#asahi-dev · 2026-05-04 06:29
yukain src/dapf.c#asahi-dev · 2026-05-04 06:28
yukaI am able to get the mtp_dart working by commenting out /arm-io/dart-aop and /arm-io/dart-isp#asahi-dev · 2026-05-04 06:27
yukaThe Neo SErrors in dapf_init_all#asahi-dev · 2026-05-03 21:27
jannauany idea why power-states are turned of despite pd_ignore_unused? turning power domains of the lpddr interfaces off has funky effects#asahi-dev · 2026-05-03 21:11
svenand maybe even patches for tipd, don't remember#asahi-dev · 2026-05-03 21:10
svenyeah, there's some initial spmi driver somewhere#asahi-dev · 2026-05-03 20:17
yukaHas anyone looked at how to use the new spmi based usb in host mode under linux on M3+?#asahi-dev · 2026-05-03 17:18
svenand maybe we need to specifically handle that#asahi-dev · 2026-05-03 17:17
svenoh... endpoint 0 probably has a max packet size of 64 bytes though#asahi-dev · 2026-05-03 17:16
sven*thought#asahi-dev · 2026-05-03 17:16
sven(i.e. request 64 bytes, notice it's longer, request >64 bytes again and we don't handle that correctly)#asahi-dev · 2026-05-03 17:16
svenmight just be the host stack doing something unexpected like only requesting 64 bytes because i through that those descriptors could be longer#asahi-dev · 2026-05-03 17:14
jannau"m1n1 uartproxy v1.6.0-rc1-dirt" works#asahi-dev · 2026-05-03 17:13
jannau"m1n1 uartproxy v1.6.0-rc1-dirty" is 32 chars so the descriptor length is over 64 bytes#asahi-dev · 2026-05-03 17:11
LarstiQsurely no limit on the number of components separated by `-`?#asahi-dev · 2026-05-03 17:10
jannauafaik the strings fields can be 255 bytes as utf-16 so that shouldn't be the issue#asahi-dev · 2026-05-03 17:08
svenor maybe it has to be split into two requests and our dwc3 code just doesn't do that correctly? :D#asahi-dev · 2026-05-03 17:07
svenmaybe there's some limit in either the usb spec or in the usb host stack? but that's weird..#asahi-dev · 2026-05-03 17:07
svenhuh#asahi-dev · 2026-05-03 16:55
jannaureproduced the USB issue. apparently caused having "v1.6.0-rc1-dirty" as BUILD_TAG instead of "v1.6.0-rc1". The build tag is added to the product descriptor. not sure why a longer string would cause this issue#asahi-dev · 2026-05-03 13:01
JamesCalligeros[m]also requires some devicetree changes but those are trivial #asahi-dev · 2026-05-03 13:01
JamesCalligeros[m]build-tested. the main changes were runtime-tested at time of ML submission. the fixups at the tip of the tree have not been tested#asahi-dev · 2026-05-03 13:00
JamesCalligeros[m]jannau: https://github.com/chadmed/linux/commits/new-audio/#asahi-dev · 2026-05-03 11:36
fl0_idjannau yeah that has been my experience too mit m3 max and looking at m3 pro trees#asahi-dev · 2026-05-03 10:21
jannausigh, ideal time for freedesktop-sdk build infra to break down: https://gitlab.com/freedesktop-sdk/freedesktop-sdk/-/work_items/1985#note_3309172658#asahi-dev · 2026-05-03 09:44
svenBut using another macOS with 26+ works #asahi-dev · 2026-05-03 09:44
svenI’ve been too busy with $work :(#asahi-dev · 2026-05-03 09:40
yukaI saw there is the new command in tuxvdmtool to put the target into debugusb mode. Is there anything public already for using the debugusb?#asahi-dev · 2026-05-03 09:25
jannauthis makes using t603x a little strange. I think I'll just use t6031 even for the multi die stuff#asahi-dev · 2026-05-03 09:22
jannaut6030 (M3 Pro) looks more like an extended M3 or prototype for M4 (6 e-cores, 2 dcpext) than a chopped down M3 Max. it even uses parent_addr 0x210000000 for MMIO instead of 0x200000000 for t603[124]#asahi-dev · 2026-05-03 08:52
nickchanthe first two regs of the /arm-io/pmgr node should point to two regions filled with state registers so that's one location to start#asahi-dev · 2026-05-03 08:45
nickchanpmgr's adt for m4/a18 since ps-regs is missing#asahi-dev · 2026-05-03 08:44
nickchanwe just don't know how to parse adt#asahi-dev · 2026-05-03 08:44
freetolgaor has anyone tried adapting m3 dts#asahi-dev · 2026-05-03 08:44
nickchanno#asahi-dev · 2026-05-03 08:44
freetolgais it known that m4 and a18 pro pmgr needs a modified driver?#asahi-dev · 2026-05-03 08:42
nickchanat least, it didn't work when I told chaos_princess to try to modify the cpu pstate definitions and I managed to reproduce that behavior later as well#asahi-dev · 2026-05-03 08:41
nickchanthe stuff in pmgr that could actually cause damage seem to be locked by iboot on m1#asahi-dev · 2026-05-03 08:39
chaos_princessyou will not set your device on fire#asahi-dev · 2026-05-03 08:39
chaos_princessit is riskless to try#asahi-dev · 2026-05-03 08:39
freetolgaif it was riskless to try i would try until i get something done#asahi-dev · 2026-05-03 08:38
freetolgapmgr is probably not my thing tho#asahi-dev · 2026-05-03 08:37
JamesCalligeros[m]easy peasy#asahi-dev · 2026-05-03 08:37
jannauJamesCalligeros[m]: doesn't really matter. if it's based on 7.1-rc1 it also has time because I won't start rebasing onto v7.1* before -rc6 (usually later)#asahi-dev · 2026-05-03 08:36
JamesCalligeros[m](non-copyrightable facts about the hardware/interfaces notwithstanding) #asahi-dev · 2026-05-03 08:35
JamesCalligeros[m]and if you do any binary re then youre not allowed to write nontrivial linux/m1n1 code unless we are sure you havent ripped off anything from apple, whether intentionally or otherwise#asahi-dev · 2026-05-03 08:34
JamesCalligeros[m]the separation is to contain the binary re so that if anyone doesnt want to see it, they dont have to miss out on anything else that could be important#asahi-dev · 2026-05-03 08:34
svenread it again then#asahi-dev · 2026-05-03 08:34
freetolgathe re policy?#asahi-dev · 2026-05-03 08:34
svenwhy not?