01:06 <kfischer> marcan: Alright, I’ve made some progress here. With the dart/dev branch and a dumb USB-C -> USB-A dongle, USB host works fine. Then with a slight modification to the device tree, I can also enable usb device mode and load the ethernet gadget in such a way that it shows up on the host as a USB device. No luck convincing macOS to actually load an ethernet driver for it though

01:13 <brentr123[m]> Sorry unrelated but happy 4th of July to all my American friends!!

02:28 <amw> kfischer: Are you going to try Linux? I believe the vuart under HV while remaining hangs on long lines so beware

02:35 <ae86> What even is linux

04:01 <marcan> amw: he's running it natiev

04:01 <marcan> *native

04:18 <amw> I'm slowly understanding the HV, samsung-uart driver. I can add debug to the Linux, m1n1 so I will slowly get to the bottom of the stuck vuart

04:23 <ae86> Why did apple implement iokit into firmware?

06:22 <amw> marcan: Tracing vuart() shows it happily writing in FIFO mode and then the writes stop but nothing has reported TxFull, but if I hack the s5l_serial_drv_data to fifosize = 160 (versus 16) it doesn't hang/block...

10:28 <sven> kfischer: fwiw, i also had trouble to get the serial gadget driver to load on macos. it didn't want to attach it to the kext. worked fine on linux though and i didn't care enough to figure out why

10:28 <sven> and nice to know that the role switching also works :-)

10:39 <amw> sven: Using the dev/dart kernel haven't been able to get any USB detected / working with linux.py or under HV run_guest.py with a disabled dwc3-0

10:40 <amw> I used a Type-C=>Type-A adapter and also a Type-C => 4 port Type A Hub

10:41 <sven> hm... strange

10:41 <sven> anything in dmesg?

10:42 <amw> Nothing at all - so I'm probably screwed up something :-(

10:42 <amw> *I've

10:42 <sven> does it show the dwc3 controllers at least?

10:43 <sven> if not, make sure both dwc3 and xhci is enabled in your configuration

10:43 <amw> No. - See the few messages I get here: https://paste.debian.net/1203360/

10:44 <sven> yeah, they might just be disabled in the kernel config

10:44 <amw> Hmm - what CONFIG's are those?

10:45 <sven> don't know by heart, but you can just search for dwc3 in make menuconfig

10:45 <sven> (i.e. press / and then enter dwc3)

10:46 <eta> (you can search?!?!)

10:46 <sven> you'll need dwc3, xhci and dart

10:46 <amw> CONFIG_USB_DWC3_HOST DWC3_HAS_XHCI CONFIG_USB_XHCI_HCD

10:46 <sven> sounds about right

10:46 <j_ey> eta: yes you can search!

10:46 <eta> this discovery is almost as great as when I figured out there's a key you can press to find module deps

10:46 <sven> :D

10:47 <amw> Hmm - non of them set .... :-D

10:47 <eta> instead of just getting the stupid "this has to be built as a module because its deps are" warning when you press Y

10:47 <sven> amw: oh, and you'll also need dart ofc

10:47 <j_ey> eta: in the search menu, you will see numbers on the left, you can press them to jump to the results

10:47 <eta> wild

11:05 <amw> Ugh it's going to take me a while to find all those options in menuconfig - tomorrow!

11:07 <sven> oh, it's simple

11:07 <sven> just press /

11:07 <sven> then you should see a search window

11:07 <sven> just type DWC3 in there

11:27 <amw> sven: Thanks - I've just been threshing around in the config => / search is *really* helpful - once I figured out how to navigate it :-)

11:27 <sven> :)

11:39 <amw> sven: YES! - I have plugged in a USB keyboard and happily typing in there.... - Thanks you SO MUCH !!!!

11:43 <sven> nice :-)

12:19 <amw> I added a photo to the Wiki - https://github.com/AsahiLinux/docs/wiki/SW:Linux#linux-usb-keyboard - thanks again!

12:26 <jn> heh, it's the good old dell keyboard with the built-in smartcard reader

12:26 <jn> i have one too

12:36 <amw> This will sound dumb - but I *never* realised it had a smart card reader till you mentioned it! I just thought it was an old keyboard ....

12:43 <JTL> amw: funny

12:58 <alyssa> sven: any idea why my kernel is failing to boot?

12:59 <alyssa> when I try it just hangs on.. the apple logo

12:59 <sven> uh

12:59 <sven> how did you run that kernel?

13:02 <alyssa> linux.py, and also tried concatenating with m1n1 and hypervisoring it

13:03 <sven> do you have a uart cable?

13:04 <alyssa> No

13:04 <alyssa> was hoping vuart would be enough if it came to that

13:05 <sven> M1N1DEVICE=/dev/tty.usbmodemP_03 python3 proxyclient/tools/linux.py Image.gz t8103-j274.dtb initramfs.cpio.gz

13:05 <sven> so that's what i use with a uart cable

13:05 <sven> make sure to disable the first dwc3 controller if you want to use vuart

13:05 <alyssa> Er ok

13:05 <alyssa> what initramfs are people using?

13:06 <alyssa> I don't usually use one, then again -ENONVME ..

13:06 <sven> some dumb busybox thing

13:06 <sven> let me see if i can just upload it

13:07 <sven> https://f.svpe.de/1f56ababce50be6e58463aef3cd8148db611e06143410e6e3aa41b5c08957639_initramfs.cpio.gz

13:08 <sven> and yeah.. nvme.. first dart, then usb, the pcie and then nvme ;)

13:08 <alyssa> sven: a binary only copy of Busybox? I thought better of you

13:08 <sven> i blame marcan! he gave that one to me

13:08 <alyssa> Please mail a DVD containing the corresponding source code to my address, ..

13:08 <sven> right, let me delete that file again :-P

13:09 <alyssa> 😋

13:29 <alyssa> sven: Maybe there's something broken with the kernel I built

13:29 <sven> do you get any output when you run it under the hv?

13:31 <alyssa> d'oh

13:32 <alyssa> nope still nothing

13:32 <alyssa> but that should be, er, closer

13:33 <sven> just to be sure: there are two ttys once the hv is running

13:33 <sven> the second one is the vuart one

16:44 <Ariadne> if people need help with busybox feel free to ping me

18:38 <marcan> amw: vuart is just not implemented, like, at all

18:38 <marcan> it's not a bug, it's just missing everything but trivial writes

18:38 <marcan> we need to implement tx complete IRQs and rx IRQs at the very least for this to work

18:39 <marcan> and rx itself of course

18:39 <marcan> linux will write one FIFO worth of characters, then wait for an IRQ

18:39 <marcan> no IRQ, no worky

18:41 <marcan> also I'm pretty sure that busybox is from a debian package or something :p

18:41 <marcan> definitely didn't compile it myself at least

18:57 <alyssa> marcan: Please send the entirety of the Debian source code as a stack of DVDs in exchange for a nominal fee I will send to you by check.

18:57 <alyssa> 😋

19:01 <marcan> need to check if japan post is *finally* shipping to the US

19:05 <alyssa> OOf