#
✕ nick:marcan
#asahi · 2021-04-17 07:07
marcanhttps://github.com/pidcodes/pidcodes.github.com/pull/635
#asahi · 2021-04-17 07:07
marcanar: you joined late, I sent a PR for a PID request earlier in the stream
#asahi · 2021-04-15 04:57
marcancool
#asahi · 2021-04-14 22:28
marcansven: the PR is just the core, right? so I can merge it as is and do my own proxy handling on top
#asahi · 2021-04-14 22:26
marcantoday most likely, assuming no major blockers :)
#asahi · 2021-04-14 12:15
marcanbest part is it's still compatible with my ancient installed m1n1, so I still don't have to bother updating it :)
#asahi · 2021-04-14 12:14
marcansven: chainload.py updated to do in-place loads, no more running out of RAM ;)
#asahi · 2021-04-14 10:01
marcanthanks :)
#asahi · 2021-04-14 09:30
marcanin the meantime I'm going to shave a yak
#asahi · 2021-04-14 09:22
marcansven: go for it, I was going to ask about that next
#asahi · 2021-04-14 09:13
marcanI'll probably rip out the earlycon thing later when I do it in a more general way, but for now that works
#asahi · 2021-04-14 09:12
marcan(made the font 8bpp because RGBA is incredibly silly for a grayscale font, dos2unix OFL license, Makefile merge fix)
#asahi · 2021-04-14 09:12
marcansven: added a few fixes on the FB PR and merged, thanks!
#asahi · 2021-04-13 22:15
marcansven: I already tried to use this to reserve ranges for linux once and realized it was reserving all of RAM :)
#asahi · 2021-04-13 17:18
marcanIIRC that thing contains all of RAM or something like that too
#asahi · 2021-04-13 17:10
marcanthe problem is I don't know how to enumerate *only* those
#asahi · 2021-04-13 17:09
marcanyes, because only stuff like the SEP carveout is actually blocked
#asahi · 2021-04-13 17:04
marcanyeah
#asahi · 2021-04-13 16:57
marcansince the ends are probably not guaranteed to be aligned to a larger block
#asahi · 2021-04-13 16:57
marcanbut it means we need the pagetables to go down to page level, probably
#asahi · 2021-04-13 16:57
marcanso we should do the same
#asahi · 2021-04-13 16:57
marcanthat's what we use to pass to linux
#asahi · 2021-04-13 16:57
marcannot in the ADT
#asahi · 2021-04-13 16:57
marcansven: the usable RAM range is in bootargs
#asahi · 2021-04-12 14:09
marcanand AIUI there are ARM cores that unconditionally guarantee TSO and under those the API could just always return that
#asahi · 2021-04-12 14:09
marcanif ARM isn't working on anything along those lines, well, they should be :)
#asahi · 2021-04-12 14:08
marcanI mean, is there significant reason not to expose it? it would seem like a fairly minimal patch, and decent x86 emulation is a significant benefit
#asahi · 2021-04-12 14:04
marcansomething like PR_SET_MEMORY_MODEL
#asahi · 2021-04-12 14:04
marcanmaz: wouldn't it just be a prctl? that's where all the other related crap is
#asahi · 2021-04-12 13:48
marcan(maz: thoughts?)
#asahi · 2021-04-12 13:48
marcansame with AMX
#asahi · 2021-04-12 13:48
marcanAF/PF and the SSE stuff is ickier; those rely on Apple-proprietary registers which again would become part of architectural state, and I'm not sure how much hate we would get for trying to introduce those into process context
#asahi · 2021-04-12 13:47
marcanTSO should be easy; that just requires flipping a bit in ACTLR_EL1 (which does mean we need to save it on context switches or whatever, but that shouldn't be a big deal since it *is* at least a standard register)
#asahi · 2021-04-12 13:46
marcanI do wonder how many of the M1 Rosetta-isms we can expose on Linux sanely
#asahi · 2021-04-12 13:45
marcanqemu does work
#asahi · 2021-04-12 13:45
marcansorear: no system stuff AIUI, only userspace
#asahi · 2021-04-12 12:44
marcanwould be nice to start building up a high-level schema to do this properly. not for actual conversion, but to get nice pretty-printed versions, and for diffing / dealing with newer devices
#asahi · 2021-04-12 12:37
marcanthis: https://github.com/Shizmob/img4/blob/master/dt.py
#asahi · 2021-04-12 12:36
marcanI've been using shiz's dt.py, which spits out this format: https://mrcn.st/p/t1926QWh
#asahi · 2021-04-12 12:29
marcanproperty values are just bytes, and the endinanness is different, so you can't know what needs to be flipped
#asahi · 2021-04-12 12:28
marcantmlind: adt to dts translation is not possible in the general sense, because there is not enough information in the format
#asahi · 2021-04-12 09:55
marcanyeah, I've seen that on some dts IIRC
#asahi · 2021-04-12 09:54
marcanshould be an easy patch though
#asahi · 2021-04-12 09:54
marcanI guess it still needs to be handled in the driver though, right?
#asahi · 2021-04-12 09:53
marcando we have a standard binding for passing macs, or is that a device-specific thing still?
#asahi · 2021-04-12 08:18
marcanI think the driver needs a few patches, but that's about it
#asahi · 2021-04-11 19:06
marcan:)
#asahi · 2021-04-11 19:06
marcanwhich I complained about
#asahi · 2021-04-11 19:06
marcanthat's what sven did in the patch above
#asahi · 2021-04-11 18:56
marcankeep in mind there's a power manager core too, I don't know how much of this is real registers and how much is somehow backed by that thing

plus ›