2022-08-16
00:06surge9n has joined #asahi-gpu
00:14surge9n has quit [Ping timeout: 480 seconds]
00:21Nspace has quit []
00:21Nspace has joined #asahi-gpu
00:39surge9n has joined #asahi-gpu
00:41uniq has joined #asahi-gpu
00:47surge9n has quit [Ping timeout: 480 seconds]
01:03uniq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
01:12surge9n has joined #asahi-gpu
01:16uniq has joined #asahi-gpu
01:20surge9n has quit [Ping timeout: 480 seconds]
01:43surge9n has joined #asahi-gpu
01:51surge9n has quit [Ping timeout: 480 seconds]
02:01uniq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
02:16uniq has joined #asahi-gpu
02:28uniq is now known as sortIuniq
02:46sortIuniq is now known as uniq
03:00skipwich has quit [Quit: DISCONNECT]
03:01skipwich has joined #asahi-gpu
03:27uniq has quit [Ping timeout: 480 seconds]
03:46uniq has joined #asahi-gpu
04:14chadmed has quit [Quit: Konversation terminated!]
04:17chadmed has joined #asahi-gpu
05:00c10l has quit [Quit: Bye o/]
05:00c10l has joined #asahi-gpu
05:58MajorBiscuit has joined #asahi-gpu
06:00Major_Biscuit has joined #asahi-gpu
06:06MajorBiscuit has quit [Ping timeout: 480 seconds]
06:23uniq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
07:11uniq has joined #asahi-gpu
07:30phireTheoretically good. I have no idea why nobody has implemented it. Maybe there is some major roadblock I'm unaware of, or maybe separate processes is good enough.
07:31uniq has quit [Ping timeout: 480 seconds]
08:24mort_I've made the mistake of making something in python, having it be too slow, then instead of rewriting it in C++ or something I multiprocess the python code, you end up with something that's still too slow but now heats up the room
08:41surge9n has joined #asahi-gpu
08:44uniq has joined #asahi-gpu
08:45phireI have a bad habbit of whenever I think performance might be an issue in the future, I write in in c++
08:47phireuseally I would have been better off waiting until preformance was actually an issue (which almost never happens), and then rewriting it. or the hot parts of it
08:53kode54though there was one time in high school where I got a random idea while I was supposed to be doing my assignments
08:54kode54I believe I wrote code to use bit rotations to rotate a horizontal raster font, to rotate it to vertical raster for a dot matrix printer
08:55phireHa! I'm kind of the opinion that everyone should know a high-preformance language, and a low-brainpower language. In my case, C++ for high preformance, python or typescript for low brainpower.
08:57phireI used to be a huge fan of c, because it was small enough to know the whole language, and most of the standard library
09:00uniq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
09:01uniq has joined #asahi-gpu
09:15surge9n has quit [Ping timeout: 480 seconds]
09:17pjakobsson has quit [Ping timeout: 480 seconds]
09:22uniq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
09:25pjakobsson has joined #asahi-gpu
09:31uniq has joined #asahi-gpu
09:42surge9n has joined #asahi-gpu
09:47uniq has quit [Quit: Textual IRC Client: www.textualapp.com]
10:25pjakobsson has quit [Ping timeout: 480 seconds]
11:28linaI think it's going to be my go-to language for things that I'd write in Python but I need better performance for
11:30mort_I found python nice to work with actually when dealing with giant data sets, it's surprisingly effective to just give yourself a ton of swap space and store all your data in python's built-in datastructures, since then you get pretty good powerful disk-backed but RAM-cached datastructures for free
11:31mort_the same would work for most other languages but not for JS sadly, since that has a much more limited address space it seems
11:32linaPython is actually pretty good at moving bulk data around, but the moment you need an inner loop, it does kind of fall apart...
11:33mort_it's my go-to tool for natural language processing stuff but it's really bad at that tbh unless all you do is use libraries which have the inner loop in C
11:33linaWhen I was using D it was a bit scary, because I was writing algorithms the same way I'd write them in Python for experiments, even poorly optimized textbook O(n^2) stuff... and then the app kept running at 60 FPS.
11:33linaThat was a bit of a revelation, because in C that would've taken much more code, and in Python it would've been way too slow...
11:34mort_but it makes sense, when you consider that operations which in low-level languages are literally a single register-register instruction is at least a hash table lookup or two lol
11:35linaYeah, hah. Though on the flipside, you can do some really crazy things with the object model. It was perfect for prototyping the GPU driver for that reason.
11:36linaSpend an hour or so to get fancy color-coded nested GPU structure diff displays, that sort of thing...
11:37mort_it will be the first major test of Rust in the kernel, right? Everything else seems to mostly have been proof of concept stuff unless I've missed something
11:37linaYeah, the first large driver, though people are working on reasonably nontrivial drivers like NVMe too.
11:39mort_for a monolithic kernel with way too much ring 0 code, rust truly seems like a good to dramatically reduce security bugs
11:48pjakobsson has joined #asahi-gpu
12:18_jannau_Sobek[m]: nothing beyond https://lore.kernel.org/rust-for-linux/CANiq72=L3soKyYPVu4pm0qzUrj7ovYfyXNGQJU1r0AAhXA63aA@mail.gmail.com/T/#m011e38cb8290531fe19aaa83d0903722959d7ee0 in public as far as I'm aware
12:19tdsrts^ has joined #asahi-gpu
12:22sventhe kernel driver is just very complex because it has to support all the random broken hardware out there and because it has to be fast
12:27svenhrm.. thinking about it some more you'd gain some rust bindings for common kernel apis like DMA
13:23linaYes, the folks writing the NVMe driver are working on PCIe and DMA abstractions too as a side effect
13:42Chainfire has joined #asahi-gpu
13:44tdsrts^ has quit [Ping timeout: 480 seconds]
13:46ChainfirePython is slow, yes, certainly much too slow for domains like drivers. But its expressiveness and resulting development speed (if used correctly - like any dynamic language you can really shoot future-self in the foot) makes it prime for productivity where man-hours are worth much more than CPU cycles
14:05mort_I never looked at dlang really, the fact that it was pushed so hard while the compiler was closed source really left a bad taste in my mouth
14:06mort_sort of like how C#'s closed source history would make me unlikely to choose it even if there are open-source implementations of it today
14:06mort_shows a deep philosophical difference between myself and whatever company or community is behind D I suppose
14:22linaBy the way, for those who missed it, I'm going to stream starting the actual GPU driver tomorrow ^^
14:33Chainfirelina> will this open the gates for compute as well, or is that something again far beyond graphics?
14:39linaCompute needs someone to care about it... I could work on adding kernel support, but without someone to first work on the OpenCL side in Mesa, that's not very useful ^^;
14:41linaIt's not especially hard or anything as far as I know, but it's something someone needs to explicitly pick up
15:13ChainfireI care more about it than graphics :') But I have neither the time nor the relevant experience to work on it, at least short term. Either way, that's good to know, thanks.
15:16ChainfireThis may perhaps be silly question, but would it be easier to support Metal than OpenCL ?
15:17danielslina: (ah sorry, I didn't spot the subtlety of compute vs. CL - core compute support is there, it's just not hooked up to Clover because Clover ... rusticl ftw!)
15:27Race has quit [Ping timeout: 480 seconds]
15:33linadaniels: Huh, last week I was looking at the kernel driver and I think the compute queue was not used? Or is that going through the 3D ones?
15:34mort_compute is just part of the normal vulkan API, right? So if vulkan gets sorted out, won't that kind of take care of compute?
15:35linaIf it's going through the 3D queue, the same trick would work on AGX without kernel driver changes I think (Apple introduced compute instantiations from render batches AIUI), but you still need someone to hook up all the userspace plumbing to make it all work...
15:36linaBut it shouldn't be that hard to support the dedicated compute queue, I just don't have any reason to do it / test code to use with it right now.
15:36linaIdeally someone would do what alyssa did for 3D, and make it all work on the macOS kernel driver first, at least to some extent. Then we can hook up the kernel side.
15:54MajorBiscuit has joined #asahi-gpu
15:56Major_Biscuit has quit [Ping timeout: 480 seconds]
16:27MajorBiscuit has quit [Ping timeout: 480 seconds]
16:33bisko has joined #asahi-gpu
17:25c10l has quit [Quit: Bye o/]
17:26c10l has joined #asahi-gpu
18:36Michael[m]123 has joined #asahi-gpu
19:04Votes78 has quit [Killed (NickServ (Too many failed password attempts.))]
19:05Votes78 has joined #asahi-gpu
19:30thinkalex[m] has joined #asahi-gpu
19:31MajorBiscuit has joined #asahi-gpu
19:46alyssa has joined #asahi-gpu
19:47alyssa(Actually, I have a Rusticlfrost branch which got with a few fails of OpenCL 3.0 conformance. on the backburner for now though.)
19:48alyssajust have a "fragment" queue and an "everything else" queue (js0 and js1 respectively ... you saw something about js2 which was a funny compute-only thing which doesn't exist since forever ago)
19:49alyssawhat's missing in panfrost.ko is plumbing to use JS2 for compute-only workloads on ancient Malis
21:32clararussell[m] has joined #asahi-gpu
21:45MajorBiscuit has quit [Ping timeout: 480 seconds]
21:48alyssa has quit [Quit: leaving]
21:53carlosstive[m] has joined #asahi-gpu
21:59kazukih[m] has joined #asahi-gpu
22:18surge9n has quit [Ping timeout: 480 seconds]
22:36Race has joined #asahi-gpu
22:45surge9n has joined #asahi-gpu
22:45RoelAlejandroPerezCandanoza[m] has joined #asahi-gpu
22:53surge9n has quit [Ping timeout: 480 seconds]
22:58Donaldbtc[m] has joined #asahi-gpu
23:16hutchinson70[m] has joined #asahi-gpu
23:17surge9n has joined #asahi-gpu
23:25surge9n has quit [Ping timeout: 480 seconds]
23:51surge9n has joined #asahi-gpu
23:51Donaldbtc[m] has quit [autokilled: This host violated network policy. Mail support@oftc.net if you feel this is in error. (2022-08-16 23:51:39)]
23:59surge9n has quit [Ping timeout: 480 seconds]