ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | Non-development talk: #asahi | General development | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-dev
sawyer has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sawyer has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
sawyer has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sawyer has joined #asahi-dev
sawyer has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<ah-> hmm, if I add this isp_dart section to my dtb any kernel stops booting: https://gist.github.com/ah-/3dc3f8f1724ed2583eb362dd5c3d68e6
<ah-> not sure why, the address looks similar to what's in the m1n1 trace. i can't see the interrupt anywhere in the trace though
<eiln> ah-: the dart address is wrong. it was supposed to be a 0x3 I think.
<eiln> looking at it now
Emantor has quit [Quit: ZNC - http://znc.in]
Emantor has joined #asahi-dev
<ah-> ha! yes you're right, I missed that the 0x2/0x3 in reg also contains the address
<ah-> boots now, but getting "deferred probe pending" for the power controllers
<eiln> ah-: are you running these under proxy? I did push a patched dt rn
<ah-> atm just by switching m1n1.bins, got proxy running too but more comfortable this way. my usb cable is a bit short
<eiln> ah-: got it. you will also need to chainload/update m1n1 with my branch, which should be right there :D
<eiln> yes
<ah-> getting FDT: 'sio' not found, 'sio' node not found, ADT: '/arm-io/isp' node not found, '/arm-io/isp' not found and then failed to prepare FDT
<eiln> ah-: ah, forgot the alias. brb
<eiln> sio is a diff hw
<eiln> ah-: pushed. the initial iommu stuff is different, and you might get translation faults / stuck dart error handler. looking into that
<ah-> what branch/repo?
<eiln> the dt in linux
jeisom has joined #asahi-dev
<eiln> ah-: whoops, actually pushed this time
gabuscus has quit []
hightower3 has joined #asahi-dev
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<ah-> hmm no luck. getting late so i'll try more tomorrow, but looks like in the adt it's isp0 not just isp
hightower4 has quit [Ping timeout: 480 seconds]
<eiln> ah-: yeah, that makes sense. ill fix that.
jeisom has quit [Ping timeout: 480 seconds]
ourdumbfuture has joined #asahi-dev
eiln has quit [Quit: WeeChat 4.0.3]
eiln has joined #asahi-dev
<ah-> ok with all /arm-io/isp replaced by /arm-io/isp0 it boots, but still "[ 15.260651] platform 292280000.power-management:power-controller@4010: deferred probe pending" in dmesg
<ah-> more tomorrow
gabuscus has joined #asahi-dev
sawyer has joined #asahi-dev
<eiln> sigh, I found it.
<eiln> isp's power domain hierarchy: 0x00 -> 0x10 -> 0x08 -> 0x28 -> 0x20 -> 0x30 -> 0x18 -> 0x38 -> 0x40
sawyer has quit [Ping timeout: 480 seconds]
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
c10l1 has joined #asahi-dev
c10l has quit [Read error: Connection reset by peer]
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
<chadmed> hmm 15" air speaker routing is wrong
<chadmed> all channels go to one of the left woofers
<chadmed> none of the kcontrols are disabled either so i assume we're just not handling the device properly in macaudio
dimilar has joined #asahi-dev
<chadmed> nvm it was just kde's volume knobs still being broken. we do need to do kcontrol fixup in macaudio though
sefidel has quit [Remote host closed the connection]
sefidel has joined #asahi-dev
dimilar has quit [Quit: ERC 5.5.0.29.1 (IRC client for GNU Emacs 29.1.50)]
sternez has joined #asahi-dev
sternez has quit [Remote host closed the connection]
sternez has joined #asahi-dev
<sternez> jannau: I woule like to make my m2 ultra's hdmi work. I noticed you've achieved big progress on m2 pro mini and studio. I can do some tests on my m2 ultra studio.
<sternez> jannau: your pull request in AsahiLinux/m1n1 mentioned that: "M2 Ultra Mac Studio and Mac Pro hopefully work after replacing dcp0/disp0 with dcpext0/dispext0.". I found that is not enough, on M2 ultra, there are no '/arm-io/lpdptx-phy0' and '/arm-io/dp2hdmi-gpio0'. but I found they are '/arm-io/lpdptx-phy1' and '/arm-io/dp2hdmi-gpio1'. after the modification, I still can not make it work.
elvishjerricco has quit [Read error: Connection reset by peer]
elvishjerricco has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
sternez has quit [Remote host closed the connection]
sternez has joined #asahi-dev
<jannau> sternez: indeed, they use die1 on the m2 ultra mac studio, sigh. you need to use dcpext4/dispext4
sternez has quit [Remote host closed the connection]
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
stipa has joined #asahi-dev
jacksonchen666 has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
crabbedhaloablut has joined #asahi-dev
WindowPa- has joined #asahi-dev
WindowPain has quit [Ping timeout: 480 seconds]
duhanebel has quit [Quit: Textual IRC Client: www.textualapp.com]
knedlik has joined #asahi-dev
yaboishady has quit []
yaboishady has joined #asahi-dev
yaboishady8 has joined #asahi-dev
yaboishady has quit [Read error: Connection reset by peer]
yaboishady8 is now known as yaboishady
yaboishady has quit []
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
loki_val has joined #asahi-dev
crabbedhaloablut has quit [Read error: Connection reset by peer]
jacksonchen666 has quit [Remote host closed the connection]
jacksonchen666 has joined #asahi-dev
jacksonchen666 has quit [Remote host closed the connection]
jacksonchen666 has joined #asahi-dev
tanty has quit [Quit: Ciao!]
midou has quit [Ping timeout: 480 seconds]
tanty has joined #asahi-dev
Luigi has joined #asahi-dev
midou has joined #asahi-dev
akemin_dayo has joined #asahi-dev
<knedlik> So now that I have the development kernel, and everything seems working, are there any other things where I can help out (maybe with programming) other than the yaks?
<knedlik> I'm generally interested in working towards the direction of helping with the GPU stuff
tanty has quit [Quit: Ciao!]
roxfan has joined #asahi-dev
knedlik has quit [Quit: knedlik]
tanty has joined #asahi-dev
Luigi has quit [Quit: Leaving]
<ah-> eiln: i'm testing now with your latest pmgr reordering in the dtb and this m1n1: https://github.com/ah-/m1n1/commits/isp-dapf
<ah-> chainloading m1n1 and linux booting with your tracer gets me this log: https://textbin.net/raw/mq3gbc3knr
<eiln> ah-: looks like the device wasn't even probed. did you uncomment the compatible under apple_isp_of_match in isp-drv.c? also you should use the second console (with the kernel msgs) if you haven't (see https://github.com/AsahiLinux/docs/wiki/Tethered-Boot-Setup-%28For-Developers%29#establishing-the-usb-connection)
<ah-> and here's the full dt from m1n1 in case it's useful: https://paste.centos.org/view/49860e7a
<ah-> I think I did enable it, this is the exact kernel i'm running: https://github.com/ah-/kernel-asahi/commits/cam1
<ah-> ooh dapf: Error getting DAPF /arm-io/dart-isp0 base address.
jeisom has joined #asahi-dev
<eiln> ah-: where do you get the dapf message?
<eiln> ah-: in https://github.com/ah-/m1n1/commit/5fa32d4ec71b95fb5f539ade42e5d19c066fc234 you copied 6 in place of 5 in src/dapf.c
ourdumbfuture has joined #asahi-dev
<ah-> yep, i'll try with 5 now. i tried 6 because it looked like there were 6 entries in reg
<ah-> saw it in the picocom-sec.sh output I think
<eiln> it' 6 so index should be 5)
<eiln> yeah, the stage 2 m1n1 output & later kernel dmesg should pop up on the second picocom console. use the tools/picocom script
lynndotpy has quit [Quit: bye bye]
lynndotpy has joined #asahi-dev
<ah-> ok, with 5 that error is gone, getting this trace: https://textbin.net/raw/fbzuxnbvri
<eiln> I see dapf & dart-isp is initialized. are you getting to the kernel? what do you get in the picocom console?
<ah-> and the corresponding trace: https://paste.centos.org/view/99d40b99
<ah-> that trace looks identical to the previous one though, apart from some timestamps/addresses changing
<eiln> ah-: sorry, can you try the updated dt?
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
ourdumbfuture has joined #asahi-dev
<ah-> hah, progress! eiln: https://paste.centos.org/view/c6e5de0b
<ah-> it's probing now and failing with "apple-isp 384000000.isp: failed to map reserved region"
<ah-> oh do I need the firmware now? I think I saw something about firmware extraction and I'm probably missing that
<eiln> no, that's the optional calibration files. this is an iommu error, give me a minute
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
<eiln> ah-: in the meantime can you run it again with the last linux commit?
<ah-> yep let me try
<jannau> and m2 pro mac mini has a picture on hdmi as well
<jannau> sternez: in the case you read this, you will have to change the write to "0x39c000010 (lpdptx-phy0[1]" at the beginning from 0x0 to 0x1
<eiln> ah-: found a possible culprit in m1n1, can you try with the recent push
<ah-> eiln: i had to change two more isp to isp0 on top of your branch: https://github.com/ah-/m1n1/commit/8f68d46d9ea0e06457ae375e5ce5dcca4ff20ede
<ah-> and looks like the same "failed to map reserved region" with new m1n1. haven't managed to fully get the new kernel with extra logging on boot yet, i'll try that next
<eiln> wait, can you add one more line to m1n1
<eiln> pushd
<ah-> hmm, don't see any new m1n1 commits
<ah-> ah now
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<eiln> is the m1 pro '/arm-io/node0' name an issue for any other hw? I'm thinkng of patching adt instead
<ah-> still looks the same, i'll run with the dbg print
<eiln> ah-: can you attach another reg trace too?
ourdumbfuture has joined #asahi-dev
<jannau> what do you mean by issue? yes, adt nodes have number suffixes on on t600x/t602x. yes, it is annoying but I don't think that patching the adt will be an improvement
<eiln> e.g. dapf.c
<jannau> a possible solution for nodes which exists only once could be an adt_path_offset that matches the prefix and returns the device if the match is unambigguous
cylm has joined #asahi-dev
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<marcan> pushed kernel/mesa updates (to dev and prod, passed a basic smoke test)
ourdumbfuture has joined #asahi-dev
eiln has quit [Quit: WeeChat 4.0.4]
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
ourdumbfuture has joined #asahi-dev
chadmed has quit [Read error: Connection reset by peer]
chadmed has joined #asahi-dev
cylm_ has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
cylm_ has quit [Quit: WeeChat 4.0.0]
utsweetyfish has joined #asahi-dev
chadmed has quit [Read error: Connection reset by peer]
chadmed has joined #asahi-dev
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
ourdumbfuture has joined #asahi-dev
roxfan has quit [Ping timeout: 480 seconds]
roxfan has joined #asahi-dev
nyx_o has quit [Ping timeout: 480 seconds]
nyx_o has joined #asahi-dev
jacksonchen666 has quit [Remote host closed the connection]
jeisom has quit [Ping timeout: 480 seconds]
nyx_o has quit [Ping timeout: 480 seconds]
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
leah2 has quit [Ping timeout: 480 seconds]
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
chadmed has quit [Ping timeout: 480 seconds]
jacksonchen666 has joined #asahi-dev
sawyer has joined #asahi-dev
ourdumbfuture has joined #asahi-dev
leah2 has joined #asahi-dev
chadmed has joined #asahi-dev
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
dabaum|2er0 has joined #asahi-dev
sawyer has quit [Quit: Textual IRC Client: www.textualapp.com]
jeisom has joined #asahi-dev
ourdumbfuture has joined #asahi-dev
<ChaosPrincess> eiln: when macos initializes the camera, it does a series of writes that look somewhat like this: https://paste.bingner.com/paste/4tv55 (two similar sequences to pmgr range, and setting a bitmask for every sequence written.) i have noticed that your 1.5fps script does not do those writes. are they not required? or am i blind and those are actually there?
<ah-> also I got a bit further with debugging my iommu map crash. so it fails on the asc-firmware map: [ 5.329788] apple-isp 384000000.isp: [apple_isp_resv_region] reserving: 0: phys: 0x10000a58000 size: 0x980000 iova: 0x0
<ah-> this address comes from segment-ranges in the adt, and I double-checked that the adt hex string does decode to exactly that address
<ah-> the address seems low to me, do we maybe have to adjust it? similar to how we had to 2| the overall isp address in the dt?
jacksonchen666 has quit [Ping timeout: 480 seconds]
<jannau> ah-: which address seems low to you? "iova: 0x0"?
<ah-> 0x10000a58000
<ah-> everything i see in the trace starts with 2/3
<ah-> i was suspicious of the 0 too, but I think for the iommu it's ok?
<jannau> yes, for the iommu/iop 0x0 is ok
<jannau> 0x10000a58000 is low considering that the memory is mapped to 0x100_0000_0000. I would expect that there's memory-region with the same address, size in the adt fetched from the device
<jannau> do you see it in the output of `experiments/dart_dump.py dart-isp0`?
<ah-> when bootiing macos? I can try in a minute
<ah-> here's the adt: https://paste.centos.org/view/dd70d76e the addr comes from ^I^I(text_phys, text_iova, _, text_size, data_phys, data_iova, _, data_size) = struct.unpac ...: k('<QQQI4xQQQI4x', b[:64])
<ah-> of isp0 segment-ranges
<jannau> no, when m1n1 is running in proxy mode
jacksonchen666 has joined #asahi-dev
chadmed has quit [Quit: Konversation terminated!]
chadmed has joined #asahi-dev
<ah-> overall dart-dump: https://paste.centos.org/view/df1315c9 I get an exception when trying to dump dart-isp0: https://paste.centos.org/view/d083e350
<ah-> there's read of a similar looking addr in the macos trace: [cpu0] [0xfffffe0013733aac] MMIO: R.8 0x385050000 (isp0[0], offset 0x1050000) = 0x102010000a58001
<jannau> so the actual mapped memory in dart-isp is from the top of the memory
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
cylm_ has joined #asahi-dev
cylm has quit [Ping timeout: 480 seconds]
<ah-> hmm, maybe the dart_dump.py dart-isp0 error means that the issue is not the address but the dart it goes through not being setup right yet
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<ah-> haha! dart_dump works if I first run p.pmgr_adt_clocks_enable(f'/arm-io/dart-isp0')
ourdumbfuture has joined #asahi-dev
<ah-> though it has a suspicious number of 0s in there: https://paste.centos.org/view/f41a47fb
ourdumbfuture has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
loki_val has quit []
ourdumbfuture has joined #asahi-dev
darkapex has quit [Remote host closed the connection]
darkapex has joined #asahi-dev
chadmed has quit [Ping timeout: 480 seconds]
<ah-> hmm, if I manually enable those clocks I can iomap_at the fw addresses in a m1n1 shell. maybe it's just that the clocks aren't enabled in linux
ellyq has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev
utsweetyfish has quit [Remote host closed the connection]
utsweetyfish has joined #asahi-dev