#asahi-dev · 2022-04-06 17:36
nicolas17oh yeah DCP#asahi-dev · 2022-04-06 17:35
nicolas17but I wasn't sure if that was right#asahi-dev · 2022-04-06 17:34
nicolas17sven: my thought was "later chips are doable, even though it may take a lot of work for some peripherals; but 32-bit is a step way too far"#asahi-dev · 2022-04-06 17:27
nicolas17but I don't actually know; how much of asahi's work would be even relevant?#asahi-dev · 2022-04-06 17:27
nicolas17someone yesterday asked about getting Linux running on A6X, I personally think it would be interesting to get it working on A-series chips, but A6X would take an absurd amount of work because it's 32-bit#asahi-dev · 2022-04-06 17:25
nicolas17now that there's more people around#asahi-dev · 2022-04-06 03:23
nicolas17ah I see it was said as a separate word earlier#asahi-dev · 2022-04-06 03:22
nicolas17lol, can't you make it trigger on the whole word only?#asahi-dev · 2022-04-06 03:21
nicolas17https://data.nicolas17.xyz/DeviceTree.p103ap.img3#asahi-dev · 2022-04-06 03:20
nicolas17rqou_: can the ADT be extracted from the ipsw too? or is that in a different format?#asahi-dev · 2022-04-06 03:08
nicolas17because otherwise it's *way* too different from the existing M1 work#asahi-dev · 2022-04-06 03:07
nicolas17I'd like to see asahi ported to (checkm8-affected) A-series chips too, but 64-bit feels like a basic requirement...#asahi-dev · 2022-04-06 03:01
nicolas17I don't know if there's even a good exploit for checkm8 for that SoC#asahi-dev · 2022-04-06 03:00
nicolas17A6X is 32-bit lol, it would need a huge amount of changes#asahi-dev · 2022-04-06 02:59
nicolas17wait what#asahi-dev · 2022-04-05 20:28
nicolas17I don't know if github would ever make raw.githubusercontent.com return 403 (blocking IPs?) or why, but if there is a fallback, we might as well use it in that case too#asahi-dev · 2022-04-05 20:28
nicolas17you may want to add -f to the curl command to make it exit non-zero on HTTP error codes too#asahi-dev · 2022-04-05 20:27
nicolas17but if the server (or the proxy intercepting it >.>) returns an HTTP error code like 403, it will exit with zero code#asahi-dev · 2022-04-05 20:27
nicolas17if curl can't connect to the server or resolve DNS or something like that, it will return a non-zero exit code#asahi-dev · 2022-04-05 20:26
nicolas17https://github.com/AsahiLinux/asahi-installer/commit/d16bc298cfaaf1018672122208a42baf393f6b2d hmm#asahi-dev · 2022-04-03 02:25
nicolas17>.>#asahi-dev · 2022-04-03 02:23
nicolas17it could also be ffmpeg's fault somehow?#asahi-dev · 2022-04-03 02:17
nicolas17speaking of mobile devices, I wonder if there's anyone interested / crazy enough to port asahi to A10 or something :p#asahi-dev · 2022-04-03 02:16
nicolas17mhm#asahi-dev · 2022-04-03 02:14
nicolas17if it's >= 1 core, it may be bottlenecking doing something dumb on the CPU...#asahi-dev · 2022-04-03 02:13
nicolas17what's ffmpeg CPU usage like?#asahi-dev · 2022-04-03 02:11
nicolas17like it's throttling rather than being inherently slow#asahi-dev · 2022-04-03 02:11
nicolas17any signs of a hard cap? (like decode never going over 160fps at all)#asahi-dev · 2022-04-03 02:07
nicolas17well you weren't testing h264 *encode*, right?#asahi-dev · 2022-04-03 02:06
nicolas17also maybe if you want all your CPU cores to be dedicated to encoding, pushing the decoder to hardware is also a speed win? idk#asahi-dev · 2022-04-03 02:04
nicolas17maybe the h264 hardware is significantly more power-efficient, but if you're doing a transcode you'd likely want best speed so that's good to know...#asahi-dev · 2022-04-03 02:03
nicolas17is cpu decoding using multiple threads?#asahi-dev · 2022-04-03 01:55
nicolas17chadmed: try "-f null -" for the output so you're *only* testing decoding#asahi-dev · 2022-04-03 01:53
nicolas17are you encoding to prores?#asahi-dev · 2022-04-03 01:53
nicolas17chadmed: sometimes hwaccel is slow because it needs to copy back to RAM, but that shouldn't be the case here, especially on M1#asahi-dev · 2022-04-01 23:14
nicolas17...I didn't even hear of half of those#asahi-dev · 2022-03-28 17:28
nicolas17smart "display controllers" are newer stuff I know nothing about :P#asahi-dev · 2022-03-28 17:27
nicolas17I know a big part of early iOS performance was offloading as much compositing/animation as possible to the GPU (LayerKit/CoreAnimation)#asahi-dev · 2022-03-28 17:21
nicolas17I'm reading chat logs and lol the DCP can alpha-composite?#asahi-dev · 2022-03-28 05:07
nicolas17maybe 'having ASTC' is a stand in for other features that do justify different files but are more awkward to fit into a filename :)#asahi-dev · 2022-03-28 05:05
nicolas17but if the 4.7" iPhone ipsw has 3 different kernels for different devices, it's no big deal, so they do that instead of 3 different ipsws#asahi-dev · 2022-03-28 05:05
nicolas17eg. separating ipsws based on iPhone screen size makes sense, that way they don't bloat 4.7" iPhones with images (like wallpapers which are *large*) that are only used on 5.5" iPhones#asahi-dev · 2022-03-28 05:02
nicolas17it seemed weird that ASTC support was a dimension to split devices by for this purpose#asahi-dev · 2022-03-28 05:02
nicolas17no, they are each for different sets of ipads#asahi-dev · 2022-03-28 05:00
nicolas17iPad_64bit_TouchID_13.7_17H35_Restore.ipsw vs iPad_64bit_TouchID_ASTC_13.7_17H35_Restore.ipsw#asahi-dev · 2022-03-28 04:59
nicolas17oddly "ASTC support" is one of the things used to separate iOS ipsws#asahi-dev · 2022-03-28 04:09
nicolas17^#asahi-dev · 2022-03-28 04:08
nicolas17oh looks like I'm even more confused#asahi-dev · 2022-03-28 04:07
nicolas17the D3D thing is probably for nvidia cards?#asahi-dev · 2022-03-28 04:07
nicolas17chadmed was probably answering your *previous* question :P