#
…

2021-02-01

00:00neunon has quit [Quit: ZNC 1.8.2 - https://znc.in]
00:00neunon has joined #asahi
00:18bsandro has joined #asahi
00:30raster has joined #asahi
00:54furkan has quit [Ping timeout: 258 seconds]
00:55furkan has joined #asahi
00:57inglor has joined #asahi
01:11alvaro_fe has quit [Ping timeout: 272 seconds]
01:12alvaro_fe has joined #asahi
01:19raster has quit [Quit: Gettin' stinky!]
01:44affieuk[m] has joined #asahi
01:46bsandro has quit [Quit: leaving]
01:56davidrysk[m]is GXF an Apple-specific ISA feature?
02:01robinp has quit [Read error: Connection reset by peer]
02:04robinp has joined #asahi
03:01inglor has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
03:11inglor has joined #asahi
03:29PhilippvK_ has quit [Ping timeout: 260 seconds]
03:29phiologe has joined #asahi
04:05Tokamak has joined #asahi
04:13rafaelmartins has left #asahi ["https://rgm.io/"]
04:13rafaelmartins has joined #asahi
04:30marvin24 has quit [Ping timeout: 260 seconds]
04:30marvin24 has joined #asahi
04:43modwizcodeAnnnd I built a second kernel from the same sources and that one won't boot (or at least print out) anything at all... very strange.
04:44modwizcodeApple breaks ARM spec it seems (if you trust the comment in qemu). Some of their extra registers use bits that are supposed to essentially form an access mask and it would require EL3 access but of course that's not the case.
04:45modwizcodeI have to patch out that access check so that it doesn't complain when I add the cp registers.
04:54Tokamak_ has joined #asahi
04:57Tokamak has quit [Ping timeout: 264 seconds]
05:01Tokamak has joined #asahi
05:02Tokamak_ has quit [Ping timeout: 246 seconds]
05:36marvin24 has quit [Ping timeout: 240 seconds]
05:37marvin24 has joined #asahi
05:57x4m3 has quit [Ping timeout: 240 seconds]
06:19Tokamak has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
06:29VinDuv has joined #asahi
06:36ephe_meral has joined #asahi
06:44modmuss50 has quit [Ping timeout: 260 seconds]
06:44brinly has quit [Ping timeout: 260 seconds]
06:44modmuss50 has joined #asahi
06:44brinly has joined #asahi
06:45robher has quit [Ping timeout: 260 seconds]
06:45diddledan has quit [Ping timeout: 260 seconds]
06:46narmstrong has quit [Ping timeout: 260 seconds]
06:46robher has joined #asahi
06:47narmstrong has joined #asahi
06:47diddledan has joined #asahi
06:57mndza has joined #asahi
07:26VinDuv has quit [Quit: Leaving.]
07:35ephe_meral has quit [Ping timeout: 240 seconds]
07:49ephe_meral has joined #asahi
07:57_whitelogger has joined #asahi
08:24ephe_meral has quit [Ping timeout: 264 seconds]
08:25phiologemodwizcode: I am about to try out your qemu patch soon but my crappy 7 years old MBP compiles soooooo slow.
08:30furkan has quit [Ping timeout: 272 seconds]
08:30chuskywalker has joined #asahi
08:41irl25519 has joined #asahi
08:44furkan has joined #asahi
08:44acelogic has quit [Remote host closed the connection]
08:45acelogic has joined #asahi
08:48irl25519 has quit [Quit: irl25519]
09:23dstzd has quit [Quit: ZNC - https://znc.in]
09:36acelogic has quit [Ping timeout: 240 seconds]
09:44a3541 has joined #asahi
09:47ephe_meral has joined #asahi
09:47never_releaseddavidrysk[m]: Yes
09:48never_releasedmodwizcode: registers that require EL3 are accessible from EL2 when EL3 isn't implemented
09:48never_releasedyou do not have to implement EL3 or EL2 in the Arm ARM
10:03marcannever_released: adhoc signed binaries now work in 1TR in the latest 11.2
10:03marcanyou were saying about apple locking 1TR down? :)
10:05marcanShiz: this is your cue to start poking more at Apple boot policy related APIs if you're up to it; if we can put together an end to end installer from 1TR this is definitely the way to go
10:11never_releasedmarcan: nice!
10:11never_releasedthis solves the problem
10:12never_releasedmarcan: remember, the BootPolicy APIs have entitlements required to access them
10:12never_releasedcheck if you have those (you probably don't)
10:13marcangood point, need to look into that
10:16thestr4ng3r has joined #asahi
10:23never_releasedmarcan: RIP about the kernel libs
10:23never_released> LDFLAGS_KERNEL_ONLY += -rdynamic -Wl,-force_load,$(KDKROOT)/System/Library/KernelSupport/lib$(CURRENT_MACHINE_CONFIG).os.$(CURRENT_KERNEL_CONFIG).a
10:23never_releasedthey moved them to the KDK
10:23never_releasedenjoy the single release with them being APSL
10:23amw has quit [Ping timeout: 264 seconds]
10:24marcanwell, beats nothing
10:24marcanis this a new release?
10:25never_releasednot visible in directory view
10:25never_releasedwas done in a sneaky way
10:27marcanI mean, at least this makes *some* sense
10:27marcanit's stupid, but at least shows *intent*
10:28marcanalso xnu-7195.60.75/doc/pac.md disappeared
10:28marcanbecause of course documentation is bad :)
10:28j`eythese releases are so weird, I wonder if they make sense internally
10:28marcanlooks like they also added __APRR_SUPPORTED__ to the censor filter
10:29marcanall those blocks are gone now
10:29never_releasedthink that's temporary
10:30marcan__APSTS_SUPPORTED__ too
10:30marcanI get the feeling they're trying to censor out all the nonstandard CPU features
10:30marcanit smells like ARM is involved somehow
10:31never_releasedI can bet that it's because of Arm, not AAPL
10:32never_releasedmaybe the Raspberry Pi target is to ensure that it always works on conventional Arm too
10:39marcanI wonder
10:46marcanlol, I tried messing with the bputil codesignature, but it's built against arm64e which is apparently limited to apple-signed stuff, so that flat out breaks everything anyway (kernel refuses to load it)
10:51furkan has quit [Ping timeout: 240 seconds]
10:52never_releasedmarcan: patch the ISA to arm64
10:52never_releasedinstead of arm64e
10:52never_releasedarm64e = PAC on, you don't need PAC on here :P
10:54raster has joined #asahi
10:58CDFH has quit [Quit: That's it, I quit]
11:00CDFH has joined #asahi
11:07marcanand yeah, bputil doesn't work without the entitlements
11:07marcanso the next question is how much can we do by shelling out to existing Apple binaries
11:07x56 has quit [Quit: Ծ-Ծ]
11:07mndza has quit [Read error: Connection reset by peer]
11:07marcanspecifically, installation-related stuff
11:08marcanonce they properly support booting/importing external macOS installs from external HDDs we should be fine
11:08mndza has joined #asahi
11:09x56 has joined #asahi
11:22never_releasedmarcan: the question is how they organise that
11:22never_releasedare the user accounts staged to SEP in that case on first boot
11:22never_releasedor on blessing
11:26ephe_meral has quit [Ping timeout: 264 seconds]
11:27marcannever_released: should be on blessing, otherwise that sounds like potentially problematic (boot an external OS and let it adopt users itself? dodgy...)
11:28marcanI'm going to do some tests with external storage
11:28marcanbut I think this doesn't work yet :)
11:32a3541 has quit [Ping timeout: 240 seconds]
11:38never_releasedmarcan: boot from external storage works already, but does it work when you bless an install from elsewhere today?
11:38marcanthat's what I mean, blessing an install from another machine
11:38marcanwhich is what we care about
11:39never_releasedI don't see why it wouldn't work
12:07furkan has joined #asahi
12:56ephe_meral has joined #asahi
12:59irl25519 has joined #asahi
12:59irl25519 has quit [Remote host closed the connection]
13:24Shizmarcan: sure
13:30marcanShiz: given the entitlements stuff it's not clear we can use it anyway, but I want to know how it works :)
13:51BaughnLogBot has quit [Ping timeout: 260 seconds]
13:58BaughnLogBot has joined #asahi
13:58Namidairo has quit [Ping timeout: 240 seconds]
14:12davidrysk[m]just so you all know, Apple has been known to yank improperly redacted source releases in the past
14:17Shizget it while it's hot etc
14:19ephe_meral has quit [Ping timeout: 265 seconds]
14:23acelogic has joined #asahi
14:32davidrysk[m]I'm concerned that we will run into the following issues: (1) Apple's performance relies heavily on these "Apple-isms"; (2) Overall security relies heavily on these "Apple-isms", e.g. without using them we are open to Spectre-like attacks
14:33marcanI doubt it
14:33marcane.g. it seems they have *specific* Spectre mitigations but Linux already has generic ones
14:33modwizcodespeaKing of spectre like attacks, ARM has some interesting barriers to just block speculation in the ISA. They seem real difficult to use correctly though
14:33marcanre (1), for userspace software there is no magic, plain ARM compute is plain ARM compute
14:34davidrysk[m]I'm thinking in the direction of how expensive context switches might be
14:34marcanI doubt there's some kind of massive gain to be had there by using special sauce
14:34marcanI mean I'm sure they have "little things"
14:35marcane.g. optimizations for objc use cases
14:35marcanbut it's not going to make linux terrible if we don't use them
14:35marcanthe CPUs *are* good
14:35modwizcodeIt would be neat to figure out those features though.
14:36davidrysk[m]btw they haven't yet released the kdk for xnu-7195.81.3
14:36davidrysk[m]I'm surprised they posted the tarball at all (even with hidden url)
14:36marcanmodwizcode: certainly
14:36marcanbut some things we also just won't be able to use out of practicality
14:36marcane.g. anything that requires recompiling all userland
14:37davidrysk[m]what kind of stuff would require recompiling all userland?
14:37seghernot using x86 anymore
14:37marcanwell, integrating custom instructions deeply, if we actually found something useful
14:37davidrysk[m]AMX2 is the main one that would be useful, and even there, Apple only uses it in their optimized blas/lapack implementations
14:38jixgiven that even apple's public toolchains don't support any of that AFAICT... I doubt there would be a need for that
14:38davidrysk[m]memory compression... probably would be something added to a variant of the zram driver
14:38modwizcodeI find it interesting how many different varients of matrix and vector enhancements to instruction sets have been made
14:39seghermarcan: and "deeply" sounds lessmlikely than it is... if you simply have some new insn that compilers can use usefully, you already get close to 100% of programs using it
14:39davidrysk[m]the only thing I can think of that comes close to recompiling userland would be PAC, but Linux has an implementation of the general/standard bits of PAC, and I am hoping (though not certain) that the Apple-specific stuff can be avoided or used entirely in kernel space
14:44jixsegher: the REed AMX docs (linked in #asahi-re) mention that moving data between AMX and the main processor is very slow. To me it seem that would make automated compiler use (or even using it at all in most general purpose code) quite difficult
14:46segheri was answering the general question :-)
14:46diddledanI'll be curious when we get that far whether the 2 screen limitation is enforced via macOS or if it is a physical limitation of the m1 chip
14:46segher"what kind of stuff would require recompiling all userland?"
14:47segheranything that uses any new instruction will not run on older hardware
14:47jixsegher: sure, I was just hoping you'd tell me that I'm totally wrong in something like that being difficult for a compiler to use and I'd learn something ;)
14:48marcandiddledan: almost certainly a physical limitation
14:48segheri suppose there *is* a way to trap and emulate those on any arm implementation, it is so very useful, but who knows
14:48marcannumber of display controllers is a hardwre thing
14:48segherjix: i don't really know what AMX is :-)
14:48diddledanthat is my only real downside of my MBA :-)
14:49j`eydiddledan: I cant even get my 1 monitor to work D:
14:51jixsegher: neither do I, but I was thinking of a set of vector/matrix instructions that are encoded inline like normal instructions, but execute on a separate core with its own cache etc (that's the impression I get from the REed docs but I might be totally wrong)
14:51marcanfwiw, apparently the internal devicename for the M1 is just H13G
14:51marcanwhich means A14X
14:51marcani.e. lineage-wise "it's just a tablet cpu"
14:51modwizcodetwo code names is the best code names
14:51marcanwhich explains some of the limitations
14:52modwizcodeThe M1 has a lot of aspects that seemed a bit rushed I think?
14:52marcanI'm guessing what Apple did was decide to beef up their latest tablet silicon enough they could stuff it in Macs and look good
14:52marcanbut it's not really a well-rounded Mac chip yet
14:53marcanhence the 16GB limitation, two thunderbolt ports, two displays, etc
14:53diddledanI wonder if the future chips will fix some of the oddities that we've encountered so far, like nGnRE being a black hole
14:53j`eymarcan: which makes the next one seem even more exciting
14:53modwizcodeI saw something that made me wonder if the DTK was just basically a repurposed iPad CPU
14:54marcanthat is well known
14:54modwizcodeoh I didn't realize that was known
14:54marcan"A12X" they called it...
14:54marcanA12X was the iPad chip
14:54krbtgtfor the low end macs it's fine
14:54marcanboth are H11G/Aruba internally
14:54krbtgtit's a tablet CPU but a really good one
14:54davidrysk[m]A12Z was the iPad chip for the early-2020 iPad
14:55davidrysk[m]A12Z is A12X with the binned GPU core unlocked and I believe slightly higher clocks
14:55krbtgti'm curious what desktop (or even higher wattage mobile) oriented silicon will look like
14:55modwizcodemarcan: Any ideas why I might get a decrompession error trying to load the kernel? I wanted to ask what your config was because my first one booted and got page faults and a second one with only a different config won't boot at all
14:55marcanah right, both were iPad chips
14:55krbtgta12z is basically just binning
14:55marcanmodwizcode: corrupted data or some kind of allocation issue?
14:55modwizcodeI guess that's more of a -dev topic anyway
14:56modwizcodeAll I can think of is that this config is less lightweight so it's probably using more memory
14:56modwizcodebut I changed my memory layout so it can't overlay anythin
14:57j`eymodwizcode: what was the config changes between the two?
14:57segherjix: ah, their convolution stuff
14:57segher"for ai"
14:57modwizcode j`ey: nothing meaningful sadly I did a diff
14:58modwizcodeI forgot to do defconfig on the first one before reconfiguring
14:59jixsegher: I think AMX is what they use for their BLAS implementation... not sure if it is the same as their "AI" stuff, but I guess if not it's probably similar in that respect
15:00modwizcodeIt's different according to the people reversing it. That AI stuff is it's own accelerator
15:05marcanyeah, the neural engine is another thing
15:19arndsegher: https://gist.github.com/dougallj/7a75a3be1ec69ca550e7c36dc75e0d6f, TL;DR essentially a vector FMA unit to speed up 8x8 DGEMM or 16x16 SGEMM matrix operations
15:21arndI tried to get it working inside of a linux guest running in qemu/hvf, but the instruction to enable it gives invalid-opcode in a VM
15:23segherjix: it is SP float only, so useless for almost all BLAS stuff
15:23jixsegher: doesn't look like it is SP float only
15:29segheralso brainfloat, and maybe some HP
15:35davidrysk[m]then how are they using it in their blas and related libs?
15:36zkrx has quit [Ping timeout: 246 seconds]
15:38segherdavidrysk: i do not know. i simply say that afaik the apple thing does not do DP float
15:38seghermaybe it does though... that would be quite surprising
15:39segher(needs about six times as much hardware, and has double the latency)
15:42jixsegher: are we looking at the same thing? (what arnd linked)
15:43arndIt looks like ANE (apple neural engine) is bfloat16 and float32, while AMX is int16/float32/float64
15:44arndANE seems to be its own processor core with shared memory, while AMX is the instruction set extension for the CPU itself and operates on vector registers
15:50jixbut from that description it also seems that AMX does have a separate load/store path (with loading stuff into AMX that is in main core L1 cache being slower than when it isn't) which was the reason I brought it up in the first place regarding potential difficulties in compilers targeting those instructions
15:52arndjix: do you have any idea how many AMX units there are? I have this suspicion that it's less than one per core, possibly only one of them in total, which would explain why it's slow to copy through L1 cache, and why it needs an explicit enable/disable instruction
15:52zkrx has joined #asahi
15:53jixarnd: nope, the only thing I did re AMX is having a quick look at the doc you linked
15:53jixbut yeah I've been wondering the same thing
15:54jixI think the enable/disable might be more to make context switches cheaper if the AMX state isn't used though
15:54arnd${coworker} who has used AMX in the past says they only run single-threaded applications
15:55arndthe context switch is definitely part of it, saving >4KB of state is expensive
15:55jixeven if there are fewer AMX cores than cpu cores, I'd expect them to have enough registers to store state for all cores... it would seem really awkward to ensure that at most one core is actually running an AMX using thread
15:58x4m3 has joined #asahi
16:11segherjix: no, i cannot keep up with the backscroll here... i didn't see arnd even! hi arnd!
16:12Tokamak has joined #asahi
16:13segherso ah, one TLA and another
16:15irl25519 has joined #asahi
16:22daniel[m]7 has joined #asahi
16:24segherjix, arnd: so AMX is for training neural nets, and ANE is for running them, it seems
16:25arndyes, pretty much, though ${coworker} uses it for running neural nets, as they are dealing with relatively small matrices
16:27arndthey got a 3x improvement just by changing from blis to Apple's blas replacement
16:28irl25519 has quit [Quit: irl25519]
16:28arndwith some hope of getting even more out of it using hand-written code specific to the application.
16:28jsseidel2 has joined #asahi
16:31jsseidel2 has left #asahi ["WeeChat 1.9.1"]
16:44segherthe very nasty thing about such accelerators (like MMA on Power) is that they implicitly change operation order, so they only can work with -ffast-math or similar
16:44irl25519 has joined #asahi
16:45segherthe nasty thing is that many algos very much *depend* on operation order
16:45segherso your choice often is "extra fast code, or correct code"
16:50irl25519 has quit [Quit: irl25519]
16:52irl25519 has joined #asahi
16:53irl25519 has quit [Client Quit]
16:54davidrysk[m]arnd: fwiw scipy yanked out support for apple's blas replacement a while ago due to apple bugs that they refused to fix at the time
16:55davidrysk[m]Apple has since fixed said bugs
16:58davidrysk[m](or am I thinking of numpy, heh)
16:58arndthe testing was done with a locally patched numpy
17:03AnCaRioN has quit [Quit: The Lounge - https://thelounge.chat]
17:04VinDuv has joined #asahi
17:05AnCaRioN has joined #asahi
17:25davidrysk[m]arnd: I was going to do testing and maybe propose they reintegrate the features, but numpy isn't passing my tests on macOS ARM
17:25davidrysk[m]I can DM if you want
17:45irl25519 has joined #asahi
17:56irl25519 has quit [Quit: irl25519]
18:07irl25519 has joined #asahi
18:09Bastian[m] has quit [Ping timeout: 244 seconds]
18:09ewlsh[m] has quit [Ping timeout: 244 seconds]
18:09bakk[m] has quit [Ping timeout: 244 seconds]
18:10ldhacker[m] has quit [Ping timeout: 244 seconds]
18:10fridtjof[m] has quit [Ping timeout: 244 seconds]
18:10Eighth_Doctor has quit [Ping timeout: 244 seconds]
18:10Alice[m] has quit [Ping timeout: 244 seconds]
18:10ponikrf[m] has quit [Ping timeout: 244 seconds]
18:10wolf511[m] has quit [Ping timeout: 244 seconds]
18:10ronyrus[m] has quit [Ping timeout: 244 seconds]
18:10fried_dede[m] has quit [Ping timeout: 244 seconds]
18:10d_u_f_f[m] has quit [Ping timeout: 244 seconds]
18:10tr0[m] has quit [Ping timeout: 244 seconds]
18:10jinen[m] has quit [Ping timeout: 244 seconds]
18:10svenpeter has quit [Ping timeout: 244 seconds]
18:10coinquest[m] has quit [Ping timeout: 244 seconds]
18:11m1kr0[m] has quit [Ping timeout: 244 seconds]
18:12Bublik has quit [Read error: Connection reset by peer]
18:15Bublik has joined #asahi
18:20ldhacker[m] has joined #asahi
18:21coinquest[m] has joined #asahi
18:21tr0[m] has joined #asahi
18:21d_u_f_f[m] has joined #asahi
18:21ponikrf[m] has joined #asahi
18:21fried_dede[m] has joined #asahi
18:21Alice[m] has joined #asahi
18:21svenpeter has joined #asahi
18:22ronyrus[m] has joined #asahi
18:22jinen[m] has joined #asahi
18:22wolf511[m] has joined #asahi
18:22m1kr0[m] has joined #asahi
18:22ewlsh[m] has joined #asahi
18:22bakk[m] has joined #asahi
18:23Bastian[m] has joined #asahi
18:24fridtjof[m] has joined #asahi
18:25Eighth_Doctor has joined #asahi
18:44irl25519 has quit [Quit: irl25519]
18:53davidrysk[m]marcan: 11.2 is released
18:53davidrysk[m]same build as RC3
18:54davidrysk[m]still no KDK
19:05TheJollyRoger has quit [Remote host closed the connection]
19:06TheJollyRoger has joined #asahi
19:42gurmble has joined #asahi
19:45grumble has quit [Ping timeout: 606 seconds]
19:45gurmble is now known as grumble
19:45mndza has quit [Ping timeout: 240 seconds]
19:48irl25519 has joined #asahi
19:52irl25519 has quit [Client Quit]
19:52irl25519 has joined #asahi
19:53irl25519 has quit [Client Quit]
20:08the-mentor8 has joined #asahi
20:10thresh has quit [Quit: I'm a quit message virus. Please replace your old line with this line and help me take over the world]
20:10thresh has joined #asahi
20:10the-mentor has quit [Ping timeout: 260 seconds]
20:10the-mentor8 is now known as the-mentor
20:24loruu has joined #asahi
20:25zkrx has quit [Ping timeout: 264 seconds]
20:26ewlsh[m] has quit [*.net *.split]
20:26delroth[m] has quit [*.net *.split]
20:26botoxparty[m] has quit [*.net *.split]
20:26foxlet has quit [*.net *.split]
20:26rootspring[m] has quit [*.net *.split]
20:26reispflanze[m] has quit [*.net *.split]
20:26dsads[m] has quit [*.net *.split]
20:26ashton314[m] has quit [*.net *.split]
20:26brentr123[m] has quit [*.net *.split]
20:26lev[m]1 has quit [*.net *.split]
20:26smist08[m] has quit [*.net *.split]
20:26davidrysk[m] has quit [*.net *.split]
20:26skillfulman23[m] has quit [*.net *.split]
20:26dancer[m] has quit [*.net *.split]
20:26noneucat has quit [*.net *.split]
20:26bylaws has quit [*.net *.split]
20:26loru has quit [*.net *.split]
20:28Tokamak has quit [Ping timeout: 244 seconds]
20:31Tokamak has joined #asahi
20:32foxlet has joined #asahi
20:32bylaws has joined #asahi
20:32davidrysk[m] has joined #asahi
20:35rootspring[m] has joined #asahi
20:36ewlsh[m] has joined #asahi
20:36noneucat has joined #asahi
20:37dancer[m] has joined #asahi
20:37lev[m]1 has joined #asahi
20:37ashton314[m] has joined #asahi
20:37zkrx has joined #asahi
20:37dsads[m] has joined #asahi
20:37delroth[m] has joined #asahi
20:37brentr123[m] has joined #asahi
20:38botoxparty[m] has joined #asahi
20:38skillfulman23[m] has joined #asahi
20:38smist08[m] has joined #asahi
20:38reispflanze[m] has joined #asahi
20:39amw has joined #asahi
20:45djhworld has quit [Quit: The Lounge - https://thelounge.chat]
20:52djhworld has joined #asahi
20:53djhworld has quit [Client Quit]
20:57djhworld has joined #asahi
21:00bsandro has joined #asahi
21:20the-mentor has quit [Ping timeout: 240 seconds]
21:28the-mentor has joined #asahi
21:37zkrx has quit [Ping timeout: 264 seconds]
21:42raster has quit [Quit: Gettin' stinky!]
21:45amw has quit [Ping timeout: 246 seconds]
21:51VinDuv has quit [Quit: Leaving.]
22:02zkrx has joined #asahi
22:11amw has joined #asahi
22:17chuskywalker has quit [Quit: Leaving]
23:03brentr123[m]so kind of off topic but still related m1 macs, anyone know why some installed applications just straight up don't have their notifications detected? for example with this app(element) the notifications are missing from system preferences, and the in app settings are on for it
23:05roxfan #asahi-offtopic
23:06roxfanor whatever the channel is for macs
23:08roxfanlooks like there's ##mac and #macosx
23:25irl25519 has joined #asahi
23:41choozy has joined #asahi
23:46irl25519 has quit [Quit: irl25519]
…