in order to not get errors?
<j`ey> looks like the mirror was having issues
<bluetail> I see
<bluetail> j`ey, I have another question. I have a USB3 storage Device and its formatted in exFAT. For some reason I cannot mount it with the default options. Do I require to download special software?
<bluetail> Or do I have to plug it into a different port? It is right now using the thunderbolt port, but its not a thunderbolt device; in fact it is a usb c device
<Tramtrist> science
<j`ey> linux supports exfat natively
<bluetail> I know. It's so awkward. Perhaps I have to mount it as root or something. But this doesnt make any sense... Does it? When I select Plasma to auto-mount the device, it should do just that. I gotta try a different physical port
<j`ey> you can check dmesg for errors
<bluetail> Normal USB Type A works. USB C doesnt
<j`ey> good
<bluetail> stat urb -71
<bluetail> but runs at usb 1.1 it seems at type A ports
<bluetail> I gotta confirm whether its actually only 11mbps
<bluetail> 12*
<bluetail> Nah. Its far faster 2gb could be written within 10s