#
✕ nick:sven
#asahi · 2021-07-25 18:10
sveni'd actually be surprised if any filesystem is enabled :D
#asahi · 2021-07-25 18:10
sveni started with a pretty minimal .config
#asahi · 2021-07-25 18:09
sveni'm also not sure i have any filesystems enabled in my config fwiw
#asahi · 2021-07-25 18:08
svenoh, they also only work once. so better plug in the usb device before booting the kernel
#asahi · 2021-07-25 18:07
svenunless you want also mess around with u-boot, then you can also make the usb a ports work
#asahi · 2021-07-25 18:07
svenyes
#asahi · 2021-07-25 18:05
svennice :)
#asahi · 2021-07-25 17:28
svenalyssa: https://f.svpe.de/99f754dcb35025b32f5fdfd093133c30ef8ad3e54a3046fc1c919e664668c59a_.config
#asahi · 2021-07-25 17:27
svenhm. let me see if i can upload my .config
#asahi · 2021-07-25 17:26
svenyou did enable ARCH_APPLE, didn't you?
#asahi · 2021-07-25 17:24
sven:)
#asahi · 2021-07-25 17:24
svenit'll die, but at least you should see something on the screen and/or your serial console
#asahi · 2021-07-25 17:23
svenyou could also try to boot the kernel without that initrd at first
#asahi · 2021-07-25 17:22
sveni think someone in here did that (amw maybe?)
#asahi · 2021-07-25 17:22
svenit should
#asahi · 2021-07-25 17:22
svenyou could just use a usb stick for the rootfs with that kernel :P
#asahi · 2021-07-25 17:21
svenlast time i tried to give it to you you wanted me to send you the busybox source code via snailmail :-P
#asahi · 2021-07-25 17:20
svenoh, you can use whatever or nothing at all i think
#asahi · 2021-07-25 16:50
svennot very important because we can just turn them all on from m1n1, but uh.. i think we should have them at some point
#asahi · 2021-07-25 16:50
sventhere's also the clock gates left to do
#asahi · 2021-07-25 16:10
svenif it was used in multiple places we could probably implement our own dma_map_ops or something for it, but meh.
#asahi · 2021-07-25 16:10
svenit's also only used inside the ANS afaict
#asahi · 2021-07-25 16:10
svenit's not really an iommu anyway
#asahi · 2021-07-25 16:10
sveni don't think so ;)
#asahi · 2021-07-25 16:08
svenbut i haven't read much of that code yet
#asahi · 2021-07-25 16:08
svenno, i think platform_bus_type is just the "virtual" top-level bus for platform devices
#asahi · 2021-07-25 16:07
svenyes
#asahi · 2021-07-25 16:05
svenbut i guess you could just call it with dev->bus for the ans device and it should do the right thing
#asahi · 2021-07-25 16:05
svenhrm, though i guess i'd have to look up how to make that work. usually bus_set_iommu is called with platform_bus_type or pci_bus_type
#asahi · 2021-07-25 16:04
svenah, true
#asahi · 2021-07-25 16:03
svenhttps://github.com/torvalds/linux/blob/9f7b640f001f9781e0803fb60e7b3e7f2f1a1757/drivers/iommu/iommu.c#L1853 bus_set_iommu
#asahi · 2021-07-25 16:02
sveni guess there was no need for more than one type so far?
#asahi · 2021-07-25 16:02
svenyeah, one type(!) of iommu
#asahi · 2021-07-25 16:01
sventhe second one might be fixable, the first one would be painful to fix
#asahi · 2021-07-25 16:01
sventwo reasons: there's only one iommu per bus and there's no domain type for "passthrough only"
#asahi · 2021-07-25 16:01
svennope
#asahi · 2021-07-25 16:00
sveneither way, we can just have a apple,mbox = <&ans_mbox 0x20>, <&ans_mbox 0x21>; or something like that in the nvme node. that should bring up ANS then. or maybe have it send a ping message to the asc and wait for the reply
#asahi · 2021-07-25 15:59
svenah, that's where it's hidden
#asahi · 2021-07-25 15:57
svenit's not in the preboot volume
#asahi · 2021-07-25 15:56
svenwhere?
#asahi · 2021-07-25 15:55
svenand i don't think we have the ANS firmware
#asahi · 2021-07-25 15:55
sveni haven't seen any traffic on them when running the hv
#asahi · 2021-07-25 15:55
svennope
#asahi · 2021-07-25 15:54
sven0x20 and 0x21 i think
#asahi · 2021-07-25 15:54
svenbut i haven't seen them being used
#asahi · 2021-07-25 15:54
sventechnically yes
#asahi · 2021-07-25 15:54
svenbut yeah, maybe just a special compatible for the mailbox driver that whitelists the shared memory in SART
#asahi · 2021-07-25 15:53
svenso the actual nvme device doesn't need SART actually
#asahi · 2021-07-25 15:53
svenit's litterally 16x2 registers. first one is the address and second one is the lengths (+some flags)
#asahi · 2021-07-25 15:52
svenSART is just a dumb DMA address filter and i haven't thought about how to implement that

plus ›