<tpw_rules>
smart linux usb folks: is there a way to see how the usb subsystem has allocated bandwidth?
timokrgr has quit [Quit: User left the chat]
timokrgr has joined #asahi
<zamadatix[m]>
lsusb -t
<tpw_rules>
that doesn't show more than nominal, does it?
timokrgr has quit [Quit: User left the chat]
timokrgr has joined #asahi
<chadmed>
it shows you the bandwidth for each port
<chadmed>
technically its the negotiated link speed but it serves the same function unless you have a particularly odd setup bottlenecked by cascaded hubs or something
<tpw_rules>
yes that's what i have
<tpw_rules>
people talk about like v4l2 won't let you open a camera if there's not enough remaining usb bandwidth, but what is the usb bandwidth?
<chadmed>
bandwidth allocation has no qos afaik so its just whichever device asked for it first
<chadmed>
if the camera is asking for 12Mbps and theres only 10Mbps left on the bus then v4l2 wont let it open
<chadmed>
i dont think its dynamic beyond that, unless one device stops using the bandwidth it asked for and "returns" it to the root hub
<chadmed>
you might be able to see that happening if you turn on usb debug in the kernel
<tpw_rules>
is there a way to see that the first camera has asked for the 2mbps?
<tpw_rules>
or is that what you meant by "bandwidth for each port"?
<chadmed>
like i said you might be able to see that in the kernel message buffer with usb debug but im not sure if theres any userspace tools that let you easily snoop on usb transactions like that
phiologe has joined #asahi
<tpw_rules>
ok
<chadmed>
it will basically just be the usb driver being like "device 0xFF asked for x Mbps, there is currently x+n Mbps not being used on this root hub by other devices, so give it x Mbps"
<chadmed>
i find it odd that youve managed to saturate an entire hub so hard that a webcam wont work, is it a fancy 4K one or something?
<chadmed>
or is the hub just crap
<tpw_rules>
yeah i need to attach several to the same computer for computer vision stuff
<tpw_rules>
i can get it to work experimentally but it would be nice to know for sure
PhilippvK has quit [Ping timeout: 480 seconds]
confusomu has joined #asahi
confusomu has quit [Remote host closed the connection]
confusomu has joined #asahi
dhromw^ has quit [Ping timeout: 480 seconds]
dhromw^ has joined #asahi
marvin24_ has joined #asahi
marvin24 has quit [Ping timeout: 480 seconds]
dhromw^ has quit [Ping timeout: 480 seconds]
dhromw^ has joined #asahi
dhromw^ has quit [Ping timeout: 480 seconds]
nicolas17 has quit [Ping timeout: 480 seconds]
the_lanetly_052 has joined #asahi
<marcan>
chadmed: this happens a lot to me, USB bandwidth shenanigans are common
<marcan>
I never bothered to look for a way to debug it but I should
<marcan>
tpw_rules: FWIW on USB3 the host controller is in charge of bandwidth management IIRC, so you could even be running into firmware bugs here
<marcan>
(on xHCI I mean)
dhromw^ has joined #asahi
<tpw_rules>
yeah these are all usb3. i just want something more quantifiable than my napkin calculations. i don't think there's any bugs. also understanding how e.g. usb hard drives factor in
dhromw^ has quit [Ping timeout: 480 seconds]
Ry_Darcy has joined #asahi
Ry_Darcy has quit [Remote host closed the connection]
Ry_Darcy has joined #asahi
<marcan>
AIUI bulk devices do not reserve any bandwidth beyond some shared pool?
<marcan>
I imagine there's some minimum reserved for polling packets though
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi
MajorBiscuit has joined #asahi
MajorBiscuit has quit [Ping timeout: 480 seconds]
MajorBiscuit has joined #asahi
Catyre has joined #asahi
Catyre has quit []
Catyre has joined #asahi
Catyre has quit []
Catyre has joined #asahi
test has joined #asahi
test has left #asahi [#asahi]
Catyre_ has joined #asahi
Catyre_ has left #asahi [#asahi]
guillaume_g has joined #asahi
psykose has quit [Ping timeout: 480 seconds]
SiddheshZantye[m] has joined #asahi
<SiddheshZantye[m]>
hi
the_lanetly_052__ has joined #asahi
<SiddheshZantye[m]>
asahi wont install for me :(
<SiddheshZantye[m]>
I get this error. Anyone know what it means?
<SiddheshZantye[m]>
rowang077[m]: will that be enough for it to work?
<rowang077[m]>
so you need whatever size you choose for linux + 38GB slack
<SiddheshZantye[m]>
oh I see
<rowang077[m]>
yes 15GB is minimum for it to work according to the tutorial.
<SiddheshZantye[m]>
will I be able to add more later?
<rowang077[m]>
but you wont have much space in linux do to stuff
<rowang077[m]>
so how much space is enough depends on what you want to do use linux for
<rowang077[m]>
I don't know if you can add more later. I actually don't have an M1 mac. I'm just following along here :p
<SiddheshZantye[m]>
lol
<SiddheshZantye[m]>
Ill try with 220
<SiddheshZantye[m]>
that would give it 25 I think
<SiddheshZantye[m]>
just tried it and that failed too
ggardet has joined #asahi
guillaumeg has joined #asahi
guillaume_g has quit [Ping timeout: 480 seconds]
<rowang077[m]>
can you send me your entire log of the install process?
ggardet has quit [Ping timeout: 480 seconds]
<ruffiely[m]>
will the m2 also work with asahi or does it require complete new code? :D
<j`ey>
ruffiely[m]: not completely new code, but some fixes and tweaks and probably a new driver ortwo
<chadmed>
i think they integrated HID on the laptops so that it no longer goes over SPI so thats going to need some tweaking
<j`ey>
(for the keyboard/trackpad)
<rowang077[m]>
I wonder how do we know/expect this? I mean the hardware is not out yet
<chadmed>
we've seen the device tree for the macbook air
eriol2 has quit []
eriol2 has joined #asahi
eriol2 is now known as eriol
eriol has quit []
eriol has joined #asahi
dhromw^ has joined #asahi
<Sobek[m]>
<chadmed> "we've seen the device tree for..." <- Wait how ?
<j`ey>
from the macos13 beta
<Sobek[m]>
Is it in macos betas / latest macos 12 ?
<chadmed>
yeah
___nick___ has joined #asahi
___nick___ has quit []
___nick___ has joined #asahi
ptudor_ has joined #asahi
ptudor has quit [Ping timeout: 480 seconds]
roxfan has joined #asahi
erincandescent has quit [Remote host closed the connection]
erincandescent has joined #asahi
lonjil has joined #asahi
<zx2c4>
hmm, any known limitations on what KVM can boot on ashai?
<zx2c4>
Does 16k page limitation trickle down into VMs?
<maz>
no
<maz>
no 64k pages though, the HW doesn't support them.
<zx2c4>
i'm tryna run windows in qemu/kvm and i cant even get past uefi initialization
<maz>
same thing here. it seems they are trying to use features that don't exist on the apple HW, such as SMC or the architected PMU.
<j`ey>
you could pass '-d int', it might be very spammy, but could possibly show if it was executing an SMC or what the last 'bad' interrupt/execption was
<maz>
j`ey: this is only valid of TCG, I think. I can't see how this would work for KVM.
<j`ey>
oh yeah, my bad.
<maz>
for example, an SMC will result in an EL1 exception, and won't be seen by KVM at all.
<zx2c4>
maz: j`ey: i see https://א.cc/kIpFrEIZ for example
<zx2c4>
"ConvertPages"
<zx2c4>
i know peopl ehave gotten qemu+hvf working on macos though
<zx2c4>
with uefi
<maz>
" ESR : EC 0x00 IL 0x1 ISS 0x00000000 "
<zx2c4>
so i'm not sure its the cpu
<maz>
undefined instruction.
<zx2c4>
O_o
<zx2c4>
oh.
<zx2c4>
Well then. How are people running this on hvf?
<maz>
you can do all sort of things if you want to sidestep the architecture, which HVF does.
<zx2c4>
Does that somehow trap and emulate?
<zx2c4>
ahh
<maz>
I'm not interested in supporting that in KVM.
<zx2c4>
Why not?
<maz>
because KVM/arm64 is architectural, and only that.
<zx2c4>
I dunno about that...
<zx2c4>
ever see ppc's kvm implementation for example?
<zx2c4>
It handles lots of interesting things
<maz>
I only care about the stuff I maintain upstream. other architecture can pile whatever crap they want, that's none of my business!
<zx2c4>
my point is that if your point is "KVM isnt supposed to do that", here's a counter example
<zx2c4>
why not make KVM more useful?
<maz>
why not make KVM more crap and unmaintainable, you mean?
<maz>
we already have to deal with a 10k page spec. adding support for something *that doesn't have a public spec* is definitely not on my radar.
<j`ey>
maz: im assuming that hvf (non architectural) will use this trap? ^
<zx2c4>
maz: anyway, if that's your perspective, it sounds like i'll never have windows running inside kvm on the m1. so i'll just sell the hardware because that's an uphill battle against "the stuff [you] main upstream" that i dont really want to fight
<zx2c4>
maintain
<maz>
j`ey: probably. assuming that's only for SMC.
<maz>
zx2c4: hey, feel free to go and maintain KVM/arm64 if that's what you want.
<zx2c4>
surely you're not offering to change MAINTAINERS after a short random conversation in an irc channel. don't be ridiculous
<maz>
well, if you think you can do a better job, I don't care.
<zx2c4>
my point is that i *dont* really want to fight you on this. so if that's a hardline perspective you take, it's obvious to me the course of action that's most reasonable is to find more compliant hardware
<zx2c4>
1) i need a windows arm64 reverse engineering platform; 2) i'll use an m1 for that because it's fast and available; 3) maz says no; 4) reevaluate step 2
<maz>
one thing that *could* be done is to forward the stuff that isn't architectural to userspace and let it deal with it. Not sure you could work around everything, but I guess that's an interesting project for someone willing to mess with that sort of stuff.
<sven>
i think that's more or less what hvf.framework on macOS does fwiw. most of the heavy lifting is done inside virtualization.framework which runs in userspace
<maz>
the crap with this is that userspace doesn't really know what sort of braindead extension it has to deal with, which negates the idea of the usrspace ABI being an abstraction layer.
<maz>
now userspace has to know "oh, I run on this particular flavour of fruity computer, let's interpret this random syndrome in a particular way".
<maz>
great for a vertical ecosystem, but pretty bad for something like Linux.
<sven>
oh, i agree that this isn't a great idea but it might be the best that can be done
pjakobsson has joined #asahi
miko has quit [Quit: WeeChat 3.5]
scoobydoo has joined #asahi
<scoobydoo>
channel list
<scoobydoo>
exit
scoobydoo has quit []
scoobydoo has joined #asahi
<scoobydoo>
hi
Ry_Darcy_ has joined #asahi
<scoobydoo>
is there anyone here?
marvin24 has joined #asahi
<scoobydoo>
hello?
marvin24_ has quit [Ping timeout: 480 seconds]
<scoobydoo>
?
<mps>
iiuc problem with qemu is that windows x86_64 guest can't boot
<scoobydoo>
mps can u read my mexes?
<mps>
mexes?
<scoobydoo>
messages
<scoobydoo>
nice
<mps>
ah, yes
<scoobydoo>
i thought i was messing some
<scoobydoo>
thanks
<scoobydoo>
so that s the channel for general questions?
<mps>
scoobydoo: /who should show users in channel
<scoobydoo>
it s not working for me
<scoobydoo>
like it takes the command but shows no output
<mps>
linux x86_64 guest works in qemu
scoobydoo has quit [Quit: leaving]
scoobydoo has joined #asahi
<scoobydoo>
\
<mps>
scoobydoo: what is your irc client
<scoobydoo>
irssi
<scoobydoo>
should i change it?
<mps>
scoobydoo: also I use irssi
<mps>
1211
<scoobydoo>
?
<mps>
12
<scoobydoo>
what is 12?
<mps>
uhm sorry. typed wrong control seqeunces
null_ is now known as null
<mps>
scoobydoo: /who works for me
<scoobydoo>
i have a full screen on this caht, where do you see the output, same page where u write text or should i change some?
* mps
wonders how much users are on channel
<mps>
scoobydoo: /who output goes to status channel, usually number 1
<j`ey>
scoobydoo: this is not a general help chat, it's a channel about the asahi project
<scoobydoo>
yeah u right j sorry, first time on irc, sorry
dhromw^ has quit [Ping timeout: 480 seconds]
Gaspare has joined #asahi
Ry_Darcy_ has quit [Remote host closed the connection]
<scoobydoo>
does anybody understood how to increase luminosuty on asahi linux on macbook pro 13 with no phisycal fucntion keys, only way i managed to increase it is to switch to macos increase it and get back
<j`ey>
scoobydoo: that's the only way for now
<scoobydoo>
ok cool
<scoobydoo>
let me ask another question, is it normal that my mac gets hot while using asahi?
<scoobydoo>
like is there any way to reduce it, i can t make the fan get to work
<AdryzzOLEDEdition[m]1>
it is normal as all the rendering is done on the CPU
<AdryzzOLEDEdition[m]1>
once it gets offloaded to the GPU it should be better
<scoobydoo>
that make sense, thanks adry
scoobydoo has quit [Quit: leaving]
luke has joined #asahi
marvin24_ has joined #asahi
marvin24 has quit [Ping timeout: 480 seconds]
psykose has joined #asahi
Gaspare has quit [Quit: Gaspare]
bisko has joined #asahi
asdflkjasdflkjrhlktrjhelkgjhdf has joined #asahi
luke has quit [Read error: Connection reset by peer]
<asdflkjasdflkjrhlktrjhelkgjhdf>
test
<asdflkjasdflkjrhlktrjhelkgjhdf>
anyone there?
<jnn>
yes
<jnn>
lots
<asdflkjasdflkjrhlktrjhelkgjhdf>
sorry this is my first time using matrixz
<asdflkjasdflkjrhlktrjhelkgjhdf>
matrix*
<lkvrsfld[m]>
Anyone already tested the new macos 13 beta and used the rosetta-mount for linux VMs?
<jnn>
the general etiquette is: 1. ask a question, 2. stay around for answers
MajorBiscuit has quit [Quit: WeeChat 3.5]
<asdflkjasdflkjrhlktrjhelkgjhdf>
will new chips (M2, M2 Pro, &c) require tweaks to asahi to work on machines with those chips or not bc they use the same instruction set?
<AdryzzOLEDEdition[m]1>
they will for a few things like input
<j`ey>
asdflkjasdflkjrhlktrjhelkgjhdf: some tweaks and a few new drivers
<asdflkjasdflkjrhlktrjhelkgjhdf>
is the same thing needed for those updating a normal linux distro to work on newer generations of other chips (e.g. intel 11th gen to intel 12th gen)?
<j`ey>
the m2 is a whole SoC, not just a CPU
<asdflkjasdflkjrhlktrjhelkgjhdf>
* 12th gen)? or is this only because of how asahi is working with apple's opaque software?
<asdflkjasdflkjrhlktrjhelkgjhdf>
ahh ok
<asdflkjasdflkjrhlktrjhelkgjhdf>
ty
<j`ey>
there may be some small tweaks related to the CPU itself, but overall that will roughly work the same, it's just some of the devices (like keyboard/trackpad) that need new work
<asdflkjasdflkjrhlktrjhelkgjhdf>
would that be difficult/take lots of time or not bc writing drivers for a keyboard is not as complicated as for graphics?
<threerik[m]>
how would I go about updating m1n1 using the installer script? more specifically what do i do here
<_jannau_>
asdflkjasdflkjrhlktrjhelkgjhdf: considerably less, also most likely it is just the transport how HID events are signalled from the keyboard controller
<_jannau_>
threerik[m]: why do you want to update m1n1 via the installer? I think m1n1 in the installer wasn't even updated yet
<threerik[m]>
im using the dev script, i thought the fixes for macOS 13 were pushed to it? or i have i misunderstood
<_jannau_>
there should be no need to update m1n1 vias the installer
<threerik[m]>
oh 🤔, how would i get everything working again? the old version doesnt work with macOS 13, so i cant boot into linux to update via pacman
<_jannau_>
I don't think the installer is updated yet but it won't fix the problem (and leave the existing installation intact)
<threerik[m]>
oh okay i see
<threerik[m]>
ill wait a bit longer then, thanks :)
<_jannau_>
I'll prepare instructions later how to restore the system to working state with macos 13 beta
guillaumeg has quit []
<threerik[m]>
alright, thanks a lot
<asdflkjasdflkjrhlktrjhelkgjhdf>
thank you all for answering my questions and for your work on the asahi project!
Gaspare has joined #asahi
<mps>
is the DCP merged to asahi kernel repo
nicolas17 has joined #asahi
<mps>
and I see kernel is rebased to 5.19 but don't see which rc
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<mps>
ok, I don't see DCP so it is not merged
<j`ey>
nah it's still 5.18 based
<mps>
ah yes, I'm blind
<mps>
so no reason to rebuild and upgrade I guess
<j`ey>
unless someone has a mac studio and wants to run alpine
null has quit [Quit: oh no.]
<j`ey>
(or wants to upgrade to macos 13)
null has joined #asahi
<mps>
j`ey: is macos 13 stable released or it is beta
<nicolas17>
macos 13 beta 1
<nicolas17>
stable will come in like september
<mps>
nicolas17: aha, I will not upgrade then, at least for now. thanks
<mps>
sent yesterday my 'destroyed' mbp to service, when it come back I could use one for experimental things
millenialhacker has joined #asahi
luke has joined #asahi
Gaspare has quit [Quit: Gaspare]
Gaspare has joined #asahi
qeeg has quit [Ping timeout: 480 seconds]
qeeg has joined #asahi
coderobe has quit [Quit: Ping timeout (120 seconds)]
coderobe has joined #asahi
gregb[m] has joined #asahi
___nick___ has quit [Ping timeout: 480 seconds]
<tpw_rules>
what is some good software to make a disk speed vs time graph? interested in filling an SSD up from empty and measuring that
the_lanetly_052__ has quit [Remote host closed the connection]