#asahi · 2026-02-06 21:18
svenhttps://github.com/torvalds/linux/blob/master/fs/fuse/cuse.c that thing#asahi · 2026-02-06 21:17
svendoesn't it have some additional features that allows creating a char device from userspace? i don't wanna write a kernel driver just to create a /dev/tty which can be used with picocom and all other tools we have#asahi · 2026-02-06 21:15
sveni think I have the protocol now, just need to write some tool that uses fuse or whatever to create a tty#asahi · 2026-02-06 21:14
svenhopefully this weekend ;)#asahi · 2026-02-06 16:08
sventhen comes m3 and then we'll tackle m4 eventually#asahi · 2026-02-06 16:08
svenit's literally that though plus that most of us are still focused on upstreaming and other features#asahi · 2026-02-06 09:52
sventhere’s no regression, you very likely just have the wrong device trees #asahi · 2026-02-05 21:22
svenfamous last words #asahi · 2026-02-05 21:22
sven*a#asahi · 2026-02-05 21:22
svenIf someone really wants to yak to shave: put the rp2040 directly onto the pcb and also adds usb hub with both the rp2040 and the DUT connected to it and a single c cable to the host only ;)#asahi · 2026-02-05 21:16
svenonly need to solder pin headers and a rp pico onto the boards they’ll send you #asahi · 2026-02-05 21:16
svenyes!#asahi · 2026-02-05 21:15
svenhuh? just send the files off to jlcpcb and let them deal with everything #asahi · 2026-02-05 21:14
svencan strongly recommend to just get https://git.kernel.org/pub/scm/linux/kernel/git/maz/cs-hw.git/#asahi · 2026-02-03 11:38
sventhere’s a few minor fixups like marking some domains as always-on and changing parents but most are just correct #asahi · 2026-02-03 11:38
svenvery reliable on m1-3, useless on m4#asahi · 2026-01-28 21:23
svenasahi-diagnose literally uses journalctl to export the kernel log for the previous and current boot#asahi · 2026-01-28 21:20
svenso then you wouldn't expect a oops or anything like that in the log, would you?#asahi · 2026-01-28 21:19
svenit's just possible that it didn't crash but froze instead #asahi · 2026-01-28 21:19
sventhen maybe take a closer look, it always includes the previous and current boot#asahi · 2026-01-28 21:13
svenenumerator: GitHub issue tracker e.g.#asahi · 2026-01-28 21:12
svenferdzo: what makes you think that?#asahi · 2026-01-28 21:11
svenasahi-diagnose output usually is a good start #asahi · 2026-01-27 11:57
svenusually doesn't matter, both @sven and sven results in a notification for me#asahi · 2026-01-24 21:01
svenalso the kernel package from some COPR rpm, nothing self-build. i've chased too many things that turned out to be not actual bugs to debug anything that i don't trust#asahi · 2026-01-24 20:58
svenno special stuff please, just the vanilla setup#asahi · 2026-01-24 20:58
svenno, please install fedora on the internal nvme#asahi · 2026-01-24 20:57
svenplease reproduce on Fedora and give me Asahi-diagnose output then #asahi · 2026-01-24 20:57
svenwhich distro is this? Fedora?#asahi · 2026-01-24 20:56
sventhat would disable usb support completely #asahi · 2026-01-24 20:56
svenuh, what?#asahi · 2026-01-16 18:24
svenlet me put it differently: if you had such a cable you would know because optical tbt cables start at >300eur or so#asahi · 2026-01-16 18:23
svenhow would that do tbt?#asahi · 2026-01-16 18:23
svenno, obviously not#asahi · 2026-01-14 09:51
svenmaybe explains why they are so cheap if they outsource that to their customers ;)#asahi · 2026-01-14 09:51
svenyeah, it seems like they have no real QA#asahi · 2026-01-14 09:50
svenit looked good even under a microscope but sometimes wouldnt work at all. flux and hot air fixed it then and it’s been reliable since then #asahi · 2026-01-14 09:49
svenI had to rework the usb connector on a CS that one of you two sent me as well to make it work reliably #asahi · 2026-01-13 20:25
svenat least until usb works#asahi · 2026-01-13 20:24
svenfor any early bringup really#asahi · 2026-01-13 20:24
svenyou also need hardware serial or a second mac for this fwiw#asahi · 2026-01-13 20:24
svenrun strings on the pmgr kext i guess#asahi · 2026-01-13 20:23
svenpitust: sure, but getting the other cores up is going to be _much_ easier than getting the mmio tracer working again#asahi · 2026-01-13 20:23
sveni'd probably run xnu with all PMGR debug flags enabled and see what it prints#asahi · 2026-01-13 20:23
svenonly the last part is related to SPTM#asahi · 2026-01-13 20:22
svenwe need to understand how to turn on power, then how to turn on the other cores and _then_ how to get the mmio tracer working again#asahi · 2026-01-13 20:22
svenno#asahi · 2026-01-13 20:21
svenso it's much too early to even try booting linux right now#asahi · 2026-01-13 20:21
svenand we also don't know how to turn on the other cores iirc#asahi · 2026-01-13 20:21
svenon m4 we don't even know how to turn on power for the various blocks