#
✕ nick:sven
#asahi-dev · 2022-10-26 16:19
svenbut on macOS I can even reproduce it with non-random data
#asahi-dev · 2022-10-26 16:19
svenyeah, can’t seem to reproduce it with a Linux host either
#asahi-dev · 2022-10-26 15:02
svenor that
#asahi-dev · 2022-10-26 15:02
svenalso wouldn't surprise me if this is a bug in apple's usb stack
#asahi-dev · 2022-10-26 15:02
sveni used a macos host fwiw
#asahi-dev · 2022-10-26 15:02
svenfun
#asahi-dev · 2022-10-26 10:41
svenugh
#asahi-dev · 2022-10-26 10:17
svenhuh, seems to happen here as well
#asahi-dev · 2022-10-25 17:29
svencan’t reproduce it here though with t8103 even without the hv
#asahi-dev · 2022-10-25 17:28
svenit’s a bit annoying that tps chip is so fragile
#asahi-dev · 2022-10-25 10:59
svenyup
#asahi-dev · 2022-10-25 09:50
svenlooks like today is a good day for solving annoying problems very quickly :P
#asahi-dev · 2022-10-25 09:50
svennice! :D
#asahi-dev · 2022-10-25 09:37
svenokay, that was a surprisingly quick fix :)
#asahi-dev · 2022-10-25 09:37
sven[ 30.067385] usb 4-1: new SuperSpeed USB device number 2 using xhci-hcd
#asahi-dev · 2022-10-25 09:28
sveni hope I can somehow fix that. otherwise we'll have to do some very ugly hacks in the DT to make sure this all runs in the right order :/
#asahi-dev · 2022-10-25 09:28
svenand ofc atcphy breaks when I move the setup to the mux set instead of the phy set_mode callback *sigh*
#asahi-dev · 2022-10-24 17:03
svenbikesheds need as many people as possible anyway!
#asahi-dev · 2022-10-24 16:00
svenand if anyone on t8103 with a display connected to atcphy1 that will negotiate link rate 0xa actually wants to try it run this before booting the kernel https://f.svpe.de/d97f0ab6924026d2377b11bdd7a1cdf68383fb6b2965ba1d912680ed168dde47_atc_dpaltmode.py
#asahi-dev · 2022-10-24 15:59
svenlol
#asahi-dev · 2022-10-24 15:58
svenhttps://github.com/svenpeter42/linux/commit/6147785db5c5255c81de76cd10af4c4434bb4a97 if anyone wants a sneak peak for the altmode mess
#asahi-dev · 2022-10-24 15:12
svenhrm, this "channel depth" thing confuses me
#asahi-dev · 2022-10-24 15:07
sven(and mailbox ofc)
#asahi-dev · 2022-10-24 15:07
svenbut it seems to be great to break regular dcp code as well ;)
#asahi-dev · 2022-10-24 15:06
sventhe dpalt stuff needs a lot more work anyway
#asahi-dev · 2022-10-24 15:05
sventhat certainly would've made this entire AFK stuff less horrible...
#asahi-dev · 2022-10-24 15:04
svenit also breaks forever when it needs to switch the link rate but that's kinda expected since I fake all that atm
#asahi-dev · 2022-10-24 15:03
svencan't even reproduce it easily
#asahi-dev · 2022-10-24 15:03
svenok :D
#asahi-dev · 2022-10-24 14:58
svenyeah
#asahi-dev · 2022-10-24 14:55
svenuh oh.. (after switching mode a few times)
#asahi-dev · 2022-10-24 14:55
sven[ 174.032396] apple-dcp 271c00000.dcp: unexpected busy command channel
#asahi-dev · 2022-10-24 13:36
svenjannau: you can also drop all dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)) since 32bit is the default dma mask anyway
#asahi-dev · 2022-10-24 13:29
sven*issues
#asahi-dev · 2022-10-24 13:29
svenit's just silly how much issue we have with that should be a irq handler with a MMIO read in a loop + a function that does a single MMIO write
#asahi-dev · 2022-10-24 13:28
svenwe should really just drop that thing
#asahi-dev · 2022-10-24 12:46
svenwasn't complaining, just something I noticed while reading the code :)
#asahi-dev · 2022-10-24 12:46
sven:D
#asahi-dev · 2022-10-24 12:40
svenjannau: fwiw, parser.c leaks memory for about every parse_string call right now
#asahi-dev · 2022-10-24 11:58
svenjust doing the dptx stuff correctly and then everything works
#asahi-dev · 2022-10-24 11:58
svennope
#asahi-dev · 2022-10-24 11:57
svenpython3 proxyclient/tools/run_guest.py -m proxyclient/hv/trace_dcp.py -S payload.macho
#asahi-dev · 2022-10-24 11:57
svenpython3 proxyclient/experiments/atc_dpaltmode.py
#asahi-dev · 2022-10-24 11:57
svenyup
#asahi-dev · 2022-10-24 11:57
sventime to get rid of atcphy.py and move that into atcphy.c now :)
#asahi-dev · 2022-10-24 11:56
svenatcphy.py + linux dcp driver
#asahi-dev · 2022-10-24 11:56
svenhttps://f.svpe.de/2a0f24d3b7ede3dac32f71ec8bb2a4bcd22983aaa7307d1e27d1bd2c7dea586a_dcpext.png \o/
#asahi-dev · 2022-10-24 11:56
svenmarcan: yup, off-by-0x10 which just happened to work because of the 0x40 alignment so far
#asahi-dev · 2022-10-24 10:01
svenit doesn’t even wrap yet
#asahi-dev · 2022-10-24 09:24
svenhuh, so my code sets the write pointer from 0xc80 to 0xd40 but in the tracer it looks like DCP tries to read past that for some reason from 0xd40 to 0xd80 and sets the read pointer to 0xd80 then and continues to increment it by 0x10 until it eventually crashes

plus ›