#
…

2025-04-04

00:08mischa85 has quit [Ping timeout: 480 seconds]
00:19chrisl has joined #asahi-dev
00:27chrisl has quit [Ping timeout: 480 seconds]
00:33mini_ has quit [Quit: ZNC closing...]
00:43mini_ has joined #asahi-dev
00:57mini_ has quit [Quit: ZNC closing...]
00:57mini_ has joined #asahi-dev
01:12pb17 has quit [Remote host closed the connection]
01:13chrisl has joined #asahi-dev
01:21chrisl has quit [Ping timeout: 480 seconds]
01:34pb17 has joined #asahi-dev
02:00mischa85 has joined #asahi-dev
02:04tobhe_ has joined #asahi-dev
02:09mischa85 has quit [Ping timeout: 480 seconds]
02:11tobhe has quit [Ping timeout: 480 seconds]
02:14chrisl has joined #asahi-dev
02:22chrisl has quit [Ping timeout: 480 seconds]
03:14chrisl has joined #asahi-dev
03:22chrisl has quit [Ping timeout: 480 seconds]
03:27DannyBee has joined #asahi-dev
03:31DannyB has quit [Ping timeout: 480 seconds]
03:35pb17 has quit [Ping timeout: 480 seconds]
03:58nora_ has joined #asahi-dev
03:59pb17 has joined #asahi-dev
04:00nora has quit [Ping timeout: 480 seconds]
04:01mischa85 has joined #asahi-dev
04:09mischa85 has quit [Ping timeout: 480 seconds]
04:21runxiyu has quit [Ping timeout: 480 seconds]
04:21runxiyu_ has quit [Read error: Connection reset by peer]
04:22runxiyu has joined #asahi-dev
04:23chrisl has joined #asahi-dev
04:29mattia013_ has joined #asahi-dev
04:29eluks has quit [Remote host closed the connection]
04:30eluks has joined #asahi-dev
04:31chrisl has quit [Ping timeout: 480 seconds]
04:46nickchandoes ACTLR TSO work for raw boot objects at least?
04:52chadmedsince its part of the arm spec i dont see why theyd gate that
04:52mattia013_ has quit [Ping timeout: 480 seconds]
04:58nickchanthat and HACR traps
04:58nickchan we use those in Linux
05:00JayBeeFOSS has quit [Ping timeout: 480 seconds]
05:01JayBeeFOSS has joined #asahi-dev
05:04umaga has joined #asahi-dev
05:06umaga is now known as yussef
05:07yussef is now known as umaga
05:10nickchanand the HACR one is *not* compliant at least the PMUv3 part
05:12mattia013 has quit [Remote host closed the connection]
05:12mattia013 has joined #asahi-dev
05:19mattia013_ has joined #asahi-dev
05:20umaga has quit [Quit: umaga]
05:20umaga has joined #asahi-dev
05:22chadmedhopefully they left it as-is since those blas jokers didnt bring it up
05:22chadmed has quit [Quit: Konversation terminated!]
05:24chadmed has joined #asahi-dev
05:34pb17 has quit [Ping timeout: 480 seconds]
05:35umaga has quit [Quit: umaga]
05:42chadmed has quit [Quit: Konversation terminated!]
05:47pb17 has joined #asahi-dev
05:48umaga has joined #asahi-dev
05:49svenanother thing that’s going to be annoying: DART etc is inside SPTM so we won’t be able to trace that
05:56nickchansven: cpu power management is probably inside sptm too
05:58nickchansigh so it's either not being able to trace some things or playing cat and mouse
06:00mattia013 has quit [Remote host closed the connection]
06:00mattia013_ is now known as mattia013
06:00nickchanand the second option means keeping some things secret which is not very open source (though it should be possible to keep the shipped binaries fully open source as long as apple did not mess with hacr or actlr tso)
06:00mattia013_ has joined #asahi-dev
06:01mattia013_ has quit [Remote host closed the connection]
06:02mischa85 has joined #asahi-dev
06:02mattia013_ has joined #asahi-dev
06:05sventhere’s also the issue that booting on >15.1 seems to fail. maybe they accidentally broke raw boot objects and didn’t notice 🙃
06:08svenbut anyway, I’m not going to play cat and mouse. I’ve done that ages ago on consoles and don’t want to ever do that again. it’s just no fun
06:08umaga has quit [Ping timeout: 480 seconds]
06:09nickchanthink some other people may do it however
06:10nickchanwe will see
06:10mischa85 has quit [Ping timeout: 480 seconds]
06:12mattia013 has quit [Ping timeout: 480 seconds]
06:12mattia013_ is now known as mattia013
06:12svenas long as they’re gonna reverse all the new hardware and write drivers then :D
06:13sventhough I guess it’s just iommus and power stuff that’s hidden in SPTM and that binary isn’t very large
06:13svenfor the rest patching our tracers into xnu might just work
06:21nickchansven: xnu patching, where should it go? (python or in m1n1)
06:22nickchanand the thing with any kind of patcher is that it will neccessary contain xnu instructions either in code or in comments
06:24nickchanwhich would be directly from binary RE given the situations
06:25nickchannot looking forwards to the copyright implications
06:26svenit’s going to be painful, yes
06:27svenhaven’t thought about it in details but probably hijack exception vectors and maybe proxy calls to SPTM to mess with page tables
06:34chadmed has joined #asahi-dev
06:35chadmedwhat actually made them do this? was it literally those guys being like "we should implement AMX in linux"
06:42svenmaybe that. maybe arm lawyer being annoyed just because they were bored. maybe both. :/
06:43mattia013 has quit [Remote host closed the connection]
06:43mattia013 has joined #asahi-dev
06:48nickchanchadmed: who is "those guys"
06:49chadmedaaaages ago some folks were asking around about adding AMX support to arm64 linux. we always maintained that we would never _ever_ support anything even remotely like that but i think upstream caught wind of it
06:49mattia013 has quit [Remote host closed the connection]
06:50mattia013 has joined #asahi-dev
06:51mattia013 has quit [Remote host closed the connection]
06:51mattia013 has joined #asahi-dev
06:51nickchani think there's also a thing where some performance counters have some edge case regarding amx
06:53nickchanah, *that* got posted onto the mailing list
06:55chadmedthen of course there was the whole apple-specific TSO discussion that kinda went nowhere which probably didnt help in that regard especially since FEAT_LRCPC was already a thing
06:55chadmedyeah which is why i think arm probably had some words to apple and they picked the nuclear option. which is just kind of really dickish of them
06:55nickchanand after looking into xnu sources it seems that a12 and possibly a13 have TSO too, but not ACTLR, it's HID9[16]
06:56nickchannot like that matters in practice already confirmed it does not exist on a11
06:56chadmedwhos going to be doing x86 emu on an iphone anyway :p
06:58chadmedwel... i guess it's good that iboot is not encrypted now :/
07:00chadmedthis doesnt affect M3, only M4 (and all new socs going forward presumably) right?
07:02nickchanchadmed: another thing
07:02nickchanfrom some of the m1n1 m4 support code i think wfi is also broken somehow
07:03nickchanand we use that and not having that would make for a really bad experience
07:04mattia013 has quit [Remote host closed the connection]
07:04chadmedyeah id rather not go back to space heater laptops
07:05chadmedtheres no way thats being trapped to sptm or anything right
07:06nickchanchadmed: i think power management things and cpu start stop things like that are handled by sptm because of state loss
07:06mattia013 has joined #asahi-dev
07:06chadmedso if sptm is torn down due to a raw boot object we just lose deep wfi?
07:07nickchanit does look like it assuming the m4 forks is not making a horribly incorrect assumption
07:07chadmedthats fucking stupid and surely not intended
07:08chaos_princessIs there any m4 fork that does chicken bits and wfi configuration correctly
07:09nickchanchaos_princess: CYC_OVRD is supposedly completely inaccessible so there is no wfi configuration to speak of
07:09nickchanand yuka's fork forced the a7-a11 code path on wfi which is a shallow wfi
07:10nickchanand that means on m4 by default wfi is already the shallow wfi
07:10chaos_princessWould be interesting to see if cyc_ovrd is inaccessible just in el2
07:12chadmedif apple hid it in gl2 that would kinda defeat the purpose of doing this to make raw boot objects only see the in-spec architecture
07:12chadmeddunno if arm would be happy about that
07:13pb17 has quit [Ping timeout: 480 seconds]
07:14nickchanchadmed: why would hiding cyc_ovrd defeat the purpose?
07:14nickchanthough I can see that will make people much more willing to play cat and mouse
07:15nickchanspace heater is unfun
07:17nickchanthere are some alternatives that have not been tried however like turning off cpu power domains in pmgr
07:32chadmednickchan: oh thats an apple register
07:32ece314378925355451680698427415 has quit [Ping timeout: 480 seconds]
07:37chrisl has joined #asahi-dev
07:39mischa85 has joined #asahi-dev
07:45chrisl has quit [Ping timeout: 480 seconds]
07:45jannau_ is now known as jannau
07:46pb17 has joined #asahi-dev
07:48mischa85 has quit [Ping timeout: 480 seconds]
08:02svenany of the deeper wfi states being broken would be very bad :(
08:06alessaba has joined #asahi-dev
08:08Mary has quit [Quit: .]
08:09Mary has joined #asahi-dev
08:10mischa85 has joined #asahi-dev
08:21mischa85 has quit [Ping timeout: 480 seconds]
08:27mazjannau: what was preventing this from working before my changes? I didn't really expect any functional difference with them.
08:28jannaumaz: it has 2 root ports which was incompatible with the bus_notifier approach
08:29mischa85 has joined #asahi-dev
08:30mazjannau: huh, right. oh well, less code anyway, and additional functionality. too bad I can't afford that box! :)
08:37mischa85 has quit [Ping timeout: 480 seconds]
08:40PaulFertser has quit [Ping timeout: 480 seconds]
08:45mischa85 has joined #asahi-dev
08:51jannauI got it refurbished for a little bit over half price. still expensive though
08:52jannaupcie setup is not fully working. the switch management endpoint is not working and the devices behind the switch only come up with PCIe gen1
08:52mazyeah. it would take my M1 studio to die on me to be able to justify it. and even then, a 128 core Altra could make more sense, given what I do with it.
08:53mischa85 has quit [Ping timeout: 480 seconds]
08:54mazI usually have to play all sort of command-line tricks to deal with PCIe switches...
08:55mazdo we have a native driver for the switch in this box? or relying on the generic Linux code?
08:56jannauyes, switchtec. it's a microchip PM40100
08:57chaos_princessjannau: oh, wow, i salute you and your wallet for that sacrifice.
08:57chaos_princessdid they fix device mappings on pcie in that box?
08:58PaulFertser has joined #asahi-dev
08:58jannauI haven't checked beyond quickly testing that old radeon hd4350
09:01jannaumaz: mem assigment breaks down with hotplug but if all devices are powered up before booting it mostly works after putting "bus-range = <0 8>;" in the device tree
09:03mischa85 has joined #asahi-dev
09:03jannaunot sure how to model the reset and power enable gpios of devices behind the switch in the devicetree
09:03mazhuh, what a mess.
09:08mazI guess this should be part of describing the slots past the switch, but how to express that numbering is a mystery. wonder if there's prior art around that.
09:08mazrobher: any clue?
09:11mischa85 has quit [Ping timeout: 480 seconds]
09:14jannauarch/arm64/boot/dts/qcom/sa8540p-ride.dts has a PM40028 switch but doesn't model devices behind the switch
09:27pb17 has quit [Ping timeout: 480 seconds]
09:39calebccff has quit [Quit: ZNC 1.8.2+deb3.1+deb12u1 - https://znc.in]
09:40calebccff has joined #asahi-dev
09:40fl0_idwhy do you think they even still allow raw / custom boot objects, if they apparentyl locked down M4 so tightly?
09:41mischa85 has joined #asahi-dev
09:51pb17 has joined #asahi-dev
09:53mischa85 has quit [Ping timeout: 480 seconds]
10:13chadmedfl0_id: they locked down apple's extensions to the ARM ISA
10:14chadmedthe platform is designed explicitly to allow custom boot objects to exist. arm got shitty with them for letting those custom objects "see" all of apples not-ARM things
10:15mischa85 has joined #asahi-dev
10:18fl0_idand these just happen to be important bits I guess
10:19svenit’s mostly annoying because our Hv for reverse engineering won’t work without these extensions
10:19svenplus deep sleep states might be affected which would actually be important for running Linux
10:20LarstiQI don't suppose the blessed Arm ISA has working alternatives for those?
10:21fl0_idsven yeah deep sleep states would be rly bad :/
10:24mischa85 has quit [Ping timeout: 480 seconds]
10:26svenyeah, the rest we can (painfully) work around probably
10:26svenbut sleep might just be a bug, ie they just disabled everything and didn’t realize that the sleep msr is very important
10:27svenor maybe it just works differently on m4. We don’t really know yet
10:28fl0_idsure. thx for explaining
10:29svenstill all pretty far away since we need to get m1/2 upstream and then ideally bring up m3 directly upstream as well
10:33mischa85 has joined #asahi-dev
10:41mischa85 has quit [Ping timeout: 480 seconds]
10:43fl0_idyeah for sure
10:44fl0_idmain issue for me is time and that my m3 is also my work machine, soo I cannot boot it into m1n1 that often
10:48chadmed has quit [Quit: Konversation terminated!]
10:49chadmed has joined #asahi-dev
10:50mischa85 has joined #asahi-dev
10:50sventime is probably the issue for all of us :D
10:55fl0_idyeah, not saying I'm unique haha
11:04mischa85 has quit [Ping timeout: 480 seconds]
11:15johey has quit [Ping timeout: 480 seconds]
11:20johey has joined #asahi-dev
11:27mischa85 has joined #asahi-dev
11:31pb17 has quit [Ping timeout: 480 seconds]
11:35mischa85 has quit [Ping timeout: 480 seconds]
11:43mischa85 has joined #asahi-dev
11:46pb17 has joined #asahi-dev
11:51mischa85 has quit [Ping timeout: 480 seconds]
11:59mischa85 has joined #asahi-dev
12:12mischa85 has quit [Ping timeout: 480 seconds]
12:22mischa85 has joined #asahi-dev
12:30mischa85 has quit [Ping timeout: 480 seconds]
12:38mischa85 has joined #asahi-dev
12:43yuyichao_ has quit [Ping timeout: 480 seconds]
12:46chrisl has joined #asahi-dev
12:47mischa85 has quit [Ping timeout: 480 seconds]
12:54chrisl has quit [Ping timeout: 480 seconds]
12:55mischa85 has joined #asahi-dev
12:56robhermaz: Certainly possible. PPC boots with all the PCI hierarchy described. The kernel can in fact generate the DT PCI nodes for you. Boot with CONFIG_PCI_DYNAMIC_OF_NODES enabled. Add what you need from the DT to your dts. Then add the GPIOs you need to handle.
12:57robhermaz: There's also a switch example on hikey970 (or 960 maybe?).
12:58john-cabaj has joined #asahi-dev
13:01robhermaz: there's also the linux,pci-probe-only flag if keeping the firmware configuration is needed.
13:04mischa85 has quit [Ping timeout: 480 seconds]
13:09pjakobsson has quit []
13:16pb17 has quit [Ping timeout: 480 seconds]
13:58pb17 has joined #asahi-dev
14:07Halian has quit [Ping timeout: 480 seconds]
14:12alessaba has quit [Quit: Connection closed for inactivity]
14:18yuyichao_ has joined #asahi-dev
14:48chrisl has joined #asahi-dev
14:55swiftraccoon has quit [Ping timeout: 480 seconds]
14:56chrisl has quit [Ping timeout: 480 seconds]
14:56mischa85 has joined #asahi-dev
15:04chrisl has joined #asahi-dev
15:04mischa85 has quit [Ping timeout: 480 seconds]
15:12chrisl has quit [Ping timeout: 480 seconds]
15:59pb17 has quit [Ping timeout: 480 seconds]
16:04mischa85 has joined #asahi-dev
16:04chrisl has joined #asahi-dev
16:12mischa85 has quit [Ping timeout: 480 seconds]
16:12chrisl has quit [Ping timeout: 480 seconds]
16:28pb17 has joined #asahi-dev
16:35mischa85 has joined #asahi-dev
16:40tanty has quit [Ping timeout: 480 seconds]
16:43mischa85 has quit [Ping timeout: 480 seconds]
16:52mischa85 has joined #asahi-dev
16:59tanty has joined #asahi-dev
17:00mischa85 has quit [Ping timeout: 480 seconds]
17:04mischa85 has joined #asahi-dev
17:05chrisl has joined #asahi-dev
17:06mischa85well if there's a will there's a way
17:07tanty has quit [Ping timeout: 480 seconds]
17:08mischa85nickchan: you mentioned trying disabling cpu power domains to gain write access to cyc_ovrd. how to? considering there's no ps-regs and the power registers are still unknown
17:10mischa85_ has joined #asahi-dev
17:13chrisl has quit [Ping timeout: 480 seconds]
17:18mischa85_ has quit [Ping timeout: 480 seconds]
17:29___nick___ has joined #asahi-dev
17:29___nick___ has quit []
17:32___nick___ has joined #asahi-dev
17:43tanty has joined #asahi-dev
17:51chrisl has joined #asahi-dev
17:52swiftraccoon has joined #asahi-dev
17:57alessaba has joined #asahi-dev
17:58kettenismaz, robher: there also is CONFIG_PCI_PWRCTL now
17:59chrisl has quit [Ping timeout: 480 seconds]
17:59kettenisI believe that is the preferred way to power on PCI devices that are soldered onto a system board
18:14pb17 has quit [Ping timeout: 480 seconds]
18:18CME has quit [Read error: Connection reset by peer]
18:18CME has joined #asahi-dev
18:33pb17 has joined #asahi-dev
18:51chrisl has joined #asahi-dev
19:00chrisl has quit [Ping timeout: 480 seconds]
19:10mischa85_ has joined #asahi-dev
19:19mischa85_ has quit [Ping timeout: 480 seconds]
19:20mischa85 has quit [Ping timeout: 480 seconds]
19:34checkfoc_us9 has quit [Remote host closed the connection]
19:36checkfoc_us9 has joined #asahi-dev
19:38cyrinux9490 has quit []
19:39cyrinux9490 has joined #asahi-dev
19:43mischa85 has joined #asahi-dev
19:50pb17 has quit [Ping timeout: 480 seconds]
19:51mischa85_ has joined #asahi-dev
19:52mischa85 has quit [Ping timeout: 480 seconds]
19:52chrisl has joined #asahi-dev
19:53WindowPain has quit [Read error: Connection reset by peer]
19:55WindowPain has joined #asahi-dev
19:59mischa85 has joined #asahi-dev
20:00mischa85_ has quit [Ping timeout: 480 seconds]
20:00chrisl has quit [Ping timeout: 480 seconds]
20:03___nick___ has quit [Remote host closed the connection]
20:06pb17 has joined #asahi-dev
20:07mischa85 has quit [Ping timeout: 480 seconds]
20:07alessaba has quit [Quit: Connection closed for inactivity]
20:10mischa85 has joined #asahi-dev
20:11mischa85_ has joined #asahi-dev
20:14tanty has quit [Quit: Ciao!]
20:18mischa85 has quit [Ping timeout: 480 seconds]
20:26tanty has joined #asahi-dev
20:27chrisl has joined #asahi-dev
20:27tanty has quit []
20:34tanty has joined #asahi-dev
20:38chrisl has quit [Ping timeout: 480 seconds]
20:41tanty has quit [Quit: Ciao!]
20:42tanty has joined #asahi-dev
20:49tanty has quit [Quit: Ciao!]
20:51tanty has joined #asahi-dev
21:22mischa85_ has quit [Ping timeout: 480 seconds]
21:25pb17 has quit [Ping timeout: 480 seconds]
21:27chrisl has joined #asahi-dev
21:31pb17 has joined #asahi-dev
21:35chrisl has quit [Ping timeout: 480 seconds]
21:48tanty has quit [Quit: Ciao!]
21:50tanty has joined #asahi-dev
21:55swiftraccoon has quit [Read error: Connection reset by peer]
21:56swiftraccoon has joined #asahi-dev
21:57red_rider has joined #asahi-dev
21:58tanty has quit [Quit: Ciao!]
22:00tanty has joined #asahi-dev
22:03tanty has quit []
22:09tanty has joined #asahi-dev
22:13swiftraccoon2 has joined #asahi-dev
22:14mischa85 has joined #asahi-dev
22:18tanty has quit [Quit: Ciao!]
22:19chrisl has joined #asahi-dev
22:20swiftraccoon has quit [Ping timeout: 480 seconds]
22:21tanty has joined #asahi-dev
22:22red_rider has quit [Quit: Leaving]
22:22mischa85 has quit [Ping timeout: 480 seconds]
22:22red_rider has joined #asahi-dev
22:23red_rider has quit []
22:25tanty has quit []
22:27chrisl has quit [Ping timeout: 480 seconds]
22:28chrisl has joined #asahi-dev
22:31tanty has joined #asahi-dev
22:34tanty has quit []
22:37tanty has joined #asahi-dev
22:41tanty has quit []
22:45tanty has joined #asahi-dev
22:48tanty has quit []
22:48tanty has joined #asahi-dev
22:48tanty has quit []
22:49tanty has joined #asahi-dev
22:52tanty has quit []
22:54tanty has joined #asahi-dev
23:00tanty has quit [Quit: Ciao!]
23:02red_rider has joined #asahi-dev
23:03tanty has joined #asahi-dev
23:04tanty has quit []
23:05tanty has joined #asahi-dev
23:07pb17 has quit [Ping timeout: 480 seconds]
23:07chadmed has quit [Quit: Konversation terminated!]
23:07chadmed has joined #asahi-dev
23:08chadmedchaos_princess: do you have any thoughts on https://github.com/AsahiLinux/speakersafetyd/pull/26? my thoughts is to just leave this alone until we rip out the black box
23:08chadmedcc flokli
23:09chadmedfor context: the black box is kinda useless now that the visense stuff is stable. id rather rip that out before merging this to eliminate some churn
23:09tubtub_ has joined #asahi-dev
23:09tubtub_ has quit [Remote host closed the connection]
23:11chaos_princessi am not clear as to how it fixes things? even with the boundary directory bit, are you still not leaving files owned by unassigned uid?
23:11tanty has quit [Remote host closed the connection]
23:11chadmedseems that way yeah
23:12chadmedthe black box is also not really "state"
23:12chaos_princessthis feels like a feature that is for files that are owned by the unit and not accessed by anything else
23:12chadmedyou wouldnt call any other log file "state"
23:13tanty has joined #asahi-dev
23:13tanty has quit []
23:15mischa85 has joined #asahi-dev
23:15red_rider has quit [Ping timeout: 480 seconds]
23:16tanty has joined #asahi-dev
23:18floklichadmed: what's in the black box
23:19flokliCan we improve logging sufficiently, pipe everything needed to the journal during the crash handler?
23:20flokliAh, now reading "visense stuff is stable, we can rip it out"
23:20chaos_princessblackbox is kinda like a core dump file
23:21chaos_princess(it is a binary)
23:22chadmedno we cant pipe everything to the journal on crash because its PCM samples :p
23:22chadmedit doesnt matter, i will be ripping it out at some point sooni(ish)
23:22chadmedthen we can look at dynamic users
23:22chaos_princesschadmed: json array of floats :P
23:23mischa85 has quit [Ping timeout: 480 seconds]
23:23chadmedchaos_princess: you only need one column type in your database and it's JSONB
23:23floklichadmed: I'm happy to update the PR to drop the blackbox functionality, or rebase on top of such a change if you want to do it.
23:24chadmedflokli: yeah hold off until i rip it out. ive got a bunch of other stuff to take care of first
23:24chaos_princessJust let it sit for now, we can rebase it later
23:24chaos_princessNot like there is a ton of changes happening to those files
23:24chadmedi need to submit v4 of the codec patches upstream today without the tdm idle mask stuff. since that was the last controversial bit hopefully those get merged without a fuss
23:24tanty has quit [Quit: Ciao!]
23:25chadmedi am deeply concerned about the fact that the upstream-approved tdm idle mask stuff ended up not working but i'll worry about that when it's worth worrying about
23:25tanty has joined #asahi-dev
23:31tanty has quit [Quit: Ciao!]
23:32tanty has joined #asahi-dev
23:34chrisl has quit [Ping timeout: 480 seconds]
23:36pb17 has joined #asahi-dev
23:57mischa85 has joined #asahi-dev
…