nicolas17 has quit [Quit: Konversation terminated!]
user982492 has joined #asahi
Graypup_ has quit [Quit: meow]
Graypup_ has joined #asahi
Ry_Darcy has joined #asahi
richardburleigh has joined #asahi
the_lanetly_052 has joined #asahi
<Skirmisher[m]>
bluh, I still need to figure out why sound won't work on my mini with the almost-identical-to-asahi kernel I built
<Skirmisher[m]>
or rather, how to fix it
<Skirmisher[m]>
`ALSA pcmC0D0p,0:: cannot preallocate for size 3145728`
the_lanetly_052___ has quit [Ping timeout: 480 seconds]
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
user982492 has joined #asahi
richardburleigh has quit [Ping timeout: 480 seconds]
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi
kameks has joined #asahi
the_lanetly_052 has quit [Ping timeout: 480 seconds]
LinuxM1 has joined #asahi
LinuxM1 has quit [Quit: Leaving]
LinuxM1 has joined #asahi
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
neatsquirrel has quit [Quit: neatsquirrel]
user982492 has joined #asahi
hectour[m] has joined #asahi
<hectour[m]>
I saw the Twitter thread about there being no virtualization support on Apple Silicon. Does this mean that using KVM to virtualize Linux VM's with Asahi Linux as the host OS also will not work?
<sven>
uhhhhh
<jannau>
hectour[m]: no, kvm on asahi linux can be used to virtualize a standard aarch64 system
<sven>
^—
<sven>
that thread was probably about virtualizing macOS
<jannau>
it's just not possible to use qemu+kvm to virtualize a Apple silicon machine capable of running macos or m1n1
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Ry_Darcy has quit [Remote host closed the connection]
<bluetail[m]>
using Asahi Linux fixes weird macOS flashes where part of the background border of another window flashes into the foreground
<bluetail[m]>
* fixes weird bug on macOS, * macOS - flashes where
<bluetail[m]>
* using Asahi Linux fixes weird bug on macOS - sometimes in macOS a part of the background border of another window flashes into the foreground
<psydroid[m]1>
macOS is one big bug, if you ask me
<ar>
most of the ones I encountered were around usb enumeration and the fact that it only does 1 attempt at link training for displays
<Dcow[m]1>
rqou_: thank you very much. bluetooth is very anticipated feature my be ;)
<Dcow[m]1>
oops, I mean "by me"
<bluetail[m]>
ar: the worst I had is when using disk utility and then replug a drive and then erase another disk and then it erases something important and I cant cancel
<bluetail[m]>
delete banana? Now deleting backup :D
<rqou_>
Dcow[m]1: it still needs a bunch of work, but the heavy lifting should be done
<Dcow[m]1>
yeah, I understand that
<ar>
bluetail[m]: the gui disk utility is just bad
<bluetail[m]>
command line / terminal / tty ftw
kenzie has joined #asahi
<mps>
bluetail[m]: hm, did you managed to install alpine linux on your m1 machine
<bluetail[m]>
I didnt pursue it
<bluetail[m]>
Still on AARCH
<bluetail[m]>
but its a bit annoying that sometimes it takes 4 reboots to boot asahi
<bluetail[m]>
* boot asahi AARCH
<bluetail[m]>
could be because I have a ton of USB devices
<bluetail[m]>
I guess it has something todo with the bootloader
<bluetail[m]>
* with the thing before the bootloader
<mps>
bluetail[m]: shouldn't be that, also I have usb hubs and other devices, always worked fine
<bluetail[m]>
it does work... somehow but it isnt that it boots it immediately
<mps>
bluetail[m]: you use grub as bootloader?
<bluetail[m]>
yes
<bluetail[m]>
the thing before the bootloader doesnt work on every reboot
<bluetail[m]>
so it tries quite a few times and then finally boots AARCH
<bluetail[m]>
is it called uboot or m1n1?
<mps>
bluetail[m]: m1n1+u-boot chained
<mps>
very strange because for me these works flawlessly
<mps>
and I don't have any idea why it is flaky for you
<mps>
u-boot by default tries to boot first from nvme
<bluetail[m]>
mps: I have two usb ssds that are formatted with ExFAT, Two Bitlocker Encrypted drives
<bluetail[m]>
> <@_oftc_mps:matrix.org> and I don't have any idea why it is flaky for you
<bluetail[m]>
* I have two usb ssds that are formatted with ExFAT
<bluetail[m]>
and I guess
<bluetail[m]>
one uninitialised drive
<bluetail[m]>
actually 2
<mps>
if I plug 3 usb disks prepared to boot asahi it always boots from nvme if I don't break u-boot manually
<bluetail[m]>
Perhaps it has todo with the uninitialised disks
<j`ey>
bluetail[m]: why not unplug everything and boot/
<j`ey>
rather than speculating!
<bluetail[m]>
oh wait
<bluetail[m]>
yea
<bluetail[m]>
ok
<bluetail[m]>
I have like
<bluetail[m]>
some bootable usb drives on my laptop and a usb switch which pushes some of those drives to the mac
<bluetail[m]>
I shouldnt remove em
<bluetail[m]>
ok, a moment, I try it out
<bluetail[m]>
mps: I tested now. It does immediately get into the GRUB bootloader but it took me two times to succesfully boot
<bluetail[m]>
it failed with error 71 on USB 1.1 something
<bluetail[m]>
I had to force shutdown via the button and then start the machine again
<bluetail[m]>
but it at least immediately booted into grub with only mouse and keyboard attached
<bluetail[m]>
* via the power button and
<bluetail[m]>
* 1.1 something and then timed out to boot
<mps>
hmm, strange
<bluetail[m]>
is there a way to dump the boot log right now as I'm in AARCH?
<j`ey>
it's just Arch, not AARCH :P
<mps>
though I installed locally built m1n1+dtbs+u-boot in ESP
<bluetail[m]>
j`ey: why do some say AARCH then? I thought it was the right way
<bluetail[m]>
oh its cause the aarch64 thingy
<bluetail[m]>
thats why I call it AARCH
<j`ey>
that's AArch64
<j`ey>
Arch Linux and AArch64 architecture :)
<bluetail[m]>
oh ic
<ey3ball[m]>
DId anyone manage to get qemu-user-static running under asahi ? I tried various builds, and it works somehow (I'm able to run some simple binaries in an ubuntu chroot), but most programs simply fail with :
<ey3ball[m]>
I tried running my own build from master as well although that's from a few days ago already
<mps>
arnd: didn't looked at this, but imo badly written application
<mps>
don't see reason why apps could be 'locked' to particular kernel PAGE_SIZE
<mps>
but yes, even some kernel drivers have this problem
<arnd>
of course they shouldn't be, the question is whether that's just a bug in minecraft or in the android runtime environment
<mps>
what happens with some other android apps
<arnd>
I need to find one I can run, this is the only one I was running on my x86 machine, and it's packaged conveniently with a launcher tool here. The launcher itself seems to work fine fwiw, it only fails to run the actual Android app
<VinDuv>
There seems to be a call to mmap in the appimage itself, but no calls to sysconf, so I guess it uses a hardcoded page size
<mps>
whenever something crash on my machine I first ascribe this to musl vs glibc incompatibility, and have some apps in queue to fix
<ey3ball[m]>
mps: do you happen to know what these patches are about, compiled with debug options and I'm getting output which looks slighly more usefull :
<mps>
ey3ball[m]: I think you don't need tcg on asahi, kvm works
<j`ey>
mps: they want x86 emulation
<mps>
j`ey: ah
<mps>
isn't fex better for this
<ey3ball[m]>
mps: isn't fex the one that was listed as "won't support 16k pages" on the wiki ?
<mps>
ey3ball[m]: didn't know this, tried to build it just once but gave up because don't have time to debug build on musl
<jannau>
I wouldn't be surprised if qemu-user has the same 4k/16k problem
<ey3ball[m]>
<j`ey> "mps: they want x86 emulation" <- Yes basically I want to run x86_64 docker images transparently, and it looks like qemu-user-static is the only viable option
<ey3ball[m]>
(or it was)
<ey3ball[m]>
Guess I should report a qemu bug then
<mps>
idk, for me qemu-static-{arm,riscv64,aarch64} works on kernels with 16 ad 4 K pages
<sven>
fex and box64 might also work until they run into an app that just assumes 4k pages somewhere
<mps>
s/static/system/
<sven>
the issue isn’t necessarily the translation, it’s that assuming 4K pages on other architectures is a perfectly valid assumption
<ey3ball[m]>
So far from distro to distro I'm getting different results with x86 translation, on an arch chroot I can't run "sh", on ubuntu i can run sh but not "ls", and alpine seems to be working relatively well so far (although I only tested basic shell and package management)
kameks has joined #asahi
<mps>
heh
<mps>
all developers should run alpine ;) (sorry from endorsing it again)
<sven>
userspace is a myth ;p
<j`ey>
run vmlinux as pid1 :o
rezn1t has joined #asahi
Ry_Darcy has joined #asahi
AntonioSchettini[m] has joined #asahi
<bluetail[m]>
vmlinuz is the compressed vmlinux right?
<ar>
yes
<Ry_Darcy>
@all With all of the discussion(s) of a 4K vs 16K vs 64K Kernel capability, what are the advantages for someone who just uses an M1 Mini for daily use, i.e. no DB, applications memory or CPU-bound of using a 16K/64K kernel?
<Ry_Darcy>
@all none methinks. Ideas?
<bluetail[m]>
stop trying to ping all LOL
<bluetail[m]>
mainly privacy and less buggy
<Ry_Darcy>
@all with such a question, it is difficult to address one person in particular... as I do not know whow this person might be. Marcan perhaps?
<sven>
i'm glad that all notification doesn't work on irc :D
<sven>
16k might be a bit faster, but dunno if it's possible to notice that during daily use
<marcan>
does the matrix IRC bridge actually have that turned on by default? that sounds like a terrible idea
<bluetail[m]>
no
<marcan>
good to hear
<Ry_Darcy>
Indeed. A qualified answer would be appreciated nevertheless... mea culpa.
<sven>
uh, did you read what i just wrote?
<bluetail[m]>
Ry_Darcy: wtf state your question better. You look down on us...
<Ry_Darcy>
Did indeed. I look down on no one.
<mps>
Ry_Darcy: 4K is used if you have software or driver which doesn't work with 16K
<bluetail[m]>
ok I just ignore you
<mps>
and 4K kernel is somewhat slower on some tasks
<mps>
well, I'm tempted to ignore all nicks with '[m]' ;)
<Ry_Darcy>
I presume the choice of kernel is then application-dependent. That is what I gathered anyway.
<mps>
jk, not really
<mps>
Ry_Darcy: I know that chromium doesn't work with 16K, and f2fs driver
<psydroid[m]1>
@lol don't ping us @oll
<Ry_Darcy>
I am an old fashioned (Mainframe) performance fan, hence the original question. I presume if I have six Firefox windows open at once - nothing else, I will notice an iota of difference - whether I am using a 4K or 16K enabled kernel.
<j`ey>
Ry_Darcy: exactly
duban[m] has joined #asahi
duban[m] is now known as duban
<Ry_Darcy>
I do not use google or it`s ilk.
<Ry_Darcy>
@j`ey thanks for the confirmation..
<j`ey>
ugh, I forgot these CPUs aren't actually magical, and some programs can still take a long time to compile :)
Ry_Darcy has quit [Remote host closed the connection]
<tpw_rules>
i mean the only time i notice 16k is better is that `time` reports lower numbers when compiling things
rezn1t has quit [Remote host closed the connection]
<j`ey>
bleh, keep getting OOMs building llvm, guess I can't build with -j8, when I only have 8GB
<j`ey>
I thought it might take only 20mins to compile, but with all the crashes and faffing around, it's been 45mins+ :|
kameks has quit [Ping timeout: 480 seconds]
<sven>
just buy a mac studio :>
<j`ey>
;_;
<bluetail[m]>
<mps> "well, I'm tempted to ignore..." <- makes sense. Hopefully people don't infest this place and make it impossible to follow topics / write gibberish / try to use us to spoon feed them
kefu_ is now known as kefu
nicolas17 has joined #asahi
<mps>
bluetail[m]: most annoying for me are replies like your last one
<bluetail[m]>
I dont understand. The OP did ask and got a qualified answer and said it wasnt, then asking for a specific person to answer. That is not nice.
<bluetail[m]>
One way to avoid toxicity is by just ignoring it
<mps>
bluetail[m]: your message is `18:02 ................... bluetail[m]| <mps> "well, I'm tempted to ignore..." <- makes sense. Hopefully people don't infest this place and make it impossible to follow topics / write gibberish / try to use us to spoon feed them`
<psykose>
yeah and that looks fine
<mps>
psykose: lol
<bluetail[m]>
Or even people who don't contribute to the topic at all.
<bluetail[m]>
I said that because it is much more likely to be connected to people who are not really into the topic when the level of entry is massively lowered
<mps>
bluetail[m]: it is annoying on irc
<bluetail[m]>
So it already had happened. I see... THats unfortunate. I really hope chats stay readable - and not with massive activity where there are burts of one message every second
<bluetail[m]>
s/one/two/, s/message/messages/
<mps>
ok
<j`ey>
sven: the macstudio is looking even more tempting since I have to rebuild LLVM with a patch and different options :|
MajorBiscuit has joined #asahi
<sven>
I suffer the same temptation and I don’t even build llvm :D
<j`ey>
Im hoping jannau can give us his thoughts at some point, on if the studio is worth it :P
svos[m] has joined #asahi
psykose has quit [Remote host closed the connection]
psykose has joined #asahi
Tano70 has joined #asahi
<mattebullet[m]>
I was trying to setup qemu usermode emulation for a amd64 program. I was deboostrapping ubuntu to setup a chroot environment and I got "qemu-x86_64-static: ./include/qemu/rcu.h:101: rcu_read_unlock: Assertion `p_rcu_reader->depth != 0' failed."
<mattebullet[m]>
Has anyone faced that issue? Any workaround?
user982492 has joined #asahi
daffy99 has joined #asahi
greguu[m] has left #asahi [#asahi]
Tano70 has quit [Quit: Leaving]
daffy99 has quit [Ping timeout: 480 seconds]
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
c10l7 has quit []
c10l has joined #asahi
MajorBiscuit has quit [Ping timeout: 480 seconds]
<nicolas17>
j`ey: distcc/icecream helps a lot :D
<j`ey>
nicolas17: not when I dont really have anything else worth dist-ing with it :P
user982492 has joined #asahi
<mattebullet[m]>
<ey3ball[m]> "Guess I should report a qemu bug..." <- Did you do that? Any update on that?
Mase3206 has joined #asahi
<ey3ball[m]>
<mattebullet[m]> "Did you do that? Any update on..." <- Not yet no, I wanted to see if I could gather some more info by randomly poking the code for a bit, but then the kids woke up :)
<mattebullet[m]>
<ey3ball[m]> "Not yet no, I wanted to see if I..." <- Thanks for the update. I had the same issue.
Mase3206 has quit [Quit: Mase3206]
Mase3206 has joined #asahi
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Mase3206 has quit []
Mase3206 has joined #asahi
MajorBiscuit has joined #asahi
MajorBiscuit has quit [Ping timeout: 480 seconds]
___nick___ has quit [Ping timeout: 480 seconds]
user982492 has joined #asahi
MajorBiscuit has joined #asahi
femto_ has joined #asahi
Mase3206 has quit [Quit: Sorry. I ran out of caffeine.]