#
✕ nick:sven
#asahi-dev · 2022-09-03 11:18
svenbypass offset iirc, doesn’t work on all darts though
#asahi-dev · 2022-09-02 20:06
svendunno about the firmware itself, but the IPC protocol is pretty insane
#asahi-dev · 2022-09-02 19:09
svenhates DCP
#asahi-dev · 2022-09-02 18:32
svenIt selected dpin0, that confirms those are for tunneling :)
#asahi-dev · 2022-09-02 18:28
svenPretty sure DCP does some magic though. I’ve copied macOS MMIO pokes completely and the dptx clock didn’t come up
#asahi-dev · 2022-09-02 18:27
svenseriously… who came up with this crap
#asahi-dev · 2022-09-02 17:12
svenmarcan: oh dear... i'm looking into this dp alt mode dcp thing and I only realized just how cursed the entire DCP protocol is now
#asahi-dev · 2022-08-31 21:33
svenoh.. wait.. that's a notificatio from DCP which gets ACKed, isn't it?
#asahi-dev · 2022-08-31 20:48
sveni've tried to match that up with your dcp.dpport.port0 messages but didn't manage to find the corresponding request
#asahi-dev · 2022-08-31 20:48
svenmarcan: i'm still a bit confused by how DCP works, how can I e.g. parse a message like this (that happens just before the display crossbar is initialized): https://f.svpe.de/7529b17d1cad6cc4c9a946c3d258a93329a34d6e83476f6cb79396177fd854ff_dpmsg.txt
#asahi-dev · 2022-08-31 20:16
svenother than there's only bluetooth missing for me (for which I should sent a v2) and atcphy/usb3/thunderbolt but I don't understand that part well enough yet :D
#asahi-dev · 2022-08-31 20:15
svenyeah, the t6k dart was a bit demotivating because there was no reply at all even after two pokes
#asahi-dev · 2022-08-31 14:46
svenhrm, or maybe that's something.
#asahi-dev · 2022-08-31 14:37
sventhere's a function that takes a DCPAVIPC::Message message in the bt
#asahi-dev · 2022-08-31 14:37
svenlooks like the crossbar configuration is triggered from DCPEXT afaict
#asahi-dev · 2022-08-31 10:12
svenany idea if there's anything inside DCP related to "enable DPTX clock"? :D
#asahi-dev · 2022-08-31 10:12
svenatcphy regs look almost the same with macos init vs. my init
#asahi-dev · 2022-08-31 10:11
svenmarcan: ^-- so that's why it probably doesn't work, the DPTX clock isn't working for some reason
#asahi-dev · 2022-08-31 10:11
svenFIFO_WR_DPTX_CLK_EN_STAT = 0x1fc (DISPEXT_0_0=0, DISPEXT_0_1=0, ...)
#asahi-dev · 2022-08-31 07:02
svengoes back to real $work
#asahi-dev · 2022-08-31 07:02
svenlet's hope marcan finds a workaround for that during his wifi investigation :D
#asahi-dev · 2022-08-31 07:02
sven:(
#asahi-dev · 2022-08-31 06:47
svenand I bet there's some magic BT command I have to send after resume
#asahi-dev · 2022-08-31 06:46
svenI just noticed there's something wrong with bt resume. the devices comes back up but e.g. my mouse doesn't reconnect for a while
#asahi-dev · 2022-08-31 06:46
svenmarcan: btw. does the hv work when you suspend macOs for you? on t8103 it always reboots when I try that
#asahi-dev · 2022-08-31 05:46
svenwe’ll need hot plug for thunderbolt anyway. and I have some hack somewhere to make it work
#asahi-dev · 2022-08-31 05:34
svenit’s just WiFi and pcie
#asahi-dev · 2022-08-31 05:33
sventpw_rules: nvme suspend/resume has been working for a long time
#asahi-dev · 2022-08-30 18:44
svenmarcan: the debug stuff is nice, still stuck at "waiting for HPD" though. I'll have to look again and see if I missed any dpphy/atcphy init
#asahi-dev · 2022-08-30 17:37
svenohhh…. that’s even better!
#asahi-dev · 2022-08-30 17:34
svenjust adopt my policy that ever question about $x progress delays $x by a month :>
#asahi-dev · 2022-08-30 17:33
svengoodnight! I’m gonna give it another try later
#asahi-dev · 2022-08-30 17:16
svensomething is missing to establish the connection
#asahi-dev · 2022-08-30 17:16
svenI’m reasonably sure I mess up the atcphy/dpphy stuff somewhere fwiw
#asahi-dev · 2022-08-30 16:31
svenwouldn’t surprise me :D
#asahi-dev · 2022-08-30 14:21
sven:D
#asahi-dev · 2022-08-30 13:11
svenlooks like i'm still missing some initialization stuff for the dpphy fwiw
#asahi-dev · 2022-08-30 13:08
sveninput from ATC's point of view
#asahi-dev · 2022-08-30 13:08
svenprobably display port input 0/1 for the usb4 tunelling mess
#asahi-dev · 2022-08-30 13:05
svenand/or this breaks on macos only
#asahi-dev · 2022-08-30 13:05
svenNameError: name 'DEFAULT_ARCH' is not defined <-- I think you forgot to push some changes
#asahi-dev · 2022-08-30 12:05
svendoesn't seem to contain the xbar
#asahi-dev · 2022-08-30 12:02
sventhe xnu function which sets up the crossbar is called connect and/or bringConnectionUp iirc
#asahi-dev · 2022-08-30 12:02
svenmaybe it'll just try to write to the same crossbar regs
#asahi-dev · 2022-08-30 12:02
svenwhich MMIO can dcp access?
#asahi-dev · 2022-08-30 12:01
svenhah :D
#asahi-dev · 2022-08-30 11:55
sventhere's e.g. some command followed by [cpu5] [DCPTracer@/arm-io/dcpext] [syslog] * [AppleDCPDPTX.cpp:361][AFK]powering nub 0x8553a8
#asahi-dev · 2022-08-30 11:55
svenhas managed to successfully ignore anything DCP so far
#asahi-dev · 2022-08-30 11:55
svenbut I don't quite understand what's going on yet
#asahi-dev · 2022-08-30 11:55
svensounds good, I do see a few commands at the beginning I think.

plus ›