#
#asahi-dev · 2026-10-08 07:58
myriaNo problem at all. I was just asking because if you're too busy I'm happy to send it myself if it works better
#asahi-dev · 2026-10-08 07:58
MichaelReeves[m]oh sorry, I've been distracted too much with p3s, I'll test Yuka's one now and then if thats good send to LKML
#asahi-dev · 2026-10-08 07:55
myriaMichaelReeves[m]: What were you plans with the M3 KVM vgic patches?
#asahi-dev · 2026-10-08 07:45
MichaelReeves[m]https://pastebin.com/raw/i2qeh1cd
#asahi-dev · 2026-10-08 07:44
MichaelReeves[m]ill put it on a pastebin
#asahi-dev · 2026-10-08 07:44
MichaelReeves[m]did need 1 line patch to Wine
#asahi-dev · 2026-10-08 07:44
JamesCalligeros[m]Nice
#asahi-dev · 2026-10-08 07:44
MichaelReeves[m]either work
#asahi-dev · 2026-10-08 07:44
MichaelReeves[m]yep tested both that and force 4K mode flag
#asahi-dev · 2026-10-08 07:44
JamesCalligeros[m]Shim launcher?
#asahi-dev · 2026-10-08 07:43
MichaelReeves[m]oh wow, now 32-bit x86 Win32 apps are working (tried a simple console and window app) through Wine+FEX
#asahi-dev · 2026-10-08 02:33
JamesCalligeros[m]my eyes were glazing over and i had a mad migraine last night so i just didnt even bother looking at it
#asahi-dev · 2026-10-08 02:32
JamesCalligeros[m]thought itd be something like this
#asahi-dev · 2026-10-08 02:32
JamesCalligeros[m]lol the fix
#asahi-dev · 2026-10-08 02:32
JamesCalligeros[m]yeah sure go for it!
#asahi-dev · 2026-10-08 02:29
MichaelReeves[m]s/but/put
#asahi-dev · 2026-10-08 02:29
MichaelReeves[m]also, I managed to reproduce the issue (SIGBUS) with a small program on the Android kernel and it happens too, so would it be OK to let the google people know about this by a comment on gerrit or best to leave it as its their project?
#asahi-dev · 2026-10-08 02:28
MichaelReeves[m]I think that just reinforces that chadmed was right and we should probably wait for the google series to get more stable and see what they are going to do with it
#asahi-dev · 2026-10-08 02:27
MichaelReeves[m]the google people made a typo and accidentally but the same thing in both paths of an if/else if branch. it just needed a single-line change to fix and now it doesnt happen anymore: https://github.com/IntegralPilot/linux/commit/7babd7ccbb0d2c68978f2bf4470fc03e97a574a3
#asahi-dev · 2026-10-08 02:27
MichaelReeves[m]also, I reproduced and found the cause of the plasmashell crashes
#asahi-dev · 2026-10-08 02:26
MichaelReeves[m]yuka: rewrote the m4 pro / m5 pro DTs from your link in upstream dchid style, they're here: https://pastebin.com/raw/eSUWyLvy
#asahi-dev · 2026-10-08 02:10
anon__And I'm happy to provide support if anyone wants to test something on m5.
#asahi-dev · 2026-10-08 02:08
anon__I don't use IRC very often, so my response and results may delay. But I will check irc logs so feel free to reach me.
#asahi-dev · 2026-10-08 00:00
anon_And I have a remaining question. I have implemented basic smp support on M5 but encountered some random data aborts. And every time it cames up, fbcon cannot preserve complete logs. How can I capture this error in a reliable way? Using USB UARTs maybe?
#asahi-dev · 2026-10-07 23:54
anon_Anyway we should not touch the power state of nvme controller and apple locks it after it boots.
#asahi-dev · 2026-10-07 23:54
anon_I suggest to find out ways to capture what darwin actually did on. But afaik m1n1 proxy mode is not viable though.
#asahi-dev · 2026-10-07 23:52
anon_For the nvme issue I've encountered, I have some new observations. It seems like we should not reset the controller on M5. Instead there is another path apple is actually using. Haven't work it out completely but I can tell that we have some positive results now.
#asahi-dev · 2026-10-07 18:42
jannausofus: the stream misses audio in the beginning and for most of the discussion. Let's hope the recordings don't have the issue. I don't think there is more information than in the slides
#asahi-dev · 2026-10-07 18:38
sofusjannau: thanks! Will watch later. I did have VV in mind when designing the driver, so porting should be straigtfull
#asahi-dev · 2026-10-07 17:36
jannauand now there is a second report
#asahi-dev · 2026-10-07 16:18
jannausize matches the one from fedora-44-gnome-202609041600.zip, which looks sane and is bit-identical to the one from the kde image. there are multiple installations, also uefi-only ones
#asahi-dev · 2026-10-07 16:13
chaos_princesscheck if the file on the esp is sane?
#asahi-dev · 2026-10-07 16:11
jannaucan anyone think of a reason why chainload of the 2nd stage boot.bin could fail on M3 imacs?
#asahi-dev · 2026-10-07 13:27
Hel3ziyuka: Ah damn, got it, that sucks :( Thanks for explaining though!
#asahi-dev · 2026-10-07 13:26
chaos_princessyep
#asahi-dev · 2026-10-07 13:25
yukaIt was discussed here I think
#asahi-dev · 2026-10-07 13:24
yukasomeone found out the TB5 atcphy is an entirely new coproc?
#asahi-dev · 2026-10-07 13:21
yukaAnd I meant on the M4 two ports were working but the third was not, because I typoed some tipd interrupt
#asahi-dev · 2026-10-07 13:20
yukas-ven could probably trace atcphy next week to get us some more insight, but either way it will require more work
#asahi-dev · 2026-10-07 13:19
yukaIt is much different
#asahi-dev · 2026-10-07 13:19
chaos_princessyes, atcphy is significantly different
#asahi-dev · 2026-10-07 13:19
Hel3ziAh lol, that explains it! Is the USB setup on M4 Pro much different from base M4, or could the nodes be ported over?
#asahi-dev · 2026-10-07 13:15
yukaM4 Pro is not the same
#asahi-dev · 2026-10-07 13:14
yukaM4 Pro DT does not have usb?
#asahi-dev · 2026-10-07 13:14
yukaEhhh
#asahi-dev · 2026-10-07 13:14
Hel3ziyuka: hit that same dead USB port on my M4 Pro. What was the typo in DT? Can test the fix right away.
#asahi-dev · 2026-10-07 13:07
jannaustream audio is fixed
#asahi-dev · 2026-10-07 13:03
linkmauveAlright, I’ll watch the recording later then.
#asahi-dev · 2026-10-07 13:02
jannauno, running problem for lpc streams this year
#asahi-dev · 2026-10-07 13:01
linkmauveCan you hear sound from the video?

plus ›