#
…

2021-09-15

02:07loki_val has joined #asahi
02:09crabbedhaloablut has quit [Ping timeout: 480 seconds]
02:46yuyichao_ has joined #asahi
02:48malvo has quit [Ping timeout: 480 seconds]
02:48malvo has joined #asahi
02:53phiologe has joined #asahi
02:56PhilippvK has quit [Ping timeout: 480 seconds]
03:23marvin24_ has joined #asahi
03:26marvin24 has quit [Ping timeout: 480 seconds]
05:27phireNames of Firestorm-next/Icestorm-next are apparnetly Avalanche and Blizzard
05:37nskl has joined #asahi
05:39nsklaus_ has quit [Ping timeout: 480 seconds]
06:00chadmedis that the name of the cores in the next A series chip?
06:26aleasto has joined #asahi
06:47chadmed has quit [Remote host closed the connection]
06:47chadmed has joined #asahi
07:18marcanand presumably M2
07:19marcanbut I expect M1X will still be icestorm/firestorm
07:25marcanstreaming a bit at https://youtu.be/kMSv_AiaH9w / #asahi-stream
08:06thunfisch has quit [Quit: frrrp!]
08:08thunfisch has joined #asahi
08:24akemin_dayo has quit [Ping timeout: 480 seconds]
08:31balrog has quit [Quit: Bye]
08:33balrog has joined #asahi
08:34borrrden has joined #asahi
09:12bgb has joined #asahi
09:15akemin_dayo has joined #asahi
09:39bgb has quit [Ping timeout: 480 seconds]
09:40j_eyalyssa's stream is in 3h https://www.youtube.com/watch?v=uTZISTjqy9Q so get your popcorn
09:40bgb has joined #asahi
09:45tom-w has joined #asahi
09:53bgb has quit [Ping timeout: 480 seconds]
10:02robinp has quit [Remote host closed the connection]
10:11tophevich[m]wow, double the streams double the fun :D
10:58chadmedperfect timing for some comfy bedtime wind down viewing :)
11:03phirealyssa: is your talk live or pre-recorded?
11:07alyssaphire: live because executive dysfunction
11:08chadmedrecording presentations sucks
11:08chadmedi had to record a research presentation today because the conference has been made virtual and it took me over 2 hours to get out a 10 minute narrated slideshow of acceptable quality
11:08chadmedjust evil stuff
11:14svengiven that i usually only finish the slides (or the demo) maybe a few hours before the presentation i'm kinda sure i'd never be able to record one in time
11:15chadmedyeah the main problem is that a good slideshow is never truly done, and so the effort of having it recorded and prepared early means that its an extra mountain of work to go back and make last minute changes
11:16chadmedlike already ive found problems with my recording but im not going to go back and fix it because it will take too much time and the audience is stuffy medical researchers who will probably not even notice
11:16chadmedbut it annoys *me* and thats what matters at the end of the day
11:43X-Scale has joined #asahi
11:45alyssachadmed: oh shoot you're reminding me I need to update my slides right now
11:46chadmedalyssa: glad i could be of *some* use :D
11:46mpaq has left #asahi [#asahi]
11:49X-Scale` has quit [Ping timeout: 480 seconds]
12:43aleasto has quit [Quit: Konversation terminated!]
12:43aleasto has joined #asahi
12:44aleasto has quit [Remote host closed the connection]
12:45aleasto has joined #asahi
12:45robinp has joined #asahi
13:05dnjmis[m]Don't worry alyssa, the drivers work we can hear you fine
13:22DarkShadow44 has quit [Quit: ZNC - https://znc.in]
13:29DarkShadow44 has joined #asahi
13:34j_eyalyssa: 👏👏👏
13:34borrrden has quit [Ping timeout: 480 seconds]
13:34psykosealyssa: great talk!
13:34awal3 has quit [Ping timeout: 480 seconds]
13:36aleasto has quit [Remote host closed the connection]
13:37fridtjof[m]alyssa: really enjoyed the talk, thank you!
13:46TheJollyRoger has quit [Read error: Connection reset by peer]
13:46TheJollyRoger has joined #asahi
13:47yuyichao_ has quit [Ping timeout: 480 seconds]
13:48aleasto has joined #asahi
13:53awal3 has joined #asahi
14:30svenhttps://github.com/AsahiLinux/linux/commits/sven/i2c-tipd-WIP this does usb hotplug in a very wrong and hacked way but it allows testing i2c (still WIP) and the tipd patches (mostly happy with those)
14:36j_eyooh, some power states thingy
14:36sventhat's not new fwiw. it's just that SPSS command for the TI chip
14:36sven(possibly "system power state switch")
14:36alyssasven: I do not envy trying to upstream that branch
14:37sventhree commits of that branch are already almost upstream
14:37svenand it'll be two series. one for i2c and another one for tipd
14:37j_eythey were reviewed-by, this morning
14:38j_eysven: can you try with pinctrl at some point, would be nice to know if there's any issues there
14:38svenand if you take a look at all those tipd commits you'll notice that they are small diffs
14:38svenj_ey: yup, i wanna see if i can do without that "irqs are overrated" commit
14:39alyssasven: 👍
14:39j_eysven: and about the power states, it sounds like the other chips do support SPSS, youre just saying 'false' to avoid changing other devices?
14:40sventhey support power states, i don't know if they support the SPSS command
14:40sveni'll write about that in cover letter
14:40svenif i can get someone to test with a regular SPSS chip that conditional can go away completely
14:40svenarg.. whatever. just get rid of the "SPSS" in that last message :D
14:41j_eyi understood :P
14:42j_eyand I like the use_int2/use_int2 thing
14:45svenyeah. i think the only behavior that's changed for regular TI chips is that the interrupts are explicitly enabled instead of just assuming they are
14:45svenand i can easily add another quirk if the maintainer would like to keep the original behavior there as well
14:52Andalu30 has joined #asahi
14:55marcanall the SMP stuff is pushed, FWIW
14:56alyssaOoh sane boot time thank g-d
14:57alyssamarcan: very nice
14:57marcanCPU freq scaling doesn't work, but apparently the pcores boot pretty fast, or macos puts them there? (just not max)
14:57marcanneed to figure out why it does that
14:57marcanit does work on bare metal
14:58marcancould be something I did wrong, or the code actively crippling itself when run as a guest
14:58alyssai've been meaning to try to understand the corellium cpufreq code
14:58marcanI'm not even sure it's, er, correct
14:58alyssa(it doesn't cherry-pick cleanly)
14:59alyssahas some reg names at least
14:59marcanyeah I know
14:59marcanI went through that in my stream at the beginning
14:59marcanbefore deciding to shave this yak
15:00j_eyremenants of the great clang-format.py "+ #TODO : turn into a real tracer"
15:06marcanj_ey: meh, not worth fixing :p
15:07j_eyfix it when resolving the TODO :P
15:40jbowen has joined #asahi
15:46tophevich[m]alyssa: Sadly I couldn't follow your talk live, thus only commenting now, but it was very informative and entertaining at the same time, nice job! Nice touch that you were able to go meta and show it from linux running on an m1 mac mini.
15:56TheJollyRoger has quit [Remote host closed the connection]
15:57aleasto has quit [Ping timeout: 480 seconds]
15:58TheJollyRoger has joined #asahi
16:19rowang077[m]Is the talk online somewhere to watch? I couldn't find it
16:20tophevich[m]it's the third talk i think
16:21rowang077[m]thx. It's 2 hours and 10 mins into the stream
16:33alyssatophevich[m]: thank you!
16:34alyssaoh nice! didn't realize the stream was available immediately
16:38tophevich[m]alyssa: this is because it is still streaming, and you can go rewind in youtube livestreams. But AFAIK they can remove the stream afterwards (and sadly that happens for conferences too often). Hopefully the keep the stream vod, and in addition publish a vod for each talk seperately.
16:40alyssasure, XDC publishes separate vods later. I guess the "all talks of the day" will be available immediately
16:40tophevich[m]I think we are in luck, I just found the ones from last year so they keep those aswell.
16:44aleasto has joined #asahi
16:47alyssayeah, got confirmation it'll be up tonight as a full video
16:48alyssaaside - I spent way too long on those slides 😋
16:48tophevich[m]well worth it (from the entertaining side of things)
16:49tophevich[m]Timesink where mostly the emojis I guess?
16:49alyssatophevich[m]: I mean
16:50alyssa...actually I made those before starting slides ;_p
16:50alyssaApple Memoji
16:50alyssaseemed topical
16:50alyssaonce I saw it had a witch hat I couldn't not
16:50tophevich[m]ahhh, clever
16:51alyssamost of the time was theming
16:51alyssaI'm used to using company templates as-is, this wasn't a company presentation though
16:51alyssaand didn't want to be Yet Another Beamer presentation
16:52jnah, the joy of presentation theming
16:52tophevich[m]yeah, they were very unique and fitting without any overload
16:53jn(side note: really cool to see people present *on* hardware, in a dual sense. both alyssa and Rob Clark did this today.)
16:54* jnmade a presentation on the Wii U's ARM core once
16:56svenoh... what was that presentation about?
16:56jnsaid ARM core, and generally how to run code on the Wii U
16:57jn(based on other people's work)
16:58jnthe slides were hand-drawn in gimp and RLE-compressed, the "next slide" button was the power button because GPIO is easy and USB is hard
16:59jnanyway, the point is: the topic was also the presentation device
16:59phirepoor USB. Everyone neglects it
16:59j_eynot sven >:D
16:59alyssajn: yo my m1 linux works better than my rk3399 linux the choice is obvious :_p
16:59jnalyssa: hehehe :)
17:00svenjn: was just curious because i did a lot of wii hacking and was also involved in the initial wiiu hack :)
17:04jnsven: my contributions to the wiiu scene are negligible. i wanted to run Linux on the PPC, but i gave up when i couldn't figure out how to start the PPC from the ARM, after using the kexec-like syscall in IOSu
17:07sveniirc you actually had to race the ppc to boot unsigned code. it's been a while.
17:14Andalu30 has quit [Ping timeout: 480 seconds]
17:24j_eymarcan: hv with smp works for me!
17:29arahael has quit [Ping timeout: 480 seconds]
17:36arahael has joined #asahi
17:38skali has joined #asahi
17:44j_eymarcan: while you were streaming I mentioned about turning the clocks on/off for the pinctrl, but after trying I dont think we can do it, since the interrupt controller is the same hw, so if we turn it off, we dont get interrrupts for the gpios
17:44roxfanpoll then?
17:45svenboth the gpio and the pinctrl driver can call clk_enable/disable independently
17:45sventhe clk framework does the required reference counting
17:46j_eyI need to look at the rockchip code a bit more, see if the GPIO side ever disable/enables the clocks
17:46marcanjn: something something ps4 ;)
17:47sveni'm not sure if doing it like that makes sense fwiw. i just know that the clock framework does all the reference counting internally :)
17:47j_eyyou could enable the clock every time a gpio was set as an interrupt maybe? that would bump the reference counts..
17:47marcanit's not even a clock!
17:48j_eys/clock/thingy/
17:48marcanlet's not bikeshed over *gpio power management* until I figure out what to even do with power management ;)
17:48svenhow's that coming along btw? :>
17:48marcansven: I made the hv do smp
17:49marcanas part of wanting to work out the cpu pstate stuff :p
17:49skali is now known as skali_
17:49skali_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
17:49marcanunfortunately it doesn't actually control them right now, it seems; either I fail or it nerfs itself in EL1
17:49j_eynow I have 7 penguins on my screen!
17:49marcanhopefully not hard to figure out though
17:49marcanj_ey: ... 7? not 8?
17:49alyssaj_ey: macbook air? :-p
17:50marcanthat's the *gpu*
17:50alyssamarcan: hypervisor steals one right?
17:50alyssamarcan: i know :-p
17:50alyssaand oh, nice
17:50svendid you check if the mini even cares about freq scaling?
17:50marcanit used to, for the watchdog, but I just made that optional
17:50svenmaybe it's just a thing for the macbooks
17:50alyssasven: I think so
17:50marcansven: yes, in bare metal it works
17:50svenah. weird.
17:50marcanthe watchdog is only really useful for me anyway
17:50j_eymarcan: uh, yes, probably 8, I was thinking the wdt was still on
17:51marcanback when I was doing early bringup
17:51j_eymarcan: i didnt actually count them :P
17:51alyssasven: like, I think some magic registers are needed just to get the performance cores to reach peak perf on the mini? i think
17:51svenalyssa: yes
17:51alyssasven: or maybe it's just wishful thinking to explain why mesa seems to build faster on macos than linux :-p
17:51skali has joined #asahi
17:51yuyichao_ has joined #asahi
17:51marcanat least macos has the pcores at close to peak though
17:51marcannot sure if that's a default or "something it does in lieu of actually making it dynamic"
17:51svenalyssa: i just thought that the mini maybe brings them up to 100% once and then leaves them along because it doesn't need to save power
17:51marcanlike they're at 3GHz instead of the max 3.2
17:51sven*alone
17:51j_eyI switched back to 1 core, it boots a little faster in m1n1/hv
17:51marcanthe ecores, otoh, are very underclocked
17:52marcanwhich explains the HV slowness until now
17:52alyssasven: given the chips lineage I suspect anything mini specific is an after thought tbh
17:52svensure, but that would be software only
17:52alyssaI know what I said ;-)
17:52svenanyway, doesn't matter since it's apparently not the case
17:52marcanj_ey: multicore is useful when you're actually *doing* stuff like macos is :p
17:53marcanlinux takes a lot fewer instructions to boot ;)
17:53j_eymarcan: :-)
17:54marcanwait hold on
17:54marcancpu scaling works now
17:54marcannevermind lol, it's that I had the interval set to 1ms
17:55marcanI'm guessing it clocks them up initially for some reason :p
17:55alyssaboot fasterr
17:55j_eycan I get a platform_device* from a device*?
17:55marcanCPU 7 frequency: 623 MHz
17:55marcanCPU 7 idle residency: 99.65%
17:55marcanCPU 7 active residency: 0.35% (600 MHz: .35% 828 MHz: .00% 1056 MHz: 0% 1284 MHz: 0% 1500 MHz: 0% 1728 MHz: 0% 1956 MHz: 0% 2184 MHz: 0% 2388 MHz: 0% 2592 MHz: 0% 2772 MHz: 0% 2988 MHz: 0% 3096 MHz: 0% 3144 MHz: .00% 3204 MHz: .00%)
17:55marcanCPU 3 frequency: 1023 MHz
17:55marcanCPU 3 idle residency: 99.28%
17:55marcanCPU 3 active residency: 0.72% (600 MHz: .05% 972 MHz: .58% 1332 MHz: .04% 1704 MHz: .03% 2064 MHz: .02%)
17:55alyssaj_ey: uh, usually you go the other way around I think
17:55marcanlooks good
17:56alyssaj_ey: but maybe to_platform_device
17:56marcanPackage Power: 97 mW
17:56marcanI will say this thing has pretty impressive power management even on the mini ;)
17:56alyssaj_ey: yeah, to_platform_device
17:56alyssamarcan: nice!
17:56svenif you _know_ that device is a platform device to_platform_device will work
17:56j_eyalyssa: thanks!
17:57svenmarcan: 97 mW. whoa.
17:58alyssasven: I mean. my 2013 intel macbook air ran on like 4 W for the whole machine, and that was mostly the backlight
17:59skali has quit [Ping timeout: 480 seconds]
17:59alyssaadmittedly that was when idle after poking powertop for a few minutes
18:00minimy mini is running a few terminals, discord, a music player and two high res external displays (4k & 5k), and is still only using ~1.2W package power :) they're quite frugal
18:00mini(and would you believe this channel pings me a lot? :p)
18:00j_eyyou pinged yourself!
18:00miniI need to mute it, but it's always interesting!
18:01alyssai haven't done anything all day but the talk and it's 2pm
18:02psykosedoing things is overrated
18:08svengah... why is my builder server so slow today :/
18:11alessandrorzz[m]Are you creating the sw In c++?
18:12j_eysven: give it a good kick
18:12sveni'm sure they'll grant me access to their data center if i tell them i need to kick a server to make it go faster :D
18:18skali has joined #asahi
18:19tophevich[m]sven: actual metal or virtual server on a badly balanced "enterprise cloud virtualization platform cluster" like vsphere?
18:22marcanbtw, now that this works, did a bit of digging and I don't see macOS setting the PSINFO regs like corellium, nor touching that voltage bit thing
18:22marcanI should check on a laptop, but I'm getting the feeling whatever they were doing is obsolete/unnecessary... (bit of a theme with them)
18:22marcanall macos is really doing is changing the target pstate bits in the cluster control register
18:22marcanit does set the psinfo regs *once*
18:23marcanwill have to run some tests with benchmarks from m1n1 to confirm that it works as intended
18:23j_eymarcan: considering a lot of this is probably from project sandcastle, I guess that makes sense
18:24marcanyeah, and that firmware changes and gets better and presumably automates more things
18:26skali has quit [Ping timeout: 480 seconds]
18:30sventophevich[m]: the latter, and i know it's slow because of that. i'm just venting :)
18:55skali has joined #asahi
19:03skali has quit [Ping timeout: 480 seconds]
19:09erincandescent has quit [Remote host closed the connection]
19:10erincandescent has joined #asahi
19:15aleasto has quit [Quit: Konversation terminated!]
19:18tophevich[m]sven: I was asking because of the kicking picture you were painting, and now I see you kicking random racks all over in my mind :P
19:20aleasto has joined #asahi
19:22skali has joined #asahi
19:29aleasto has quit [Remote host closed the connection]
19:30skali has quit [Ping timeout: 480 seconds]
19:32jbowen has quit [Ping timeout: 480 seconds]
19:46jbowen has joined #asahi
19:46sventophevich[m]: :D
19:49skali has joined #asahi
19:57skali has quit [Ping timeout: 480 seconds]
20:22psykose has quit [Ping timeout: 480 seconds]
20:22psykose has joined #asahi
20:25skali has joined #asahi
20:33skali has quit [Ping timeout: 480 seconds]
21:02skali has joined #asahi
21:10skali has quit [Ping timeout: 480 seconds]
21:36jbowen has quit [Quit: leaving]
21:38skali has joined #asahi
21:46skali has quit [Ping timeout: 480 seconds]
21:54alyssamarcan: cpufreq came early in the corellium tree which means it's probably from sandcastle
21:57alyssa(thunderbolt, obviously different story)
22:14skali has joined #asahi
22:22skali has quit [Ping timeout: 480 seconds]
22:24yuyichao_ has quit [Ping timeout: 480 seconds]
22:42skali has joined #asahi
22:50skali has quit [Ping timeout: 480 seconds]
22:58ella-0 has joined #asahi
23:18skali has joined #asahi
23:26skali has quit [Ping timeout: 480 seconds]
23:55skali has joined #asahi
…