#
…

2026-02-06

00:03slendidev has joined #asahi-dev
00:05alicela1n has joined #asahi-dev
00:09nepeat has quit [Quit: ZNC - https://znc.in]
00:09nepeat has joined #asahi-dev
00:18alicela1n has quit [Quit: Textual IRC Client: www.textualapp.com]
00:18alicela1n has joined #asahi-dev
00:29c10l has quit [Quit: Bye o/]
00:37fl0_id__ has quit [Ping timeout: 480 seconds]
00:38c10l has joined #asahi-dev
00:38chadmed has quit [Quit: Konversation terminated!]
00:39chadmed has joined #asahi-dev
00:48alicela1n has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
00:55mixi has quit [Ping timeout: 480 seconds]
01:06mixi has joined #asahi-dev
01:21fl0_id__ has joined #asahi-dev
01:28chadmed has quit [Read error: Connection reset by peer]
01:29chadmed has joined #asahi-dev
01:59fl0_id__ has quit [Ping timeout: 480 seconds]
02:00chadmed has quit [Quit: Konversation terminated!]
02:00chadmed has joined #asahi-dev
02:18chadmed has quit [Read error: Connection reset by peer]
02:19chadmed has joined #asahi-dev
02:23hdbngr has joined #asahi-dev
02:32hdbngr has quit [Ping timeout: 480 seconds]
02:35hdbngr has joined #asahi-dev
02:39pb17 has quit [Ping timeout: 480 seconds]
02:45pb17 has joined #asahi-dev
02:47chadmed has quit [Quit: Konversation terminated!]
02:47chadmed has joined #asahi-dev
02:49hdbngr has quit [Ping timeout: 480 seconds]
03:16tobhe_ has joined #asahi-dev
03:18Sai_ has quit [Remote host closed the connection]
03:18Rock has joined #asahi-dev
03:19Rock is now known as Guest1647
03:19freetolga has quit [Quit: Connection closed for inactivity]
03:20chadmed has quit [Read error: Connection reset by peer]
03:21chadmed has joined #asahi-dev
03:23tobhe has quit [Ping timeout: 480 seconds]
03:23luma has quit [Ping timeout: 480 seconds]
03:25john-cabaj has joined #asahi-dev
03:40chadmed has quit [Read error: Connection reset by peer]
03:41chadmed has joined #asahi-dev
03:51hexdump02 has joined #asahi-dev
03:52chadmed has quit [Read error: Connection reset by peer]
03:53hexdump01 has quit [Ping timeout: 480 seconds]
03:53chadmed has joined #asahi-dev
04:03luma has joined #asahi-dev
04:07alicela1n has joined #asahi-dev
04:07alicela1n has quit []
04:21slendidev has quit [Read error: Connection reset by peer]
04:21slendidev has joined #asahi-dev
04:35hdbngr has joined #asahi-dev
04:35pb17 has quit [Ping timeout: 480 seconds]
04:37nora_ has joined #asahi-dev
04:39nora has quit [Ping timeout: 480 seconds]
04:41pb17 has joined #asahi-dev
04:42john-cabaj has quit [Ping timeout: 480 seconds]
04:44hdbngr has quit [Ping timeout: 480 seconds]
04:44prike286 has joined #asahi-dev
04:50prike28 has quit [Ping timeout: 480 seconds]
04:50prike286 is now known as prike28
04:55marvin24 has quit [Ping timeout: 480 seconds]
04:56chadmed has quit [Quit: Konversation terminated!]
04:56chadmed has joined #asahi-dev
05:08HitokiriKincho has joined #asahi-dev
05:21hdbngr has joined #asahi-dev
05:29hdbngr has quit [Ping timeout: 480 seconds]
06:00hdbngr has joined #asahi-dev
06:06croissant has quit []
06:08hdbngr has quit [Ping timeout: 480 seconds]
06:15croissant has joined #asahi-dev
06:28Guest1647 has quit [Remote host closed the connection]
06:29Rock has joined #asahi-dev
06:29Rock is now known as Guest1656
06:35gargantua has quit [Ping timeout: 480 seconds]
06:42pb17 has quit [Ping timeout: 480 seconds]
06:43pb17 has joined #asahi-dev
07:03HitokiriKincho has quit [Quit: Leaving...]
07:13hdbngr has joined #asahi-dev
07:22hdbngr has quit [Ping timeout: 480 seconds]
07:22hdbngr has joined #asahi-dev
07:27seb_ has quit [Quit: Konversation terminated!]
07:30hdbngr has quit [Ping timeout: 480 seconds]
08:09hdbngr has joined #asahi-dev
08:18seb_ has joined #asahi-dev
08:30Rock has joined #asahi-dev
08:30Guest1656 has quit [Read error: Connection reset by peer]
08:31Rock is now known as Guest1669
08:37hdbngr has quit [Ping timeout: 480 seconds]
08:42hdbngr has joined #asahi-dev
09:00hdbngr has quit [Ping timeout: 480 seconds]
09:16hdbngr has joined #asahi-dev
09:23hdbngr has quit [Remote host closed the connection]
09:23hdbngr has joined #asahi-dev
09:31hdbngr has quit [Ping timeout: 480 seconds]
09:47seb_ has quit [Quit: Konversation terminated!]
10:24hdbngr has joined #asahi-dev
10:32hdbngr has quit [Ping timeout: 480 seconds]
10:38arryn has joined #asahi-dev
10:41arrynHello everyone! Besides the drivers and the device tree is there any documentation on the devices, their registers and so on? I'm currently poking around a little, writing my own code for testing, trying to poke at the platform a little so that down the line I can maybe help reverse engineering a little.
10:43noopwafelsven/whoever, how do I make usb do usb things without atcphy, what am I missing
10:44sveneasiest way is to add empty tunables, it should fall back to usb2 only then
10:45noopwafelaha :D so "don't". thanks!
10:45svenlol :D
10:45noopwafeltracing the likely devices seems to only show _one_ tunable being applied
10:46svenyou could also use the generic snps,dwc3 or whatever compatible in the dwc3 node and remove all tipd and phy references
10:46svenbut then it’s gonna work once
10:46svenMaybe
10:47svenI can send you xnu bootargs later
10:47svenIt’ll tell you which tunables it applies
10:47noopwafeloh nice! for now I tried to add the new tunables (they changed all the naaames) to m1n1..
10:48svenyeah, that’s what I assumed from looking at the adt ages ago
10:48svenbut other than tunables it’s the same driver essentially
10:48arrynnoopwafel: I saw this but it seems veeeeery incomplete
10:50hdbngr has joined #asahi-dev
10:51mazarryn: incomplete because things need to be reverse engineered?
10:51noopwafelif you're looking for particular pieces then often the irc logs can be helpful.
10:53arrynmaz: I am mostly talking about M1/M2 things right now that already have drivers the (asahi) linux repos. I was just wondering if someone took the time to document stuff somewhere without people needing to read the driver.
10:54svenNot really, there’s usually some python scripts but that’s a similar level of documentation as the drivers
10:55arrynnoopwafel: I am not reaaally looking for anything in particular. What I'm curious about is currently "How does the keyboard work? What devices does it need to function?" etc sort of questions. Basically information gathering stage so that I can understand how more things work beyond what the devicetree shows me
10:55arrynsven: you mean the ones in m1n1/proxytool/experiments and such?
10:56arrynyeah I saw those
10:56sventhat’s all the documentation there is
10:58arrynI'm wondering would it be beneficial to document the hardware in text? (like the link noopwafel sent) Maybe I can try to do that a little and then contribute like that for now? I'm sure it would help other people that are like me out
10:59svenit's definitely can't hurt and if you think it helps other people out go for it
10:59hdbngr has quit [Ping timeout: 480 seconds]
10:59arrynalso because reading linux driver code is not really easy for me haha
10:59svendebug=0x14e serial=3 apcie=0xfffffffe wdt=-1 AppleSynopsysDRDUSBXHCI-debug=0xffff AppleT8103TypeCPhy-debug=0xffff AppleT8103USB40XHCITypeCPort-debug=0xffff AppleT8103USBXDCI-debug=0xffff AppleTypeCPhy-debug=0xffff AppleTypeCPhyCIOInterface-debug=0xffff AppleTypeCPhyDisplayPortInterface-debug=0xffffffff AppleTypeCPhyInterface-debug=0xffff
10:59svenAppleTypeCPhyUSBInterface-debug=0xffff AppleUSBHostResourcesTypeC-debug=0xffff
10:59svennoopwafel: ^--
11:00svennot sure what's actually required there but that should give you very verbose usb debug output
11:02noopwafeldid anyone work out what apcie=0xfffffffe does
11:03svenno idea, that commandline has just been growing from when we did the initial bringup years ago
11:03noopwafelthe one relevant mention in irc logs is someone in 2022 asking "what does the "apcie=0xfffffffe" boot arg exactly do?"
11:03mazarryn: in my modest experience, code is far less ambiguous then English. but if you think this is useful, go forth and write things up!
11:03noopwafelfor any future irc log viewers: I dropped it a while ago and have not seemed to have any issues :)
11:04noopwafelfor docs, a bit more of a high level viewpoint (how do all the higher-level pieces fit together) would have been useful to me
11:05hdbngr has joined #asahi-dev
11:05arrynmaz: That's true BUT! it (often) requires the knowledge of how these kinds of devices work on a deeper level to begin with. That's the knowledge I'd like to gain here
11:06arrynnoopwafel: yeah something like that. So far my experience with drivers is very limited and mostly things that QEMU emulates
11:06pb17 has quit [Ping timeout: 480 seconds]
11:07arrynAnd since QEMU is often very lenient with mistakes my understanding is a little loose in my opinion
11:08arrynAlso learning how to use the m1n1 tooling is also a little complex for people coming in the first time. But I have to say once you figure it out it's a very convenient experience
11:09arrynThe first time I put my own code on it I was messing with a usb stick and uboot commands because I couldn't figure out how to load a raw binary over m1n1 hhh
11:11noopwafeloh, nice, not only tunable info from AppleT8122TypeCPhy but also the other register writes.
11:18sventhat's how all the #defines in atcphy come from
11:19sven*where
11:20hdbngr has quit [Ping timeout: 480 seconds]
11:25hdbngr has joined #asahi-dev
11:31integralpilot has joined #asahi-dev
11:35hdbngr has quit [Remote host closed the connection]
11:35hdbngr has joined #asahi-dev
11:37integralpilotwould support be accepted in mainline m1n1 for booting a zircon/fuchsia kernel? i've been experimenting with it a bit recently for fun
11:38chadmedwhat does it entail
11:43integralpilotusing their BSD licenced boot image library, https://fuchsia.googlesource.com/fuchsia/+/5f5e2c49e15a/src/firmware/lib/zbi/ (simple enough to roll our own if _really_ needed)
11:44hdbngr has quit [Ping timeout: 480 seconds]
11:46chadmedpersonally id rather not roll in any google code handed down to us from up on high only by the good graces of their benevolent product managers
11:46chadmedlast time we did that we ended up with sommelier and that did not go very well imho
11:46chaos_princesscan't it boot from efi?
11:46integralpilotthey have a U-Boot downstream fork they maintain that supports booting it
11:47chadmedlmao how preposterous
11:48chaos_princessi'd suggest using that then, as m1n1 is perfectly capable of loading u-boot already
11:48chadmedmy concern here is that they just one day decide to change how all of this works and pull the rug out
11:49chadmedjust chainload their u-boot fork
11:49chadmed^ yeah that
11:50integralpilotit's what i've been doing so far, guess i'll keep with that.
11:51integralpilothaven't actually tried loading the kernel yet, am trying to add AIC3 support rn
11:51integralpilotfuchsia is interesting in that UART, IC and timer are all kernelspace, every other driver is userspace
11:52NickChan[m]so like a bit of microkernel
11:52Guest1669 has quit [Remote host closed the connection]
11:52Rock has joined #asahi-dev
11:52chadmedzircon is nominally a microkernel yeah
11:53Rock has quit [Remote host closed the connection]
12:03pb17 has joined #asahi-dev
12:06chaos_princessoh, so it has terrible perf, got it (/j)
12:08Rock has joined #asahi-dev
12:09Rock is now known as Guest1687
12:09svenwe shouldn't add special support for any OS, just load to u-boot and take it from there
12:12Guest1687 has quit [Remote host closed the connection]
12:12freetolga has joined #asahi-dev
12:23integralpilot has quit [Quit: integralpilot]
12:26Sai_ has joined #asahi-dev
12:26hdbngr has joined #asahi-dev
12:28Sai_ has quit [Remote host closed the connection]
12:38hdbngr has quit [Ping timeout: 480 seconds]
12:39hdbngr has joined #asahi-dev
12:42Sai_ has joined #asahi-dev
12:44Sai_ has quit [Remote host closed the connection]
12:57hdbngr has quit [Ping timeout: 480 seconds]
12:59Sai_ has joined #asahi-dev
13:03Sai_ has quit [Remote host closed the connection]
13:06hdbngr has joined #asahi-dev
13:17Rock has joined #asahi-dev
13:18hdbngr has quit [Remote host closed the connection]
13:18Rock is now known as Guest1693
13:18Guest1693 has quit [Remote host closed the connection]
13:18hdbngr has joined #asahi-dev
13:21steven has quit []
13:22steven has joined #asahi-dev
13:27pb17 has quit [Ping timeout: 480 seconds]
13:34Sai_ has joined #asahi-dev
13:35Sai_ has quit [Remote host closed the connection]
13:35hdbngr has quit [Ping timeout: 480 seconds]
13:53pb17 has joined #asahi-dev
13:58hdbngr has joined #asahi-dev
13:59oliver has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
14:02oliver has joined #asahi-dev
14:07hdbngr has quit [Ping timeout: 480 seconds]
14:22noopwafelsven: do you expect to see pipehandler debug in there too btw?
14:23svenshould be in there i think
14:25sventhat commandline should literally enable every single debug thing i was able to find related to usb :D
14:25Rock has joined #asahi-dev
14:26Rock is now known as Guest1700
14:26noopwafel(I don't see pipehandler mmio either)
14:26Guest1700 has quit [Remote host closed the connection]
14:28hdbngr has joined #asahi-dev
14:36noopwafelah no I'm doing a silly thing of course. don't mind me!
14:36noopwafel(typoed the T8122 variant..)
14:36hdbngr has quit [Ping timeout: 480 seconds]
14:37Sai_ has joined #asahi-dev
14:39Sai_ has quit [Remote host closed the connection]
14:42sveniirc it really should just be different tunables
14:53Sai_ has joined #asahi-dev
14:55Sai_ has quit [Remote host closed the connection]
14:58noopwafelbtw, we ignore the usb-drd tunables..? the example in tunables.h is /arm-io/usb-drd0 - not needed?
15:06noopwafel(ah I guess 'yes' based on irc logs)
15:10Sai_ has joined #asahi-dev
15:12Sai_ has quit [Remote host closed the connection]
15:13Sai_ has joined #asahi-dev
15:17svenyes, one of them is e.g. just "set this documented dwc3 register to host mode" which the driver does anyway
15:18sveni've tried applying them to see if they help with the 5s delay for usb3 devices but they didn't change anything
15:21hdbngr has joined #asahi-dev
15:21Sai_ has quit [Remote host closed the connection]
15:21john-cabaj has joined #asahi-dev
15:37mripard has quit [Quit: WeeChat 4.8.1]
15:50hdbngr has quit [Ping timeout: 480 seconds]
15:52Rock has joined #asahi-dev
15:53Rock is now known as Guest1706
15:53Guest1706 has quit [Remote host closed the connection]
15:58hdbngr has joined #asahi-dev
16:02freetolgahow can i capture logs on macos lol
16:02freetolgai saw some debug flags in the irc logs
16:02freetolgawhich drivers do you want me to show debug output for right now
16:03freetolgaalright so unless me or someone else with an m4 has another mac and a bunch of spare time its never going to happen
16:04sveni'm confused, what is never going to happen?
16:04freetolgaanything usable on any m4 device
16:04svensomeone with reverse engineering knowledge needs to work on that, yes
16:04svengetting debug logs is not the bottleneck there
16:06freetolgaah ok so its simillar to when m1 was new
16:06freetolgasome big shift happened and we need a developer who wants to lock in for a while
16:06svenplease keep this discussion to #asahi since it's not related to development
16:09Sai_ has joined #asahi-dev
16:09Sai_ has quit [Remote host closed the connection]
16:11hdbngr has quit [Ping timeout: 480 seconds]
16:15deanrock097 has quit [Quit: The Lounge - https://thelounge.chat]
16:16deanrock097 has joined #asahi-dev
16:19hdbngr has joined #asahi-dev
16:25Sai_ has joined #asahi-dev
16:26nickchani did find a solution for t8015 shutdown compatible issue looking at other drivers all i need to do is declare another array of mfd_cell for t8015
16:26nickchanthough the more i look into it the more cursed it gets
16:26nickchannow i am in the power off simulation behavior stage (physically cannot be off when plugged in) where China devices *should* behave differently than other devices
16:26Sai_ has quit [Remote host closed the connection]
16:27nickchanwhich is whether offline charging should have a battery indicator (like android) or not
16:27nickchanChina devices should have a battery indicator in offline charging and other devices do not
16:27nickchanand this is determined by a value written into pmu nvram on shutdown
16:28hdbngr has quit [Ping timeout: 480 seconds]
16:28nickchanso basically now i need to get the region code into FDT to handle things 100% correctly which is even more complexity
16:28lynndotpy7 has quit [Quit: bye bye]
16:29lynndotpy7 has joined #asahi-dev
16:29lynndotpy7 has quit []
16:30lynndotpy7 has joined #asahi-dev
16:35hdbngr has joined #asahi-dev
16:45Sai_ has joined #asahi-dev
16:45Sai_ has quit [Remote host closed the connection]
16:48hdbngr has quit [Ping timeout: 480 seconds]
16:53Sai_ has joined #asahi-dev
16:56noopwafelhmmm, okay, after realizing _I need XHCI in my kernel_, usb works.
16:57hdbngr has joined #asahi-dev
16:59Sai_ has quit [Remote host closed the connection]
17:00noopwafelusb3 BIST dance doesn't succeed..
17:01alicela1n has joined #asahi-dev
17:03svenit won't work with incorrect tunables iirc
17:03svenor maybe something else changed around there
17:03svensometimes usb3 still works even if that thing fails
17:03sveni suspect it's some kind of bug workaround anyway
17:09WindowPain has quit [Quit: Ping timeout (120 seconds)]
17:09hdbngr has quit [Ping timeout: 480 seconds]
17:09WindowPain has joined #asahi-dev
17:11gargantua has joined #asahi-dev
17:13seb_ has joined #asahi-dev
17:14Rock has joined #asahi-dev
17:14Rock is now known as Guest1712
17:14Guest1712 has quit [Remote host closed the connection]
17:17hdbngr has joined #asahi-dev
17:25hdbngr has quit [Ping timeout: 480 seconds]
17:30Sai_ has joined #asahi-dev
17:31Sai_ has quit [Remote host closed the connection]
17:32noopwafelmildsunrise: thoughts re: upstreaming your spmi work? (aka "do you mind if rework your series")
17:34noopwafelsven: the dance is there in the trace but looks quite different.. but yes, I should double-check I got the tunables right first
17:35svenhow exactly does it fail?
17:35svenat the pipehandler lock? or while waiting for some other bit
17:35hdbngr has joined #asahi-dev
17:47Sai_ has joined #asahi-dev
17:50Sai_ has quit [Remote host closed the connection]
18:26noopwafelsven: ACIOPHY_TOP_PHY_STAT_LN0_UNK0
18:26noopwafelbut I should double-check things before you waste time :)
18:26sventhat's usually not caused by tunables fwiw
18:27svenbut who knows with m3
18:44pb17 has quit [Ping timeout: 480 seconds]
18:47Meleody has joined #asahi-dev
18:48Meleodyhi, just a quick question, is merging neural engine support still planned?
18:50alicela1n has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
18:52MeleodyI guess I'll just wait for the next force push to merge and build myself
18:56Meleodybut it would be nice to see this merged so that other software can start building around it.
18:57knorkle has quit [Quit: knorkle out]
18:58knorkle has joined #asahi-dev
18:58pb17 has joined #asahi-dev
18:58sveni don't think anyone has worked on the neural engine in quite a while
18:59Meleodytrue, but I guess it works and I would like to use it, I see no downside in merging it.
19:00svenwe'd be adding unmaintained code to our downstream tree which is precisely what we've been working against in the past months
19:00svenif someone is interested in taking over, rebasing on the latest -wip and working on making it upstreaming ready that would be great though
19:05chaos_princessMeleody: are you willing to write the userspace bits?
19:07MeleodyI'm no kernel dev. I like the word userspace however
19:07Meleodyand I was planning on using it on a regular basis myself for running a phase vocoder for singing speech synthesis.
19:08chaos_princessare you aware of the amount of work that would take?
19:08Meleodyhonestly? no
19:09chaos_princessabout as much as a gpu driver
19:10Meleodyare you sure? the pr says the driver is pretty simple
19:10chaos_princessthat is for the kernel bits
19:14chaos_princessif you want to see an example, look at the "rocket" driver in mesa
19:20hdbngr has quit [Ping timeout: 480 seconds]
19:21noopwafelsven: https://noopwafel.net/asahi/bist_trace_tmp.txt - some of the bit offsets are the same, but - new dance?
19:21jannausomeone was interested in looking into the ane driver not that long ago. not sure if there anything came out of it
19:23svennoopwafel: looks like it changed from a quick glance
19:28hdbngr has joined #asahi-dev
19:31noopwafeloh I messed up the correlations in there. but thx :)
19:36hdbngr has quit [Ping timeout: 480 seconds]
19:39deanrock097 has quit [Quit: The Lounge - https://thelounge.chat]
19:40deanrock097 has joined #asahi-dev
19:44hdbngr has joined #asahi-dev
19:46MeleodyOne last question regarding the AME: what counts as userspace? Implementing a whole ML Framework is a lot of work for very little gain, but never upstreaming the simple parts allowing basic access is just a shame imo.
19:46deanrock097 has quit [Quit: The Lounge - https://thelounge.chat]
19:48deanrock097 has joined #asahi-dev
19:51chaos_princessuserspace = "Implementing a whole ML Framework"
19:52hdbngr has quit [Ping timeout: 480 seconds]
19:57hdbngr has joined #asahi-dev
19:57MeleodyI just think being pragmatic and sticking to the scope of what the ANE Driver provides atm would be benficial imo.
19:58hdbngr has quit []
19:58MeleodyAnd I don't mean to say that it should be unmaintained, it still needs work for M2 and stuff.
20:00chaos_princesswhat does it provide atm?
20:03Meleodypractical stuff? https://github.com/eiln/ane-ex & https://github.com/ggml-org/whisper.cpp/pull/1021. Technically? A way to run hardware microcode for ML Models generated on Mac OS
20:04svenit's simple really: if you want to get this into the tree work on getting it upstream. it'll end up in our tree soon after it hits -next
20:04chaos_princessjfyi, with my experimentation, on a 1-hr audio file, using ane saves like a second or two compared to gpu only
20:04alicela1n has joined #asahi-dev
20:05chaos_princessas ane is only capable of running the encoder part of the whisper model, while the runtime is dominated by the decoder
20:07yanovskyy has joined #asahi-dev
20:07Meleodytbf the speedup depends on the application. I'll watch the linux kernel space more closely and see how it could be upstreamed.
20:07svenfrom what i remember the current driver has quite a bit of hacks, especially regarding the iommu handling
20:07chaos_princessyour main issue is that there is no ecosystem to use those things on linux
20:08svenyeah, and that
20:08svenand if this follow the gpu rules you need an open-source userspace with a stable UABI before the kernel module will be accepted
20:11MeleodyI can work without an ecosystem with the tools that are there. Also the Neural Engine isn't a gpu so why should it follow gpu rules just cause it does matrix stuff?
20:12chaos_princessbecause it is in drm/accel
20:12MeleodyAnother thing to consider is SME on the M4. But I guess it's gonna be a long road until then.
20:12chaos_princessit is a thingy that does dma and computation based on a program you load into it
20:12chaos_princesssounds pretty gpu-like :P
20:13chaos_princessthere is most likely no need to consider sme on m4, as we will get that for free from generic arm64
20:13alicela1n has quit [Ping timeout: 480 seconds]
20:16yanovskyy has quit [Remote host closed the connection]
20:16jannauand open user user space is not just the program which loads pre-compiled microcode but the compiler
20:17MeleodyFine, no Neural Engine for me ig :(
20:18chaos_princessjust use the gpu?
20:20MeleodyVulkan and OpenGL support rn, correct?
20:20jannauthe ane driver programs DART directly instead of using linux' DMA API
20:21chaos_princessopencl is there, may be more appropriate
20:21jannauand OpenCL. the GPU has a mayor quirk though. There is no preemption for compute jobs
20:26Meleodylooks like the choice is between pytorch with opencl or vulkan
20:31nicolas17 has joined #asahi-dev
20:33Meleodyboth of which I'll have to compile from source
20:46Meleody has quit [Quit: leaving]
20:50CME has quit [Ping timeout: 480 seconds]
20:51CME has joined #asahi-dev
21:00seb_ has quit [Quit: Konversation terminated!]
21:00seb_ has joined #asahi-dev
21:22hdbngr has joined #asahi-dev
21:35seb_ has quit [Quit: Konversation terminated!]
21:40noopwafelhm, wrote new BIST dance and .. it also fails waiting for a status register :)
21:44svenlooks like the usb curse got to you now@
21:54jannausven: argh, I can't read. good catch
21:54svensaw it earlier today and was confused for a moment as well because i though i had already fixed it :/
21:55sven*thought
21:56jannauI was confused by the commit date in the message and thought this was an old report. I checked that fixup was only added after that
22:01Rock has joined #asahi-dev
22:02Rock is now known as Guest1735
22:02Guest1735 has quit [Remote host closed the connection]
22:08pb17 has quit [Remote host closed the connection]
22:09pb17 has joined #asahi-dev
22:17Sai_ has joined #asahi-dev
22:20Sai_ has quit [Remote host closed the connection]
22:24hdbngr has quit [Quit: Lost terminal]
22:41mesa has quit [Read error: Connection reset by peer]
22:41mesa has joined #asahi-dev
22:50Sai_ has joined #asahi-dev
22:53Sai_ has quit [Remote host closed the connection]
22:58alicela1n has joined #asahi-dev
23:04crabbedhaloablut has quit []
23:07crabbedhaloablut has joined #asahi-dev
23:08Rock has joined #asahi-dev
23:08Rock is now known as Guest1740
23:09john-cabaj has quit [Ping timeout: 480 seconds]
23:12Guest1740 has quit [Remote host closed the connection]
23:25Sai_ has joined #asahi-dev
23:26Sai_ has quit [Remote host closed the connection]
23:32Sai_ has joined #asahi-dev
23:35Sai_ has quit [Remote host closed the connection]
23:36amarioguy has quit [Remote host closed the connection]
23:44lukeleppan has quit [Quit: The Lounge - https://thelounge.chat]
23:46pb17 has quit [Ping timeout: 480 seconds]
23:47nnym has quit [Read error: No route to host]
23:47nnym has joined #asahi-dev
23:48Sai_ has joined #asahi-dev
23:49lukeleppan has joined #asahi-dev
23:51Sai_ has quit [Remote host closed the connection]
…