#asahi · 2023-07-11 17:21
mazjannau: I'm reworking a bunch of stuff -- v3 has proved to be difficult to manufacture (only one good board out of 5)#asahi · 2023-06-27 14:29
mazcheers.#asahi · 2023-06-27 14:28
mazlflde: feel free to propose an update to the documentation.#asahi · 2023-06-27 14:27
mazwell, there is tons of "plain and simple" things that could be advertised. bizarrely, some of us prefer focusing on writing code...#asahi · 2023-06-27 14:26
mazdottedmag: we seem to get that question every 5 hour or so...#asahi · 2023-04-13 08:27
mazmarcan: I'm all for keeping the kernel driver simple and rely on userspace to deal with the outcome (that's what i do for chromebooks).#asahi · 2023-04-13 08:24
mazmarcan: can't this be dealt with in userspace with something like keyd?#asahi · 2023-04-04 13:59
mazonly if you consider that time is finite.#asahi · 2023-04-04 13:58
mazthe usual answer to that sort of question is that it gets exponentially delayed...#asahi · 2023-03-28 14:14
mazI'd definitely expect that. my mini with 11.something is unlikely to ever see a monitor attached to it.#asahi · 2023-03-28 14:13
mazeven better!#asahi · 2023-03-28 14:12
mazas long as the GPU driver stays optional, I'm OK with that... ;-)#asahi · 2023-03-28 13:36
mazsame thing here. the machines have whatever macOS version was on it when they were shipped, and it is unlikely they will ever see an upgrade.#asahi · 2023-03-27 10:20
mazbecause it simplifies the guest by only having PV drivers?#asahi · 2023-03-27 10:20
mazbecause it just works?#asahi · 2023-01-26 07:30
mazand really, I'm not keen on it being merged anywhere before it goes upstream. people who want to play with it will have to build their own kernel (will eat your data, and all that).#asahi · 2023-01-26 07:28
mazbalrog: if you're interested in FEAT_NV2, you're welcome to help reviewing the patches. they aren't going to review themselves, nor bemerged by default! ;-)#asahi · 2023-01-03 10:20
mazeric_engestrom: I don't think they are. there is no firmware to actually handle the storage, AFAIK.#asahi · 2022-12-28 14:19
mazdo you have a pointer to a discussion?#asahi · 2022-12-28 14:19
mazmarcan: not that I can remember, but I also never really cared about s2idle...#asahi · 2022-12-15 11:07
mazit doesn't.#asahi · 2022-12-15 09:55
mazKVM will give you arm64 virtualisation. not emulation.#asahi · 2022-12-15 09:54
mazCyrinux: there is nothing to retry. you're trying to run x86 emulation. without extra support for this mythical TSO mode, it will always be slow as hell.#asahi · 2022-12-15 09:48
mazCyrinux: KVM works on M1/M2. it just won't help when it comes to x86 emulation.#asahi · 2022-12-07 09:30
mazChainfire: I had issues with X not coming up, solved by not compiling the efifb driver... YMMV.#asahi · 2022-10-31 11:06
mazkit_ty_kate: been using FF 105 (debian package) for some time without issue. this is in a 4kB guest though, which could hide a number of sins.#asahi · 2022-10-13 15:22
mazmarcan: I'll go back to my cave. at least, the only idiot there is me.#asahi · 2022-10-13 15:21
mazthis is all wrong on so many levels... :-/#asahi · 2022-10-13 14:16
mazthe fact that it works on x86 isn't much of a proof of nothing being wrong, I'm afraid.#asahi · 2022-10-13 14:15
mazI can't imagine how the kernel finds the initramfs if nothing tells it where it is. the EFI stub in the kernel expects to find the address in the DT.#asahi · 2022-09-27 14:46
mazfor some definition of "work", yes.#asahi · 2022-09-09 13:06
mazI don't really see how we can answer macos feature questions here...#asahi · 2022-09-01 08:32
maz(re smc driver fix)#asahi · 2022-09-01 08:31
mazj`ey: yup, looks good to me.#asahi · 2022-08-29 16:03
mazj`ey: ah, awesome news! only need to recalibrate to this timezone, and parse the messages from the past week... ;-)#asahi · 2022-08-26 16:19
mazTLBs (can't even spell this morning)#asahi · 2022-08-26 16:19
mazmeh. I'll go back to TBLs... :D#asahi · 2022-08-26 16:18
mazright. I guess this is an overly complicated tagging system for each packet being sent down that pipe, and the other guy demuxes it...#asahi · 2022-08-26 16:12
mazis DP tunelling related to the 'chaining' of TB devices?#asahi · 2022-08-26 16:12
mazyeah, this is pretty ugly :-/#asahi · 2022-08-26 16:10
mazwho needs display when you can have PCIe!!!!!!! :D#asahi · 2022-08-26 16:08
mazsven: TB or 4k? hmmm...#asahi · 2022-08-26 16:06
mazright. I should try and use this, as I'd like some perf number between 4k and 16k outside of a VM...#asahi · 2022-08-26 16:05
mazah yes, I find the IOMMU API horrible. but really, we need something in, and maybe use this as leverage to improve this stuff.#asahi · 2022-08-26 16:04
mazor am I looking at something else?#asahi · 2022-08-26 16:03
mazrmk: wot? I don't think I said much on this...#asahi · 2022-08-15 15:47
maz(and I believe they are)#asahi · 2022-08-15 15:46
mazyup. but so would be Apple if they are serious about running VMs.#asahi · 2022-08-15 15:45
mazto be clear, none of the implementations were as relaxed as the spec allowed it. but the wq bug is indeed a really nice investigation.