#
✕ nick:sven
#asahi-dev · 2022-10-24 08:24
svenguess i'll rebase and send another version later today
#asahi-dev · 2022-10-24 08:24
svenhuh, bluetooth just disappeared from patchwork *sigh*
#asahi-dev · 2022-10-24 08:18
svenyeah, same for bluetooth. who knows if there's someone crazy enough to connect that pcie card to big endian machine with a pcie bus
#asahi-dev · 2022-10-24 08:12
svenmarcan: I kinda like the __leXX types because that way it's obvious when some buffer is sent across to some other processor vs. when it's just some internal data structure
#asahi-dev · 2022-10-24 08:08
svenso the previous problem was that I messed up the afk dp->ap call handling. looks like now I also broke the ringbuffer handling itself :D
#asahi-dev · 2022-10-24 08:08
sven[ 2.980643] apple-dcp 271c00000.dcp: RTKit: Message (id=1): ringbuffer abort at iop_ringbuffer.h:856 <-- uh oh
#asahi-dev · 2022-10-23 14:14
svenbut i already don’t see all the dptx calls that I expect in Linux
#asahi-dev · 2022-10-23 14:14
svendcpext.py is kinda broken now :D
#asahi-dev · 2022-10-23 14:06
svenlooks like it gets color modes though. guess something on the dptx side is broken then
#asahi-dev · 2022-10-23 14:06
sven 00000a40 02 00 00 09 49 44 02 00 40 00 00 04 48 00 00 00 |....ID..@...H...|
#asahi-dev · 2022-10-23 14:06
sven 00000a30 65 01 00 00 40 00 00 04 37 b5 03 00 00 00 00 00 |e...@...7.......|
#asahi-dev · 2022-10-23 14:06
sven 00000a20 08 00 00 02 0d 00 00 01 05 00 00 09 53 63 6f 72 |............Scor|
#asahi-dev · 2022-10-23 14:06
sven 00000a10 0a 00 00 09 43 6f 6c 6f 72 4d 6f 64 65 73 02 00 |....ColorModes..|
#asahi-dev · 2022-10-23 14:05
svener.. wait.. that's not actually the mode ids
#asahi-dev · 2022-10-23 14:05
svenokay, macos uses Message: A412 (IOMobileFramebufferAP::set_digital_out_mode(unsigned int, unsigned int)): (in 0x8, out 0x4)
#asahi-dev · 2022-10-23 14:01
svenuhhhh.... now DCP crashes when running macos
#asahi-dev · 2022-10-23 13:58
svenset_digital_out_mode(arg0=82, arg1=38) <-- that's what happened when I used this capture card with dcpext.py
#asahi-dev · 2022-10-23 13:58
svenhm... I guess something with the DPTX code is broken
#asahi-dev · 2022-10-23 13:55
svenhuh
#asahi-dev · 2022-10-23 13:54
sven(also, lol, I just realized I reimplemented dcp's parse.c in much dumber because I didn't realize it's the same format)
#asahi-dev · 2022-10-23 13:52
svenah
#asahi-dev · 2022-10-23 13:51
svenright, but how can I get that from a trace?
#asahi-dev · 2022-10-23 13:51
svenhttps://f.svpe.de/94d44bd9344a281e7c5cd73c726a292f590999d85f854473c74db961ff73dc70_dcp.log.txt
#asahi-dev · 2022-10-23 13:49
svenwhere's dcpav_prop?
#asahi-dev · 2022-10-23 13:47
svenurgh..
#asahi-dev · 2022-10-23 13:47
svenpython: Relative module names not supported
#asahi-dev · 2022-10-23 13:45
svenok
#asahi-dev · 2022-10-23 13:44
svenwhat's the "extra tracer"?
#asahi-dev · 2022-10-23 13:43
svenhrm, let me see if I can get the tracer to work
#asahi-dev · 2022-10-23 13:42
svenhttps://f.svpe.de/8b427ff54cc0acf56d3d1a93f70eebc1ae1c807c045e7463685d63e0f0a16ded_modes.txt
#asahi-dev · 2022-10-23 13:39
sventrying to dump the raw dcp->chunks.data, dcp->chunks.length now
#asahi-dev · 2022-10-23 13:35
svenin parse_color_modes
#asahi-dev · 2022-10-23 13:35
svenhuh, looks like dcp_parse_foreach_in_array(handle, outer_it) never does a single iteration
#asahi-dev · 2022-10-23 13:19
svenlet me also add tracepoints to the color modes
#asahi-dev · 2022-10-23 13:19
svenlooks like they all have best_color_mode = -1
#asahi-dev · 2022-10-23 13:19
sveniomfb_parse_mode_success: id: 70d, best_color_mode: -1, virtual: 0, score: 119
#asahi-dev · 2022-10-23 13:05
svennot sure yet, went for lunch and now i'm gonna add more tracepoints to see what's going on
#asahi-dev · 2022-10-23 12:36
svenprobably because the rx one doesn’t have those additional context switches in there
#asahi-dev · 2022-10-23 12:36
svenfor DCP syslog the rx path seems to be much faster than the tx one which results in that overflow
#asahi-dev · 2022-10-23 12:32
svenive hacked the define to 200 for now to debug DCP
#asahi-dev · 2022-10-23 12:32
svenyeah, it should be simple enough
#asahi-dev · 2022-10-23 12:30
svenand apparently the three(?) context switches in the tx path are enough to overflow that queue
#asahi-dev · 2022-10-23 12:30
svenyup, global define
#asahi-dev · 2022-10-23 12:09
svenhuh, looks like parse_mode fails for all of them
#asahi-dev · 2022-10-23 12:02
svenhrm, when tracing I can see the dcp->ap call that sets all modes but for some reason they aren't propagated to drm correctly I think
#asahi-dev · 2022-10-23 11:22
svenmarcan: ^-- so I can reliably overflow the mailbox queue with just lots of dcp debug spam. kinda surprised lina hasn't run into this with the gpu yet
#asahi-dev · 2022-10-23 11:21
sven[ 2.610175] apple-mailbox 271c08000.mbox: Try increasing MBOX_TX_QUEUE_LEN *sigh*
#asahi-dev · 2022-10-23 11:18
sven(or actually dart->fake_ttbr[sid])
#asahi-dev · 2022-10-23 11:18
svenor maybe even stuff it into something in struct apple_dart. dart->fake_ttbr = domain; :-P
#asahi-dev · 2022-10-23 11:17
svennot great.. but oh well

plus ›