#asahi · 2021-06-04 23:09
svenso.. uh.. i think i maybe know how to fix it without resetting dwc3#asahi · 2021-06-04 23:09
svenhuh...#asahi · 2021-06-04 23:08
svenwhen i disconnect the cable it goes from 0x1c1000f to 0x1d1000f and i don't see that in the mmiotrace#asahi · 2021-06-04 23:08
svenbut something strange happens in usb_regs.atc + 0x08#asahi · 2021-06-04 23:08
svenno :D#asahi · 2021-06-04 22:55
svenawesome, thanks!#asahi · 2021-06-04 22:44
sveni imagine that might bring the usb pd chip into a weird state#asahi · 2021-06-04 22:39
svendid you re-enable the hpm nodes as well?#asahi · 2021-06-04 22:35
svenhm, quite a few atc-phy pokes in there#asahi · 2021-06-04 22:29
svenbloom: just ask if you need more stupid ideas! :P#asahi · 2021-06-04 22:29
svenjannau: thanks! looks great already. can you maybe also plug the device in once more at the end? (i.e. to have a connect -> disconnect -> connect cycle)#asahi · 2021-06-04 22:24
svenoh, i guess you could just disable the debug serial when launching the hypervisor and then that should definitely work#asahi · 2021-06-04 22:21
svenif yes echo "hi" > dev/cu.debug-console might just work#asahi · 2021-06-04 22:21
svencan you still write to /dev/cu.debug-console when kprintf is enabled?#asahi · 2021-06-04 22:15
svencan you also trace irqs? or is that completely left to xnu right now?#asahi · 2021-06-04 22:14
svenmaybe during plugging/unplugging some simple device#asahi · 2021-06-04 22:14
svenusb-drd1 and atc-phy1#asahi · 2021-06-04 22:08
sven:)#asahi · 2021-06-04 22:08
svenhey, stand in line please! i asked for usb replugging first!#asahi · 2021-06-04 21:20
svenfast enough to replug a usb device? ;)#asahi · 2021-06-04 20:52
svenprobably the first four are just the "normal" interrupts but the last one looks supicious#asahi · 2021-06-04 20:50
sven(well, that or some magic bit that needs to be poked somewhere in the phy regs)#asahi · 2021-06-04 20:49
svenwould be lovely if you could figure out that usb replug issue. i suspect it's either the i2c chip or possibly one of the other interrupts in the drd node#asahi · 2021-06-04 20:49
sven:D#asahi · 2021-06-04 19:53
svennot sure if i used the right kernel version though#asahi · 2021-06-04 19:52
svenfirst time i tried i hit some weird pmgr panic#asahi · 2021-06-04 19:52
svenyeah, i just need to make it work#asahi · 2021-06-04 19:39
svensure#asahi · 2021-06-04 19:36
sveni'd rather just figure out how to do it properly#asahi · 2021-06-04 19:34
sven*not#asahi · 2021-06-04 19:34
svenhm, ok. so it's probably no the usb pd chip then#asahi · 2021-06-04 19:26
svenwell, unplug and then reattach#asahi · 2021-06-04 19:25
svenbtw. did anyone check if you can unplug devices on the usb-c ports when using the corellium kernel?#asahi · 2021-06-04 19:24
svenyes, still not something i'd comfortably enable in production#asahi · 2021-06-04 19:23
svenessentially any device attached to thunderbolt would have access to all memory#asahi · 2021-06-04 19:22
svenj_ey: just google thunderbolt dma attack or something like that#asahi · 2021-06-04 19:22
sven<sven> it should actually been *without :D <-- scratch that. i'm just confused.#asahi · 2021-06-04 19:20
sven*this way#asahi · 2021-06-04 19:20
sveneither way, we can be confident now that the dart bindings work with way and will continue to work if we can teach the kernel to handle 16k iommu pages with 4k cpu pages#asahi · 2021-06-04 19:19
svenit should actually been *without :D#asahi · 2021-06-04 19:19
sven:D#asahi · 2021-06-04 19:19
svenyeah, thunderbolt with bypass mode is a very bad idea#asahi · 2021-06-04 19:17
svenfeels much less like a hack now#asahi · 2021-06-04 19:16
svenalright, looks i can extract the bypass configuration from dma-ranges and i can also use def_domain_type to enforce bypass mode from within the DART driver if booting a 4k pagesize kernel :)#asahi · 2021-06-03 16:32
svenmakes sense#asahi · 2021-06-03 16:26
svenhrm. so essentially the pcie hw provides functionality that's similar to what's described in iommu-map, doesn't it?#asahi · 2021-06-03 16:21
svenprobably needs more magic pokes before it's available#asahi · 2021-06-03 16:21
sveni wanted to look at it but the MMIO region only returned 0xffff.. for all reads#asahi · 2021-06-03 16:20
sveni think i saw in the ADT that it has at least more sids#asahi · 2021-06-03 16:15
svenor is the pcie node for thunderbolt completely different anyway?