#
…

2022-07-15

00:03quarkyalice_ has quit [Remote host closed the connection]
00:04quarkyalice_ has joined #asahi-dev
00:05millenialhacker has joined #asahi-dev
00:10doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
00:11quarkyalice_ has quit [Remote host closed the connection]
00:12quarkyalice_ has joined #asahi-dev
00:12Catyre has quit [Remote host closed the connection]
00:12Catyre has joined #asahi-dev
00:13millenialhacker has quit [Ping timeout: 480 seconds]
00:23quarkyalice_ has quit [Remote host closed the connection]
00:24quarkyalice_ has joined #asahi-dev
00:30nicolas17 has quit [Quit: moving router to UPS]
00:33doggkruse has joined #asahi-dev
00:35quarkyalice_ has quit [Remote host closed the connection]
00:38Catyre has quit [Remote host closed the connection]
00:39Catyre has joined #asahi-dev
00:42Catyre has quit [Remote host closed the connection]
00:42Catyre has joined #asahi-dev
00:45gladiac is now known as Guest5037
00:45gladiac has joined #asahi-dev
00:47gladiac has quit []
00:48gladiac has joined #asahi-dev
00:50Guest5037 has quit [Ping timeout: 480 seconds]
00:55gladiac is now known as Guest5040
00:55gladiac has joined #asahi-dev
00:55doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
00:57Catyre has quit [Remote host closed the connection]
00:58Catyre has joined #asahi-dev
01:00Guest5040 has quit [Ping timeout: 480 seconds]
01:12Catyre has quit [Ping timeout: 480 seconds]
01:17nicolas17 has joined #asahi-dev
01:57chadmed has joined #asahi-dev
01:57chadmedamarioguy: the reason to not do that is as you say, because it would never fly upstream
01:58chadmedgiven that the goal is to eventually "stop existing" as it were, it would make no sense to spend time developing something with literally zero chance of being accepted upstream
02:32Catyre has joined #asahi-dev
02:43chadmed has quit [Quit: Konversation terminated!]
02:48Mary has quit [Quit: The Lounge - https://thelounge.chat]
02:48Mary has joined #asahi-dev
02:49Catyre has quit [Remote host closed the connection]
02:49Catyre has joined #asahi-dev
02:53chadmed has joined #asahi-dev
02:58phiologe has joined #asahi-dev
03:03Catyre has quit [Ping timeout: 480 seconds]
03:07millenialhacker has joined #asahi-dev
03:08chadmed has quit [Quit: Konversation terminated!]
03:15millenialhacker has quit [Ping timeout: 480 seconds]
03:36kov has quit [Quit: Coyote finally caught me]
03:42doggkruse has joined #asahi-dev
03:50millenialhacker has joined #asahi-dev
03:58millenialhacker has quit [Ping timeout: 480 seconds]
04:02svenyou need SPRR when you enable GXF and that’s not going to be possible upstream since it completely changes the meaning of pagetable permissions
04:02svenand you also don’t have different pagetables in GL2 which is one of the things that makes the runtime service easier
04:15nicolas17 has quit [Read error: Connection reset by peer]
04:15nicolas17 has joined #asahi-dev
04:43chadmed has joined #asahi-dev
04:57doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
05:23jluthra has quit [Ping timeout: 480 seconds]
05:35Catyre has joined #asahi-dev
05:43jluthra has joined #asahi-dev
05:54Catyre has quit [Remote host closed the connection]
05:54Catyre has joined #asahi-dev
06:02doggkruse has joined #asahi-dev
06:02doggkruse has quit []
06:03Catyre has quit [Ping timeout: 480 seconds]
06:08Catyre has joined #asahi-dev
06:13chadmedFDT support has been merged into the kde system info thing :)
06:14chadmedwith the m1n1 PR this means that asahi desktop users should be able to see their serial number (and other stuff) from linux
06:15the_lanetly_052 has joined #asahi-dev
06:22millenialhacker has joined #asahi-dev
06:30millenialhacker has quit [Ping timeout: 480 seconds]
06:31Catyre has quit [Remote host closed the connection]
06:32Catyre has joined #asahi-dev
06:40Catyre has quit [Ping timeout: 480 seconds]
06:47Catyre has joined #asahi-dev
06:53the_lanetly_052 has quit [Ping timeout: 480 seconds]
06:58Dcow[m]what other stuff?
06:59millenialhacker has joined #asahi-dev
06:59chadmedcurrently, the model from the devicetree and the version of u-boot they're using
07:00chadmedas cool as it would be, adding m1n1 specific stuff would probably require carrying downstream patches to kinfocenter which would be annoying to maintain
07:01chadmedi went with serial number, model and uboot version because theyre canonical devicetree properties that many different systems will have
07:07marcanany concerns about pushing that dev release to prod soonish? I'm out today, but I can probably do something tonight
07:07marcansven: might want to fix that LE thing, but I won't cut new images for that probably, just push out a day 0 kernel update for new installs
07:07millenialhacker has quit [Ping timeout: 480 seconds]
07:07sventhe LE thing is fixed in the current bluetooth-wip fwiw
07:08svenit also sends the proper calibration data for M2 now though I'm not sure that makes any difference
07:09sven(and maybe fixes 4377 but who knows)
07:11nicolas17chadmed: got a screenshot btw?
07:12chadmedthis includes the MIDR_EL1 kf5solid patch that has also been merged
07:12Dcow[m]very nice
07:13chadmedyeah its not a huge thing in the grand scheme of things but its a nice QoL improvement at least and will benefit other systems too
07:13Catyre has quit [Remote host closed the connection]
07:14Catyre has joined #asahi-dev
07:18marcansven: great, will pull it later
07:18sveni've changed the name from 43xx to 4377 fwiw since there are tons of 43xx chips from broadcom that also do bluetooth but are just UART
07:18marcanyeah, makes sense
07:18sven(and because alyssa complained that BCM43XX looks ugly :D)
07:19marcanmight cut new images then anyway if I have time, not like it takes much
07:19marcanalright, off to fly in a Cessna :p
07:19marcan(my brother is a pilot now)
07:19svenhave fun!
07:20sventhose are much more fun that one of these big planes :>
07:20JTLa) didn't know you had a brother b) have fun
07:30Catyre has quit [Ping timeout: 480 seconds]
07:32_jannau_marcan: no concerns for m1* devices here. m2 mtp firmware has become a little annoying. Sometimes it takes more than 5 reboots for it to work
07:34Catyre has joined #asahi-dev
07:35j`eymarcan living it up in a private jet :P
07:37Catyre has quit [Remote host closed the connection]
07:45dhewgbcm43xx sounds old, that's also what the original wii had, right?
07:47JTLdhewg: didn't old powerbooks also use it for wireless?
07:48JTLI vaguely remember having to mess with broadcom blobs trying to get wifi working under linux with powerbook g4
07:57kloenk has quit [Remote host closed the connection]
07:58kloenk has joined #asahi-dev
07:59chadmedyeah its a bit ridiculous that they still use the 43xx nomenclature tbh
07:59chadmedmost things about broadcom are a bit ridiculous though so...
08:00dhewgmaybe, I dunno, but I do remember the wii. That was the b43 legacy driver with the openfwwf firmware through. I hope this isn't as ancient as that ;)
08:02the_lanetly_052 has joined #asahi-dev
08:03svenmany of broadcom's devices start with 43, i just didn't realize and that's why I called the driver bcm43xx
08:07millenialhacker has joined #asahi-dev
08:15millenialhacker has quit [Ping timeout: 480 seconds]
08:20L1Q has joined #asahi-dev
08:21leitao has joined #asahi-dev
08:22leitao has quit [Read error: No route to host]
08:23leitao has joined #asahi-dev
08:24leitao has quit []
09:45sirn has joined #asahi-dev
09:45sirn has quit [Remote host closed the connection]
09:45kettenischadmed: the consensus in the ARM/UEFI universe seems to be to provide information about firmware/hardware through SMBIOS
09:46walid has joined #asahi-dev
09:46kettenisu-boot supports this, but only implements the traditional "32-bit" interface
09:47ketteniswhere pointers to the relevant information blobs are 32-bit
09:48kettenisthis doesn't work on Apple hardware since there is no memory at addresses below 4G to store the SMBIOS tables
09:48sirn has joined #asahi-dev
09:48kettenisthat's why I disabled SMBIOS support in u-boot
09:49kettenissomebody who cares could probably implement the 64-bit interface for that and populate the tables with interesting stuff
09:58nicolas17 has quit [Read error: Connection reset by peer]
09:59nicolas17 has joined #asahi-dev
10:05chadmedyeah i understand that and kinfocenter already supports SMBIOS/DMI but there are really just loads of devices that dont use it anyway. a certain variety of wildly popular SBCs for example
10:10chadmedlike there are just so many arm platforms, including this one, that will never fully conform the BSA or any variety of SystemReady for various reasons
10:12chadmedespecially when you can achieve ~equivalent functionality to exposing SMBIOS tables with 50 lines of code in this case
10:16kettenisfair enough; just wanted to make sure you were aware of this in case you wanted to go further down that path
10:17chadmedit would be helpful for the BSDs i guess since afaik theres no filesystem where the FDT is exposed but id personally be worried about scope creep and duplicated efforts if we got stuck into it
10:18chadmedand at the end of the day we'd just end up with a peculiar mix of SBBR and EBBR infrastructure. like if we implement SMBIOS support, how long before we end up saying "fk it" and doing ACPI too
10:18chadmedall on top of u-boot :P
10:18mazchadmed: try and fit AIC in ACPI first...
10:20chadmedscratch up the SPI tracks on the motherboard and tack on a 64Mb flash chip for the tables
10:23kettenisfdt is exposed through ioctls in OpenBSD (the same way as the OpenFirmware device tree is exposed on sparc64 and the old Apple powerpc machines)
10:23kettenisthink it is the same way on other BSDs
10:24kettenisanyway, exposing this information in a GUI isn't something I particularly care about
10:26mps`cat /proc/device-tree/model` is fine for me
10:26chadmedyeah fair, i just thought it was some nice low hanging fruit to marginally improve quality of life for desktop users
10:29chadmedimo its really the little things like this which contribute most to how janky people feel their experience is
10:30chadmedpeople go into it knowing which features will be missing but when they go to see something like 10 Unknown Processors @ 0 GHz it can be really offputting
10:30mpschadmed: you are right (but personally I don't care because I prefer text)
10:32chadmedi mean these problems exist in text too but yeah
10:34sirn has quit [Quit: ZNC 1.8.2 - https://znc.in]
10:34sirn has joined #asahi-dev
10:35sirn has quit [Read error: Connection reset by peer]
10:36sirn has joined #asahi-dev
10:45mpsyes, not yet all data but will be in future, probably
10:46jakebot6 has quit [Quit: Ping timeout (120 seconds)]
10:46jakebot6 has joined #asahi-dev
10:53chadmedlscpu should get support for all the current devices in the next util-linux
10:56mpschadmed: https://tpaste.us/4og1 that is what lscpu shows on my m1 mbp
10:57chadmedyeah it was missing pro, max and ultra
10:58chadmedbecause they are different part IDs it would show nothing for CPU part
10:58mpscpufetch -s legacy shows this https://tpaste.us/MkNk
10:59mpsSoC: unknown and Technology: unknown
11:01chadmedsupport was added in may there, so just after the last release
11:11mpsah, I didn't tried new kernels
11:13chadmedi cant actually make sense of how this thing actually identifies arm cpus
11:13chadmedit just has an enormous enum list of all the socs it supports
11:13chadmedi cant find anywhere where it actually tries to match the part id in midr_el1 to anything
11:15chadmedoh thats because it only does apple chips on macos lmao
11:24chadmedthis is the most overengineered cpu information listing tool
11:24chadmedreminds me of enterprise fizzbuzz
11:34kov has joined #asahi-dev
11:35chadmed has quit [Quit: Konversation terminated!]
12:02millenialhacker has joined #asahi-dev
12:10millenialhacker has quit [Ping timeout: 480 seconds]
12:16millenialhacker has joined #asahi-dev
12:19the_lanetly_052 has quit [Ping timeout: 480 seconds]
12:24millenialhacker has quit [Ping timeout: 480 seconds]
12:58millenialhacker has joined #asahi-dev
12:59MajorBiscuit has joined #asahi-dev
13:06millenialhacker has quit [Ping timeout: 480 seconds]
13:30amarioguychadmed: gotcha that makes sense
13:39millenialhacker has joined #asahi-dev
13:40Gaspare has joined #asahi-dev
13:41walid has quit [Ping timeout: 480 seconds]
13:47millenialhacker has quit [Ping timeout: 480 seconds]
13:58MajorBiscuit has quit [Quit: WeeChat 3.5]
14:03Gaspare has quit [Ping timeout: 480 seconds]
14:27the_lanetly_052 has joined #asahi-dev
14:27mps has quit [Remote host closed the connection]
14:27mps has joined #asahi-dev
14:38mps_ has joined #asahi-dev
14:39mps has quit [Ping timeout: 480 seconds]
15:13millenialhacker has joined #asahi-dev
15:14zzywysm_ has joined #asahi-dev
15:20Gaspare has joined #asahi-dev
15:21zzywysm has quit [Ping timeout: 480 seconds]
15:22millenialhacker has quit [Ping timeout: 480 seconds]
15:23zzywysm_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
15:38zzywysm has joined #asahi-dev
15:53millenialhacker has joined #asahi-dev
15:54sirn has quit [Quit: ZNC 1.8.2 - https://znc.in]
15:57sirn has joined #asahi-dev
16:02Gaspare has quit [Quit: Gaspare]
16:07zzywysm has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
16:10millenialhacker has quit [Read error: Connection timed out]
16:12amarioguy has quit [Ping timeout: 480 seconds]
16:12jluthra has quit [Remote host closed the connection]
16:13jluthra has joined #asahi-dev
16:22zzywysm has joined #asahi-dev
16:37svenlooks like the bt function can be reset just like the WiFi one in the brcm Android driver. It just doesn’t come back up afterwards yet :(
17:11millenialhacker has joined #asahi-dev
17:15Gaspare has joined #asahi-dev
17:19millenialhacker has quit [Ping timeout: 480 seconds]
17:21millenialhacker has joined #asahi-dev
17:28jannausven: apple magic keyboard/mouse work over bluetooth, completes the imac support. not sure how many people buy apple keyboards/mice with mac mini/studio
17:29millenialhacker has quit [Ping timeout: 480 seconds]
17:43kettenisI don't think I'll implement support for bluetooth in u-boot ;)
17:54jannaureconnecting whenever used (how without keyboard) would probably annoying anyway
17:55jannauluckily the keyboard is useable as usb keyboard when the charging cable is connected
17:56jannaunot designed by the person who put the charging connector for the mouse on the bottom
17:58kettenisjannau: ah cool, I didn't realize that
18:02Catyre has joined #asahi-dev
18:11Catyre has quit [Remote host closed the connection]
18:11Catyre has joined #asahi-dev
18:11Catyre has quit [Remote host closed the connection]
18:29doggkruse has joined #asahi-dev
18:37doggkruse has quit [Ping timeout: 480 seconds]
18:56millenialhacker has joined #asahi-dev
18:57svenjannau: nice!
18:58sveni didn't even realize that this would "complete" imac support :D
18:59ketteniswell, there still is the pcie su3 controller issue ;)
19:00sveni only consider bluetooth done anyway once it's in torvald's tree ;)
19:00svenand then there's still ATC/thunderbolt...
19:05millenialhacker has quit [Ping timeout: 480 seconds]
19:09jannauhah, I wish. "complete" in the sense that there is a useable system without extra parts/how a mac os user would use the system
19:26nicolas17complete? so we have GPU now? ;) *hides*
20:26Gaspare has quit [Quit: Gaspare]
20:30mps_ is now known as mps
20:32doggkruse has joined #asahi-dev
20:34Gaspare has joined #asahi-dev
20:40doggkruse has quit [Ping timeout: 480 seconds]
20:40nicolas17 has quit [Ping timeout: 480 seconds]
21:18nicolas17 has joined #asahi-dev
21:19nicolas17 has quit []
21:19nicolas17 has joined #asahi-dev
21:20Gaspare has quit [Quit: Gaspare]
21:28Gaspare has joined #asahi-dev
22:14Gaspare has quit [Quit: Gaspare]
22:32zzywysm has quit [Quit: Textual IRC Client: www.textualapp.com]
22:32zzywysm has joined #asahi-dev
22:49the_lanetly_052__ has joined #asahi-dev
22:54nabby^ has joined #asahi-dev
22:55the_lanetly_052 has quit [Ping timeout: 480 seconds]
23:08psykose has quit [Remote host closed the connection]
23:09psykose has joined #asahi-dev
23:12nabby^ has quit [Ping timeout: 480 seconds]
23:14___nick___ has joined #asahi-dev
23:15___nick___ has quit []
23:17___nick___ has joined #asahi-dev
23:19teach3e[m]@marcan I was reviewing the recent M2 streams trying debug something I'm running into booting a guest and noticed a potential issue. Here is the point in the stream: https://youtu.be/SidIJkC5YN0?t=7051. You copy the general PCORE chicken bits, which includes reg_set(SYS_IMP_APL_HID11, HID11_DISABLE_LD_NT_WIDGET). But after some edits it gets pushed below the line of new vs old code and then deleted here (see lines 37,54):
23:32nabby^ has joined #asahi-dev
23:32nabby^ has quit [Remote host closed the connection]
23:41nehsou^ has joined #asahi-dev
23:54hays has joined #asahi-dev
…