#asahi-dev · 2026-06-08 19:37
svenare you sure...#asahi-dev · 2026-06-08 19:35
svenignore the dts files in u-boot's tree if they are still there#asahi-dev · 2026-06-08 19:35
svenit's the dtb passed by m1n1#asahi-dev · 2026-06-08 19:35
svenyeah#asahi-dev · 2026-06-08 19:35
svenhuh?#asahi-dev · 2026-06-08 19:16
svenwell, get it working once :D#asahi-dev · 2026-06-08 19:16
svenyou should be able to get it working with a snps,dwc3 or however that compatible was called and no connection to any phy or tps chip#asahi-dev · 2026-06-08 18:41
svenluckily not a big deal though, we'll eventually be able to virtualize sptm on m2 and that should be similar enough anyway#asahi-dev · 2026-06-08 18:41
svenkinda expected that to happen sooner or later#asahi-dev · 2026-06-08 09:37
svenmaybe move this to -off topic #asahi-dev · 2026-06-08 09:34
svenin real life security through obscurity buys you time especially when combined with real security #asahi-dev · 2026-06-08 09:30
svenit’s a very black box to us #asahi-dev · 2026-06-08 09:30
sveneven the sep crashes are encrypted with a key we don’t have #asahi-dev · 2026-06-08 09:25
svenand there’s ~nothing known about its peripherals #asahi-dev · 2026-06-08 09:25
svensep also has its keys hidden in hardware engines#asahi-dev · 2026-06-08 09:11
sven*sigh*#asahi-dev · 2026-06-08 09:11
svenI’d just file a radar for tso and see what happens #asahi-dev · 2026-06-08 09:09
sventhere are arm machine with tso by default #asahi-dev · 2026-06-08 09:08
sventso disabled might be an accident #asahi-dev · 2026-06-08 08:55
sveneverything that’s not architectural arm stuff#asahi-dev · 2026-06-08 08:53
svenI really don’t buy the “some people tried to use amx or whatever and that pissed arm off” story #asahi-dev · 2026-06-08 08:53
sven+do#asahi-dev · 2026-06-08 08:53
svennickchan: i suspect arm wasn’t happy about m1 already but there’s just nothing they could at that point #asahi-dev · 2026-06-08 08:46
sventhey cannot disable more stuff on m1/2#asahi-dev · 2026-06-08 08:37
sventhat required new hardware #asahi-dev · 2026-06-08 08:37
sventhey can’t disable sysregs on m2/3#asahi-dev · 2026-06-08 07:29
svenwould make sense that it needs some communication with sep #asahi-dev · 2026-06-08 07:28
svencould be that txm talks to sep before and sets some stuff up#asahi-dev · 2026-06-08 07:27
svendoes xnu crash or complain?#asahi-dev · 2026-06-08 07:26
svenI don’t know what exactly is disabled though but this never was a problem on m1/2#asahi-dev · 2026-06-08 07:26
svenit disables some drm stuff but xnu should be aware of that #asahi-dev · 2026-06-08 06:29
sventhe harder part was getting the setup iboot does done correctly because it loads xnu and sptm and txm segments at specific memory locations and passes them using the adt #asahi-dev · 2026-06-08 06:28
svenhopefully they’ll release their code soonish #asahi-dev · 2026-06-08 06:28
svenah yeah, someone got sptm running on m2 with nested virt #asahi-dev · 2026-06-08 06:11
svenHm?#asahi-dev · 2026-06-07 21:23
svenvirtualizing it on m2/m3 is annoying but doesn't require any patching#asahi-dev · 2026-06-07 21:23
svenand sptm really only matters for m4+#asahi-dev · 2026-06-07 21:21
sven"early" as in a few years before m1 iirc#asahi-dev · 2026-06-07 21:21
svenoriginally just for PPL which was also present on M1#asahi-dev · 2026-06-07 21:20
svengxf has been there since some early iphone#asahi-dev · 2026-06-07 17:40
sventhat's not gonna work#asahi-dev · 2026-06-07 17:40
sven26.5 on m2 already has sptm iirc#asahi-dev · 2026-06-07 17:35
svenwhich macos version? you should also see some serial output from xnu at that point i think#asahi-dev · 2026-06-07 17:33
svenand t6022 is then just 2x t6021 dies with a very high speed connection inbetween so that one likely uses the same kernel as well#asahi-dev · 2026-06-07 17:33
svent6020 and t6021 are rather similar, one is a effectively a cut-down version of the other#asahi-dev · 2026-06-07 17:32
svenit probably uses the t6020 kernel#asahi-dev · 2026-06-07 14:42
sveni guess we really should add apple,firmware-abi to any new thing we upstream to prevent this nvme mess :(#asahi-dev · 2026-06-07 14:21
sveni still don't get why they had to change those MMIO locations with a firmware update :/