marcan changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | 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
Dcow has joined #asahi
Dcow has quit [Quit: My Mac Mini has gone to sleep. ZZZzzz…]
nobodynada has quit [Ping timeout: 480 seconds]
nobodynada has joined #asahi
nobodynada has quit [Ping timeout: 480 seconds]
nobodynada has joined #asahi
bgb__ has joined #asahi
bgb has joined #asahi
bgb_ has joined #asahi
bgb___ has joined #asahi
bgb_____ has joined #asahi
bgb__ has quit [Ping timeout: 480 seconds]
bgb has quit [Ping timeout: 480 seconds]
bgb_ has quit [Ping timeout: 480 seconds]
bgb___ has quit [Ping timeout: 480 seconds]
bgb_____ has quit [Ping timeout: 480 seconds]
riker77_ has joined #asahi
riker77 has quit [Ping timeout: 480 seconds]
riker77_ is now known as riker77
bgb has joined #asahi
bgb_ has joined #asahi
bgb__ has joined #asahi
bgb___ has joined #asahi
bgb____ has joined #asahi
bgb has quit [Ping timeout: 480 seconds]
bgb_ has quit [Ping timeout: 480 seconds]
bgb_____ has joined #asahi
bgb__ has quit [Ping timeout: 480 seconds]
bgb___ has quit [Ping timeout: 480 seconds]
bgb____ has quit [Ping timeout: 480 seconds]
bgb has joined #asahi
bgb_____ has quit [Ping timeout: 480 seconds]
bgb has quit [Ping timeout: 480 seconds]
bgb_ has joined #asahi
bgb has joined #asahi
nobodynada has quit [Ping timeout: 480 seconds]
bgb_ has quit [Ping timeout: 480 seconds]
bgb_ has joined #asahi
bgb has quit [Ping timeout: 480 seconds]
bgb has joined #asahi
nobodynada has joined #asahi
bgb_ has quit [Ping timeout: 480 seconds]
bgb has quit [Ping timeout: 480 seconds]
nobodynada has quit [Ping timeout: 480 seconds]
bgb__ has joined #asahi
bgb has joined #asahi
bgb_ has joined #asahi
bgb___ has joined #asahi
bgb____ has joined #asahi
nobodynada has joined #asahi
bgb__ has quit [Ping timeout: 480 seconds]
bgb has quit [Ping timeout: 480 seconds]
bgb_____ has joined #asahi
bgb_ has quit [Ping timeout: 480 seconds]
dsrt^ has joined #asahi
pwn has joined #asahi
bgb___ has quit [Ping timeout: 480 seconds]
bgb____ has quit [Ping timeout: 480 seconds]
bgb_____ has quit [Ping timeout: 480 seconds]
phiologe has joined #asahi
PhilippvK has quit [Ping timeout: 480 seconds]
dsrt^ has quit [Ping timeout: 480 seconds]
dsrt^ has joined #asahi
skipwich has joined #asahi
marvin24 has joined #asahi
marvin24_ has quit [Ping timeout: 480 seconds]
pwnblanks has joined #asahi
pwn has quit [Ping timeout: 480 seconds]
linearcannon has quit [Read error: Connection reset by peer]
<fridtjof[m]>
(I'd make a PR for the blog post, but not sure what the best way to go about it is, just changing the post without any semblance of history feels wrong
Dcow has joined #asahi
<j_ey>
fridtjof[m]: I ticked off PCIe, since that's in 5.16
pwnblanks has joined #asahi
<j_ey>
and USB2 and Ethernet
<fridtjof[m]>
ah, neat!
<sven>
type c is also "done" for usb2/3 but will need more work for thunderbolt/displayport
<j_ey>
sven: yeah, i wasnt sure about those sub-points on that list
<sven>
the first two are essentially automatic. for the third we won't use extcon but the typec / usb-role-switch infrastructure
<maz>
I need to find a moment to tackle the PCIe hot-plug stuff.
<sven>
and plug detect works for usb2/3 host/device mode but i think the driver needs additional work to also identify thunderbolt/displayport iirc
<maz>
somehow, the PMU was more fun.
<sven>
i'm still not looking forward to the mess that will be thunderbolt
<maz>
indeed. I started reading the USB4 spec, and it's not fun.
<j_ey>
oh, hot plug doesnt work on the PCIe stuff either?
<maz>
j_ey: well, we take an interrupt, and do nothing except on the first probe.
pwn has quit [Ping timeout: 480 seconds]
<maz>
which is fine for USB and Ethernet, less so for wifi, and definitely bad for TB.
pwnblanks has quit [Remote host closed the connection]
pwnblanks has joined #asahi
<sven>
thunderbolt will at least require us to pass through tunables from the ADT to the atc-phy driver because different ones need to be applied whenever USB3 or TBT is enabled, add support for a co-processor, turn the NIH driver to a platform device and somehow tie all that together with pcie and the usb pd chip
sailorek1234 has quit [Quit: sailorek1234]
bgb has quit [Ping timeout: 480 seconds]
bgb has joined #asahi
pwnblanks has quit [Ping timeout: 480 seconds]
<marcan>
sven: oh, USB3 works now?
<marcan>
maz: fwiw on the new machines PCIe devices need to be enabled via SMC
<marcan>
so that's either a driver dep problem (should SMC be a power domain provider?) or a hotplug problem
<sven>
“works”
<sven>
I have some disgusting hack to make it work which is essentially just put the write32s from a hv trace in there
<sven>
but iirc it needed some tunables from sysconf
bgb_ has joined #asahi
<maz>
marcan: depends on what events the PCIe RC gets. if we can react to an interrupt and talk to SMC, that's fine. otherwise, we need to prod it before enabling the link...
<sven>
or maybe that was just for display port :/
<marcan>
maz: the SMC controls power to the devices; without power the devices are dead and there is no link
<marcan>
so the question is how would PM be handled for this
<sven>
that sounds very similar to how WiFi is enabled
<marcan>
yes, WiFi is the same
<marcan>
for some things like USB ports and the SD card reader I don't think we can ever truly power them down, since you need to react to external hotplugs
<marcan>
for WiFi you could power it down when it's off
bgb has quit [Ping timeout: 480 seconds]
<maz>
OK, so this needs to be modelled as a power gate, or something along those lines, I guess.
<marcan>
but I don't think macOS ever turns these things off either
<marcan>
so e.g. if hotplug works we could just have SMC itself have some init list to turn them on and that's that
pwn has joined #asahi
dsrt^ has joined #asahi
bgb has joined #asahi
bgb_ has quit [Ping timeout: 480 seconds]
Dcow has quit [Quit: My Mac Mini has gone to sleep. ZZZzzz…]
pwn has quit [Ping timeout: 480 seconds]
Dcow has joined #asahi
sailorek1234 has joined #asahi
pwn has joined #asahi
<kettenis>
maz: my approach so far has been to add a "pwren-gpios" property, which is what is used in sifive,fu740-pcie.yaml
Dcow has quit [Ping timeout: 480 seconds]
bgb_ has joined #asahi
bgb has quit [Ping timeout: 480 seconds]
<maz>
kettenis: if the SMC access can be modelled as such, I'm good with it.
sailorek1234 has quit []
<marcan>
it's a gpio behind smc, so yes, we can make it a gpio controller
gpanders[m]1 is now known as gpanders
gpanders has quit [Quit: Reconnecting]
gpanders has joined #asahi
gpanders has quit []
gpanders has joined #asahi
dsrt^ has quit [Remote host closed the connection]
bgb has joined #asahi
yuyichao has quit [Ping timeout: 480 seconds]
bgb_ has quit [Ping timeout: 480 seconds]
pwn has quit [Ping timeout: 480 seconds]
yuyichao has joined #asahi
<pitust[m]>
How far does linux get on m1 (non-pro/max)?
<j_ey>
pitust[m]: depends
<pitust[m]>
hmm, okay. what does it depend on?
<j_ey>
what branches/patches you use, what m1 machine you have
<pitust[m]>
ah okay, i understand
<j_ey>
if you need usb3 or usb2, thunderbolt etc
<_jannau_>
with t6000/bringup and the right dtb framebuffer, smp, and pinctrl/usb-pd/usb2 works
waagrr has quit [Read error: Connection reset by peer]
waagrr has joined #asahi
<_jannau_>
should work on all hardware, tested so for only on 14" M1 Pro and Max
<j_ey>
non-pro/max, the older m1s
<_jannau_>
ah, comment still holds except that the code is already in Linus' master branch and working PCI (usb-a and ethernet on the mac mini)
<pitust[m]>
no wifi, i assume?
<j_ey>
wifi works
<pitust[m]>
oh okay
<j_ey>
(with patches)
<nskl>
marcan said in his last stream (toward the end) that there are a few loose ends, that are mostly done, but still need some work. he cites wifi among them. there are also keyboard, touchpad, and installer too need to become a bit more user friendly and figure a way to automaticaly import all the needed firmwares
<nskl>
i've seen report saying that wifi, keyboard and touchpad too, all can be made to work, but they are in separate trees, requires patching or some special steps to have them. a consolidation effort is needed to tie all these little pieces together, to have them included, checked and done.
<nskl>
(not to mention sound support, which is still at POC status, last time i've heard about it).
X-Scale` has joined #asahi
X-Scale has quit [Ping timeout: 480 seconds]
<marcan>
wifi is at the PoC stage in the sense that the way it's currently done is not upstreamable, and the standard for the project is upstreamability
<marcan>
that's basically why I don't run those patches; it means IO
<marcan>
There is one further overarching practical issue with RMS: regardless of what anyone thinks of any particular issue, he just can't shut up. His entire personal website full of political opinions just serves to generate division, if he is considered a leader of the movement. It is just not possible to be an effective leader of a global movement while having an unfiltered firehose of political takes, ...
<marcan>
... because by definion he will clash with almost everyone on some subset of issues. While it is important that we throw our weight behind specific social topics we consider important (e.g. the way I have defended trans rights and promoted trans people on Twitter), a never-ending ticker of hot takes just
<marcan>
oh fuck
<marcan>
I was not intending to paste that
<marcan>
lol
<marcan>
who invented shift-insert as paste anyway? ...
* marcan
wishes he could just disable that, he *never* wants it
<j_ey>
yesterday I pasted a 1000line trace from m1n1 into bash
<marcan>
ahaha
<marcan>
(yes I'm talking about RMS again... someone emailed me about this... at least it's not on twitter now :p)
<j_ey>
ctrl-b ] instead of ctrl-b [ in tmux zz
<brentr123[m]>
Zsh is better then bash
<marcan>
nskl: I was going to say it means I'm motivated to clean this stuff up if nobody else does it first
<j_ey>
marcan: the patches are definitely good enough for tinkering with!
<marcan>
tinkering yes :)
* sorear
imagines you can do something in xkb
<daniels>
you could indeed make insert two-level and make the second level map to NoSymbol
<daniels>
p sure that's in /usr/share/X11/xkb/symbols/pc