#asahi-dev · 2021-02-12 19:24
modwizcodehttps://twitter.com/modwizcode/status/1360308258519605254?s=20 I got the AIC emulation working for real finally#asahi-dev · 2021-02-11 19:08
modwizcodeI mean at the moment I think that would be premature. I have a lot on my plate. I suspect it'd be better overall to do something like testing of features that marcan implements.#asahi-dev · 2021-02-11 19:03
modwizcodeWish I had a Mac to work on stuff, probably for the best though. I have enough distractions#asahi-dev · 2021-02-11 18:58
modwizcodeor just unknown at least#asahi-dev · 2021-02-11 18:57
modwizcodeThe other bits (besides the self bit) in the ACK register are unused then?#asahi-dev · 2021-02-11 18:57
modwizcodeI was going to update the doc after I verified things further with you/emulation#asahi-dev · 2021-02-11 18:57
modwizcodeRight#asahi-dev · 2021-02-11 18:57
modwizcodeI think the confusion is that the register in linux that's IPI_SEND you originally called FLAG_SET and IPI_ACK was FLAG_CLEAR, but in reality you write the other bit in the ACK reg, so it's not like SET/CLEAR#asahi-dev · 2021-02-11 18:55
modwizcodeWith IPIs it doesn't tell you the sender though does it?#asahi-dev · 2021-02-11 18:54
modwizcodeDoes reading the AIC "reason" reg mask IPIs automatically or are they masked immediately? If so do you know if it masks only IPIs from the initial IPI it's handling or all of them?#asahi-dev · 2021-02-11 18:52
modwizcodedevice tree#asahi-dev · 2021-02-11 18:49
modwizcodeThere's no fake pci on the M1#asahi-dev · 2021-02-11 18:49
modwizcodeEveryone loves a new connector#asahi-dev · 2021-02-11 18:49
modwizcodecertainly not#asahi-dev · 2021-02-11 18:47
modwizcodeAnd AMD systems with TB are still rare if not nonexistant.#asahi-dev · 2021-02-11 18:47
modwizcodeI'm assuming they let the publicity out before the lawyers actually finished hammering out the details#asahi-dev · 2021-02-11 18:46
modwizcodeThe board layout on those cards looks really nice#asahi-dev · 2021-02-11 18:45
modwizcodethose are some truly tiny fiber optic cables in those#asahi-dev · 2021-02-11 18:45
modwizcodeheh#asahi-dev · 2021-02-11 18:44
modwizcodeI figured they just used an SFP with a special port#asahi-dev · 2021-02-11 18:43
modwizcodeoh?#asahi-dev · 2021-02-11 18:43
modwizcodeIt would be in the SFP yes#asahi-dev · 2021-02-11 18:43
modwizcodewell but that wasn't the point of light peak tho#asahi-dev · 2021-02-11 18:42
modwizcodeThe problem is of course power#asahi-dev · 2021-02-11 18:42
modwizcodeIt was supposed to be plastic fiber based I think#asahi-dev · 2021-02-11 18:42
modwizcode(the datacenter ones not the plastic ones)#asahi-dev · 2021-02-11 18:41
modwizcodeI mean I've handled more than a few optical fiber cablely things, they're not *that* fragile#asahi-dev · 2021-02-11 18:41
modwizcodeWas that before GbE fiber specs came out?#asahi-dev · 2021-02-11 18:40
modwizcodeActually that will probably be the next upgrade to USB-C, a fiber but only in some cables ;)#asahi-dev · 2021-02-11 18:40
modwizcodeThat's cursed#asahi-dev · 2021-02-11 18:40
modwizcoderight#asahi-dev · 2021-02-11 18:40
modwizcodeOhhh#asahi-dev · 2021-02-11 18:40
modwizcodeReally?#asahi-dev · 2021-02-11 18:40
modwizcodeI'd love to know just how hard it was to make those optical TB3 cables though#asahi-dev · 2021-02-11 18:39
modwizcodeTB was supposed to support both still but I think only those weird corning cables exist#asahi-dev · 2021-02-11 18:39
modwizcodethat would have been terrible but it sounds so futuristic and cool#asahi-dev · 2021-02-11 18:38
modwizcodeIt very much reads as "marketing said we had to make this not work, but I have an AMD computer" or something lol#asahi-dev · 2021-02-11 18:38
modwizcode"the engineers wanted to use them for something else"#asahi-dev · 2021-02-11 18:38
modwizcodeHuuuuh I was not aware of that lol#asahi-dev · 2021-02-11 18:37
modwizcodeEven after opening up TB3 I think there were still issues with being forced to use Intel controllers#asahi-dev · 2021-02-11 18:37
modwizcodeThey wouldn't let you use non-intel silicon even if you wanted to I think#asahi-dev · 2021-02-11 18:36
modwizcodeI think it was a licensening requirement#asahi-dev · 2021-02-11 18:36
modwizcode(which is why they all have the same weird bugs!)#asahi-dev · 2021-02-11 18:36
modwizcodeI thought that all existing designs were IP licensed so this is the first I've heard#asahi-dev · 2021-02-11 18:36
modwizcodeIs it the first TB device in existance with a non Intel derived controller?#asahi-dev · 2021-02-11 18:35
modwizcodeheh#asahi-dev · 2021-02-11 18:35
modwizcodeOh believe me I have#asahi-dev · 2021-02-11 18:35
modwizcodeI'm amazed that works at all honestly#asahi-dev · 2021-02-11 18:34
modwizcodeTB just tunnels PCIe TLP right?#asahi-dev · 2021-02-11 18:33
modwizcodeLots of people do that kind of thing with AXI although that's admittedly a bit different, still ends up with things like routing a memory controller to AXI only to immediately send it to another IP block to unwrap AXI.