#asahi-re · 2022-04-06 14:39
sventhat doesn't do xart though iirc#asahi-re · 2022-04-06 14:39
svenohh... right, i actually wrote something to boot sep from m1n1#asahi-re · 2022-04-06 14:39
svenerm, s/write-only/read-only/ :D#asahi-re · 2022-04-06 14:38
svenyou need to map the firmware, boot SEP, and then handle requests from the xart endpoint#asahi-re · 2022-04-06 14:38
svenTZ0 is already setup, that's write-only by the time we get control#asahi-re · 2022-04-06 14:37
svenuh, not really#asahi-re · 2022-04-06 14:29
sveniBoot shuts SEP down again and we have to re-start it and handle the xART endpoint#asahi-re · 2022-04-06 14:28
svenno#asahi-re · 2022-04-06 14:24
sveni also have some description of the xART format somewhere required to boot it#asahi-re · 2022-04-06 14:23
svenhttps://github.com/AsahiLinux/m1n1/pull/85/files don't remember why I never followed up, but that was a very early tracer#asahi-re · 2022-04-06 12:35
svens/with/and/#asahi-re · 2022-04-06 12:34
sventhere's a difference between criminal with civil law#asahi-re · 2022-04-06 12:33
svennot convinced about that, but i'm also not a lawyer#asahi-re · 2022-04-06 12:24
sventhey can pretty much decide to stop doing business with you for whatever reason#asahi-re · 2022-04-03 11:04
sventhe easiest way is probably to run macOS, connect some thunderbolt device and then break in the hypervisor and read it from there#asahi-re · 2022-04-03 11:03
svenyes#asahi-re · 2022-04-03 10:50
svenbut if you do that without ATCPHY or handling that CIO co-processor you get ~2 seconds until the entire SoC reboots#asahi-re · 2022-04-03 10:49
sventechnically it only requires you to poke some "thunderbolt clock" or something that's different from the normal PMGR stuff#asahi-re · 2022-04-03 10:49
svengetting the thunderbolt DART up requires ATCPHY to be up and that CIO co-processor to be running as well#asahi-re · 2022-04-03 10:49
sventhat's more complex#asahi-re · 2022-04-03 09:33
svenI’ve never seen a bit to indicate “don’t add this to the tlb” but “accesses to this page bypasses all caches” is pretty common#asahi-re · 2022-04-03 07:19
svenwhich wouldn't be great if we accidentally set that for all pages :D#asahi-re · 2022-04-03 07:19
svenuh, uncacheable might also mean that access to that page bypasses all caches#asahi-re · 2022-04-02 14:58
svendynamic kext loading can't work because of this kernel-readonly-region thing#asahi-re · 2022-04-02 14:58
svenall those kexts are merged into this kernelcache thing and loaded at once#asahi-re · 2022-04-02 14:57
svenSystem/Library/Extensions should always exist, but it's effectively unused anyway#asahi-re · 2022-03-28 10:23
sveniirc it supports one more pagetable level now, and presumably a bigger iova space with that#asahi-re · 2022-03-28 10:23
svenI very briefly looked at that t8110 dart kext and it seems that it changed quite a bit #asahi-re · 2022-03-28 10:22
sventhe best DART docs we have is probably the Linux driver and that dart.py :D#asahi-re · 2022-03-28 10:21
svenand the registers look a bit different #asahi-re · 2022-03-28 10:21
sventhere’s an older version from the iPhones that only does 4K pages #asahi-re · 2022-03-28 10:21
sventhat’s probably for the DART from long before m1 was a thing :D#asahi-re · 2022-03-19 08:00
svenand there is that someone :)#asahi-re · 2022-03-19 07:59
sven*ran#asahi-re · 2022-03-19 07:59
sveniirc someone else ram into a similar panic a few days ago#asahi-re · 2022-03-16 08:34
svenand from what I can tell just constant values that they'd usually stuff into a tunable#asahi-re · 2022-03-16 08:33
svenwith fun registers like LANE0_AUSPMA_TX_SHM_BLK_TXA_IMP_REG2.PMA_TXA_MARGIN or LANE0_AUSPMA_RX_SHM_TJ_BLK_RXA_VREF_CTRL22.PMA_RXA_VREF_BIAS_SEL#asahi-re · 2022-03-16 08:33
svendisplay port bring up is ~200 lines in the mmio trace and a bunch of that seems to be some kind of "write magic value to magic register"#asahi-re · 2022-03-02 14:37
svenah.. that actually fits with that general control register as well#asahi-re · 2022-03-02 12:20
svenalso doesn't really matter, but i'm curious what that thing might be#asahi-re · 2022-03-02 12:20
svenmarcan: does SIDDQ in relation to some kind of "general USB2 PHY control/reset register" that also contains other reset bits tell you anything? googling resulted in something called "IDDQ" / quiescent current but that's way outside stuff i'm familiar with#asahi-re · 2022-02-28 15:03
svenchances are that one's similar to t6000/t8103 as well then#asahi-re · 2022-02-28 15:02
svenpossibly :D#asahi-re · 2022-02-28 15:02
svenlol#asahi-re · 2022-02-28 15:02
sventhere should also be T8103TypeCPhy and T6000TypeCPhy#asahi-re · 2022-02-28 15:01
sveni guess so#asahi-re · 2022-02-28 14:57
svenmost weird one to me is AUS_CMN_SHM_BLK_VREG_REG_0.VREG_TRIM. dunno what CMN (common? not very helpful though) or SHM could be.#asahi-re · 2022-02-28 14:46
svens/same/name/#asahi-re · 2022-02-28 14:46
sveneven if that "PLL" wasn't in the same those very much sound like clock tunables#asahi-re · 2022-02-28 14:45
svenmarcan: ^-- those are the fuses from xnu debug output for atc-phy1