#asahi · 2021-04-30 16:29
modwizcodeI wonder if a feature like the GX implementation cuold be generically useful if it was introduced as a core architecture feature#asahi · 2021-04-30 16:27
modwizcodeI think the default on lockdown mode is fairly permissive iirc?#asahi · 2021-04-29 21:44
modwizcodethanks#asahi · 2021-04-29 21:42
modwizcodesven: can you give me a ping when the GX stuff goes up and then if you find out more about the SART?#asahi · 2021-04-29 21:36
modwizcodeah okay#asahi · 2021-04-29 21:36
modwizcodedo the APRR registers still exit too in addition to the GX thing?#asahi · 2021-04-29 21:34
modwizcodeah#asahi · 2021-04-29 21:33
modwizcodeminus the use of cpu instructions#asahi · 2021-04-29 21:33
modwizcodebecause that sounds basically identical to APRR#asahi · 2021-04-29 21:33
modwizcodeNot knowing anything (and having forgotten a lot about APRR) is there a reason they changed how that was happening?#asahi · 2021-04-29 21:32
modwizcodeIt's more generic then?#asahi · 2021-04-29 21:32
modwizcodeGX mode basically prevents you from changing the page tables right?#asahi · 2021-04-29 21:31
modwizcodethat might explain the why#asahi · 2021-04-29 21:31
modwizcodedo we at least have any idea *why* they would bother with two (or three even?) distinct mapping formats#asahi · 2021-04-29 21:30
modwizcodeah#asahi · 2021-04-29 21:30
modwizcodeor whatever it's called#asahi · 2021-04-29 21:30
modwizcodehow about the SART?#asahi · 2021-04-29 21:30
modwizcodecool#asahi · 2021-04-29 21:29
modwizcodewell I guess I won't bother trying to pull up the wiki then#asahi · 2021-04-29 21:28
modwizcodethat's great!#asahi · 2021-04-29 21:22
modwizcodesven: how far did you get on guarded exec stuff?#asahi · 2021-03-24 18:57
modwizcodenight!#asahi · 2021-03-24 02:23
modwizcodewait marcan you wrote the qemu xhci code? Heh#asahi · 2021-03-24 02:23
modwizcodewait marcan you wrote the qemu xhci code? Heh#asahi · 2021-03-01 15:38
modwizcodeHeh that's kind of devious#asahi · 2021-03-01 15:36
modwizcodethanks#asahi · 2021-03-01 15:27
modwizcodemarcan: can you link to the commit that explains the dsb thing?#asahi · 2021-02-28 18:38
modwizcodeIt's very strange that it's not properly documented though, that seems like the most important guidence to give to people#asahi · 2021-02-28 18:37
modwizcodeTheoretically I think that if dmb could even make that gurantee you'd need a full dmb sy, I don't think oshst would be enough either... I'm assuming that luck is what make it enough and that memory ops just complete way too quick or the cache coherance compensates somehow.#asahi · 2021-02-27 15:33
modwizcodeI guess the cache coherence stuff could theoretically not complete before the device write/read does? That still seems weird (and extermely unlikely)#asahi · 2021-02-27 15:31
modwizcodeI found a random presentation on dmb vs dsb on arm and it implies my understanding is correct. dmb is enough to order the operations right, but because completion might not have happened there's a data race between the operations being observed remotely.#asahi · 2021-02-24 17:16
modwizcodeI hope that's wrong though because that contradicts a sane interpretation of DMB that affects outer shareable regions like device memory...#asahi · 2021-02-24 17:07
modwizcodewrt device accesses that is.#asahi · 2021-02-24 17:06
modwizcodeMy interpretation seems to indicate that DMB is safe for orderings between both on the same core, but DSB is required for multicore because you need to be sure the writes have "completed" which seems to be the only gurantee of external visibility?#asahi · 2021-02-24 17:04
modwizcodeThe ARM is annoyingly vague on that specific issue too.#asahi · 2021-02-24 16:51
modwizcodeYeah I wasn't clear in my answer and even if I was I think I was wrong.#asahi · 2021-02-24 16:33
modwizcodemarcan: afaik dmb is only guranteed as a memory ordering op. i.e. do not issue memory ops out of order. It offers no assurances beyond that. ARM docs usually use it as part of locking, but from what I can tell it offers no meaning in the face of io.#asahi · 2021-02-24 16:31
modwizcodeWell that was a fun few days of chat to catch up on#asahi · 2021-02-24 16:31
modwizcodeheh#asahi · 2021-02-21 19:26
modwizcodeNintendo in general is pretty litigious and doesn't seem to care a lot about project specifics as much.#asahi · 2021-02-21 18:27
modwizcodeThat's the trick they use to avoid having to explicitly say what they don't want done. No chance of arguing, just "oh it violates the license"#asahi · 2021-02-19 00:56
modwizcodeI thought PPL was a kernel integrity support feature, or did I confuse thase?#asahi · 2021-02-18 21:30
modwizcodeThose docs are slightly more detailed than I would expect apple to normally give#asahi · 2021-02-18 21:30
modwizcodeThat express cards with power reserve thing is actually a pretty cool idea#asahi · 2021-02-17 20:26
modwizcodeI was yes#asahi · 2021-02-17 19:22
modwizcodeinteresting#asahi · 2021-02-17 18:52
modwizcodeIt's interesting that apple has (in addition to their generic Hypervisor acceleration extension framework) a second API just for easily bootstrapping Linux VMs, but it's specific to Linux VMs.#asahi · 2021-02-16 17:39
modwizcodei am aware#asahi · 2021-02-16 17:38
modwizcodeHow are they removing the APRR stuff from the core of XNU? or is it still in there but stripped out before release? APRR seems pretty low level#asahi · 2021-02-16 17:33
modwizcodenight