#asahi-dev · 2026-08-11 19:51
sveni'll get some sleep and stare at it again tomorrow#asahi-dev · 2026-08-11 19:51
sventhis must be the fabric firewall the slop generators always talk about /s#asahi-dev · 2026-08-11 19:47
chaos_princessno change#asahi-dev · 2026-08-11 19:34
svenchaos_princess: now this really shouldn't make a difference, but i'm kinda out of idea at this point: https://f.svpe.de/99fe30ff297fe3d18e7d347080279f27036bffe01b4f1bfc0500fc57bfa187f9_0002-HACK.patch#asahi-dev · 2026-08-11 19:23
chaos_princessno change#asahi-dev · 2026-08-11 19:16
svenshouldn't matter, but just with e2e#asahi-dev · 2026-08-11 19:16
chaos_princessshould i try this w/ or w/o e2e?#asahi-dev · 2026-08-11 19:15
svenchaos_princess: https://f.svpe.de/f1881bb46acce261fdbb1865e5d3093f5157cd31169501371ab71ce7f8bcd9c2_0001-HACK.patch#asahi-dev · 2026-08-11 19:07
svensame story every time :D#asahi-dev · 2026-08-11 19:07
svenlol, and ofc now i see a difference#asahi-dev · 2026-08-11 19:00
svenfor some reason it just triggers no interrupts for ring1#asahi-dev · 2026-08-11 18:57
svenchaos_princess: i dunno, this all looks very identical to me :-(#asahi-dev · 2026-08-11 18:34
sventhat actually looks like either a bug in the tipd driver or my tbt-switch thing because those modes don't make sense#asahi-dev · 2026-08-11 18:33
svenprobably won't make a difference, there's something wrong with the cable mode#asahi-dev · 2026-08-11 18:32
RafayI am not running the latest commit, let me fetch that.#asahi-dev · 2026-08-11 18:32
Rafay[ 193.553395] thunderbolt-apple-acio 501ac0000.nhi: TBT cable: cable_mode 0x0, device_mode 0x0, enter_vdo 0x3, orientation 2 -> cable info 0x603#asahi-dev · 2026-08-11 18:16
svenso 0x413 is correct and 0x603 wrong#asahi-dev · 2026-08-11 18:16
svenand macos writes [cpu2] [ACIOTracer@/arm-io/acio1] CTL TX WRITE route=0x0 port=0 space=SWITCH offset=0x36 len=1 seq=2 data=[0x00000413]#asahi-dev · 2026-08-11 18:16
sven[cpu4] [ACIOTracer@/arm-io/acio1] CTL TX WRITE route=0x0 port=0 space=SWITCH offset=0x36 len=1 seq=0 data=[0x00000603] <-- that's the VSE being written and after that nothing happens#asahi-dev · 2026-08-11 18:15
sventhe issue there is just that the apple cable VSE is wrong#asahi-dev · 2026-08-11 18:15
svenRafay: are you running the latest commits from the tbt-reset-wip branch? if yes, do you see a "TBT cable: cable_mode " entry in dmesg?#asahi-dev · 2026-08-11 18:04
sven*spot#asahi-dev · 2026-08-11 18:04
sventhere has to be some difference somewhere and the closer the trace are the easier it will be to stop#asahi-dev · 2026-08-11 18:04
svenyes#asahi-dev · 2026-08-11 18:03
chaos_princessshould i make one w/ e2e enabled?#asahi-dev · 2026-08-11 18:03
sventhat should still work ofc, but that makes the traces hard to compare#asahi-dev · 2026-08-11 18:02
sveni think it's still disabled#asahi-dev · 2026-08-11 18:02
sven[cpu1] [ACIOTracer@/arm-io/acio1] TX ring 1 flags = 0x80000000 (ENABLE=1, RAW=0, NO_SNOOP=0, E2E_FLOW_CONTROL=0, ISOCH_ENABLE=0)#asahi-dev · 2026-08-11 18:00
chaos_princessshould be disabled unless i did sth wrong#asahi-dev · 2026-08-11 18:00
svenchaos_princess: is that still with e2e enabled or disabled?#asahi-dev · 2026-08-11 17:55
sventhanks! let me look at chaos' traces first and then yours#asahi-dev · 2026-08-11 17:18
Rafaysven: here's your tracer on my machine if it helps. Linux: https://pastebin.com/V8kfgXPH Mac: https://pastebin.com/rGZNU87Y#asahi-dev · 2026-08-11 17:03
svenideally both on macos and on linux once#asahi-dev · 2026-08-11 17:01
svenack#asahi-dev · 2026-08-11 16:59
chaos_princessack, give me a bit#asahi-dev · 2026-08-11 16:57
svenchaos_princess: can you run https://f.svpe.de/f9237287e3af548f79fb0d8ceaa85c49b7f8f6ed27a36d07d15d6f86d1292b21_trace_acio.py.txt under the hv up until the point where you run ping and the packets then disappear?#asahi-dev · 2026-08-11 16:55
RafayOh, got it.#asahi-dev · 2026-08-11 16:54
svenso probably#asahi-dev · 2026-08-11 16:54
svenmy kernel branch does not contain much of the asahi-wip stuff#asahi-dev · 2026-08-11 16:53
Rafaysven: Is this lspcie output expected? I think some nodes are missing. https://pastebin.com/PeSMcLTQ#asahi-dev · 2026-08-11 16:32
yukadice roll#asahi-dev · 2026-08-11 16:21
chaos_princessyuka: is m4 smp reliable w/ that, or still a dice roll?#asahi-dev · 2026-08-11 16:07
svenack#asahi-dev · 2026-08-11 16:07
Rafayt8112-j473#asahi-dev · 2026-08-11 16:06
RafayYes.#asahi-dev · 2026-08-11 16:06
sventhat's base M2, right?#asahi-dev · 2026-08-11 16:06
svennice, thanks!#asahi-dev · 2026-08-11 16:06
Rafaysven: Here's the trace under linux. https://gist.github.com/rafayahmed317/d7d8768cec63178c0c15a8983af7d553 . Just to remind you, on my setup, thunderbolt_net doesn't create a device and xdomain doesn't initialize. You already have the mac trace: https://gist.github.com/rafayahmed317/e083de8e478830dc37969ca0b36750e7