#
…

2023-04-13

01:21A_L_I_C_E has quit [Remote host closed the connection]
01:33A_L_I_C_E has joined #asahi-dev
01:42hightower3 has joined #asahi-dev
01:49hightower2 has quit [Ping timeout: 480 seconds]
02:00gabuscus has quit []
02:16gabuscus has joined #asahi-dev
02:45newiz has joined #asahi-dev
03:50cylm has joined #asahi-dev
05:53kedde has quit [Ping timeout: 480 seconds]
06:17marcanjannau: Argh, I already had the installer changes locally, just didn't want to push to stop people from doing silly things..
06:36chadmed_ has quit [Remote host closed the connection]
06:37nyilas has joined #asahi-dev
06:37Hibyehello has joined #asahi-dev
06:38kedde has joined #asahi-dev
06:39Hibyehello_ has quit [Ping timeout: 480 seconds]
06:40newiz has quit [Quit: Konversation terminated!]
07:04MajorBiscuit has joined #asahi-dev
07:09mazeric_engestrom: you need to select the *PMU* that matches the core you are using.
07:10mazthe default is probably the PMU type that matches cores 0-3, which leads to nothing on core 4...
07:17gladiac has joined #asahi-dev
07:17jannaumarcan: np, changes are not complicated and I had them most of them since the m2 mac mini arrived
07:20nyilas has quit [Remote host closed the connection]
07:20jannaudid you see https://github.com/AsahiLinux/linux/pull/134 ? the ps_isp_sys change is maybe worth looking into further
07:21jannauwith the pcie fix the bringup branch works on m1/m1 max/m2 with dcp firmware 12.3 and 13.2
07:23nyilas has joined #asahi-dev
07:25Nae has joined #asahi-dev
07:30jannaumaz: would it make sense to support something like https://lore.kernel.org/lkml/YE+avkajxbGJMOoM@krava/T/ ? that looks to me like more intuitive default behavior
07:31cylm has quit [Ping timeout: 480 seconds]
07:40MajorBiscuit has quit [Quit: WeeChat 3.6]
07:48bps has joined #asahi-dev
07:50mazjannau: can't see why not, given that this a pure userspace approach. I'm just worried that it will lead to "interesting" results on implementations that have events numbers that count vastly different things on different micro architectures...
07:51mazfor the Apple stuff, it's probably OK. but the ARM implementations are... insane.
07:56jannaushouldn't that be still ok as long as perf doesn't aggregate counters from different PMUs? might be impossible to interpret the results but shouldn't be worse than not counting anything due to running on the wrong CPU
08:02jn has quit [Read error: No route to host]
08:03jn has joined #asahi-dev
08:03mazjannau: if we don't get aggregation, it should be OK indeed.
08:28kaprests has quit [Ping timeout: 480 seconds]
08:31kaprests has joined #asahi-dev
08:38bpye has quit [Ping timeout: 480 seconds]
08:47bpye has joined #asahi-dev
08:51nsklaus has joined #asahi-dev
09:13kettenishmm, I'm probably at the point where I want to do a "proper" uefi install on my m2 pro mini
09:14kettenisor is it easy to produce the firmware init ramdisk cpio by running one of the python scripts on macOS?
09:16kettenisin any case I just pushed the M2 Pro/Max U-Boot support to the asahi branch
09:17jannautest.sh should work locally on macos
09:19jannauthat will install a broken m1n1 +dtb + u-boot boot.bin but I assume you're able update it
09:20chadmedi shoehorned 13.2 support into the installer for the m2 pro mini a while ago but its horrid and uses a stale branch of m1n1
09:20chadmedworks though!
09:21kettenisI did something similar, but initailly just did the proxy install
09:22kettenisbut I should be able to upload the cpio using the proxy shouldn't I?
09:44Nae has quit [Quit: Connection closed for inactivity]
09:47cylm has joined #asahi-dev
09:48bpye has quit [Ping timeout: 480 seconds]
09:48bps has quit [Ping timeout: 480 seconds]
09:54cylm_ has joined #asahi-dev
09:56cylm has quit [Ping timeout: 480 seconds]
10:06bpye has joined #asahi-dev
10:08DarkShadow44 has quit [Quit: ZNC - https://znc.in]
10:08DarkShadow44 has joined #asahi-dev
10:10Ariadne has quit [Remote host closed the connection]
10:11Ariadne has joined #asahi-dev
10:20stipa is now known as Guest10828
10:20stipa has joined #asahi-dev
10:22Guest10828 has quit [Read error: Connection reset by peer]
10:25robinp has quit [Quit: Connection closed for inactivity]
10:34chadmed has quit [Ping timeout: 480 seconds]
10:37chadmed has joined #asahi-dev
10:58bps has joined #asahi-dev
11:17bcrumb has joined #asahi-dev
11:30bcrumb has quit [Quit: WeeChat 3.8]
11:31linuxgemini1 has joined #asahi-dev
11:33cynthia has quit [Remote host closed the connection]
11:34cynthia has joined #asahi-dev
11:35linuxgemini has quit [Ping timeout: 480 seconds]
11:43cynthia has quit [Ping timeout: 480 seconds]
11:53cynthia has joined #asahi-dev
12:17marcanlet me just push the installer to dev, no point in making everyone's life difficult with this many people interested
12:24mazmarcan: now you're making it way too tempting for me to go and grab an M2 pro...
12:24marcanpushed alx.sh/dev, you probably want to use 13.3 firmware since it sounds like that's going to be the new target
12:25marcanobviously it will install a broken u-boot/kernel if you do a normal install, but good enough to get a rootfs on there and then develop in tethered mode
12:25marcanplain m1n1 should work though, and I assume everyone here knows how to install custom m1n1/uboot builds :p
12:26jannaumaz: your use case (serial, ethernet, usb3) is already supported :)
12:26marcanstill no display out, mine's still in a box pending that (it's going to be a server but I don't want to put it into prod until it's officially supported...)
12:27chadmedmaz: the thing is stupid fast and everything you need already works
12:27mazstop it, all of you! :D
12:28kettenismeanwhile, I have successfully ported the pcie changes to u-boot
12:28kettenismarcan: found a minor issue in your driver while doing so
12:28marcanmy drivers, broken? never!
12:29marcan(still no idea how MTP works on M2...)
12:29kettenisthe second phy_lane_ctl poke still uses the hardcoded register number
12:30cynthia has quit [Ping timeout: 480 seconds]
12:30kettenisso, no, not that one jannau ;)
12:31kettenis(I fixed that one without realizing it was broken)
12:32marcanjannau: grabbed that one
12:32marcankettenis: that one's just a bug on paper since there is only one nonzero value, but fixed :)
12:33kettenishey I didn't call it a bug ;)
12:36jannaumarcan: there are still devicetree fixes in https://github.com/AsahiLinux/linux/pull/134
12:37cynthia has joined #asahi-dev
12:37jannauthe isp power-domain on j474 might justify a second look
12:41marcanjannau: looks reasonable to me
12:46minecrell has quit [Read error: Connection timed out]
12:46kettenisI made a start looking at bcm4388; it seems the bcm4389 android driver has some bcm4388 bits in it
12:46minecrell has joined #asahi-dev
12:48kettenisI'm probably going to concentrate on asmedia fwload support in u-boot before pursuing that further though
12:49marcanI would hope the wifi is mostly just another chipid/otp offset/tcm offset
12:49marcanif I end up having to implement more firmware APIs for it... :/
12:50marcan(at least for feature parity, wifi 6 who knows, we don't even have full WPA3 support yet)
12:50kettenisdefinitely a different otp offset
12:50marcanI never heard back from the broadcom guy about whether my WPA3 SAE patch worked on anything non-apple...
12:51marcanI know it doesn't work on some raspberry pi thing, but that could just be firmware
12:51marcanmeanwhile still not a single email from cypress
12:52kettenisI found the 0x15 "system vendor" tag, but didn't manage to work my way back to the start of the OTP area
14:03j`ey_ is now known as j`ey
14:18kedde has quit [Ping timeout: 480 seconds]
14:53gladiac has quit [Quit: k thx bye]
15:18leviz has joined #asahi-dev
15:21leviz has quit []
15:21leviz has joined #asahi-dev
15:22leviz has quit []
15:34gladiac has joined #asahi-dev
16:00cylm_ has quit [Quit: WeeChat 3.6]
16:05jn has quit [Remote host closed the connection]
16:06jn has joined #asahi-dev
16:12MajorBiscuit has joined #asahi-dev
16:13MajorBiscuit has quit []
16:13MajorBiscuit has joined #asahi-dev
16:27MajorBiscuit has quit [Ping timeout: 480 seconds]
16:30bps has quit [Ping timeout: 480 seconds]
16:31raveling has quit [Ping timeout: 480 seconds]
16:53raveling has joined #asahi-dev
17:05raveling has quit [Ping timeout: 480 seconds]
17:11A_L_I_C_E has quit [Remote host closed the connection]
17:12raveling has joined #asahi-dev
17:16MajorBiscuit has joined #asahi-dev
17:23lawrence8 has joined #asahi-dev
17:23raveling has quit [Ping timeout: 480 seconds]
17:26A_L_I_C_E has joined #asahi-dev
17:29lawrence has quit [Ping timeout: 480 seconds]
17:29lawrence8 is now known as lawrence
17:29A_L_I_C_E has quit [Read error: Connection reset by peer]
17:30A_L_I_C_E has joined #asahi-dev
17:31raveling has joined #asahi-dev
17:39raveling has quit [Ping timeout: 480 seconds]
17:39leviz has joined #asahi-dev
17:39leviz has quit []
17:47A_L_I_C_E has quit [Quit: Quit]
17:47A_L_I_C_E has joined #asahi-dev
17:49raveling has joined #asahi-dev
17:57raveling has quit [Ping timeout: 480 seconds]
17:59kettenis has quit [Ping timeout: 480 seconds]
18:02kettenis has joined #asahi-dev
18:06raveling has joined #asahi-dev
18:08avi has joined #asahi-dev
18:09avi has quit []
18:11MajorBiscuit has quit [Quit: WeeChat 3.6]
18:14raveling has quit [Ping timeout: 480 seconds]
18:24A_L_I_C_E has quit [Read error: Connection reset by peer]
18:25A_L_I_C_E has joined #asahi-dev
18:42pthariensflame has joined #asahi-dev
18:45gladiac has quit [Quit: k thx bye]
19:00pthariensflame has quit [Quit: Textual IRC Client: www.textualapp.com]
19:03ChaosPrincessso, whats the deal with devices having two darts
19:03ChaosPrincessdispdfr has two, the adt shows reg = [Container:addr = 0x0000000028304000 size = 0x0000000000004000, Container:addr = 0x0000000028300000 size = 0x0000000000004000]
19:04ChaosPrincessi did dart = DART.from_adt(u, "arm-io/dart-dispdfr") in m1n1, and then iova = dart.iomap(0, buf, fb_size)
19:04ChaosPrincesswrote iova to reg, everything works
19:04ChaosPrincesson linux i have in adt iommus = <&displaydfr_dart_1 0>; it has reg = <0x2 0x28304000 0x0 0x4000>;
19:05ChaosPrincesswriting iova to register, getting a translation fault from dart at 228300000
19:05ChaosPrincesswhat am i doing wrong
19:15raveling has joined #asahi-dev
19:15jannauwe usually ignore the second instance. except for dwc3. you did that implicitly with the python code
19:19jannaudef from_adt(cls, u, path, instance=0, **kwargs):
19:20jannau def from_adt(cls, u, path, instance=0, **kwargs):
19:20jannau dart_addr = u.adt[path].get_reg(instance)[0]
19:25ChaosPrincessYea i saw that python uses the first instance, the problem is that i am getting translation faults from second one
19:27jannaueither don't add the second instance to the dt or if that doesn't work add both iommus to displaydfr, cmp. with dwc3
19:35A_L_I_C_E has quit [Read error: Connection reset by peer]
19:41A_L_I_C_E has joined #asahi-dev
19:45A_L_I_C_E has quit [Remote host closed the connection]
19:46svenif you get translation faults on the second dart something tried to dma through it.
19:46sventhat usually only happens when that dart is actually used
19:46svendo you recognize the fault address?
19:47sventhose two darts are very suspicious because they’re just next to each other
19:47svenare you sure they are darts and not dart plus what apple calls SMMU or something like that?
19:49svenLook at the dart node, there’s an instance property
19:50svenfor that dispdfr thingy it claims dart and smmu
19:50svenso that second thing is that apple SMMU (not to be confused with arm‘s architectural SMMU)
19:50svenjust ignore it
19:50jannauwe use only instance 0 for dart-disp0/dispext*
19:52sventhat also seems to be dart plus smmu
19:53svenjust look at that instance property in the ADT. It’d a bit weird but if that thing contains the SMMU fourcc and the two regs of the dart are separated by 0x4000 the second thing isn’t a DART
19:54svenI never figured out exactly what it is but it seems to expose the most recently accessed pagetable entries inside it’s MMIO
19:54svenSo it’s somehow related to the dart itself but no idea what exactly it does or why it’s separate
19:54svencould just be a dart debug interface that’s not synthesized for each dart
20:41bps has joined #asahi-dev
20:45kettenisthe model names for the M2 and M2 Pro mini in the dev installer are wrong
20:45kettenismanaged to work around it
21:08nyilas has quit [Remote host closed the connection]
21:41nsklaus has quit [Quit: ZZZzzz…]
21:46bluetail42 has joined #asahi-dev
21:49kedde has joined #asahi-dev
21:53bluetail4 has quit [Ping timeout: 480 seconds]
21:54raveling has quit [Ping timeout: 480 seconds]
22:14A_L_I_C_E has joined #asahi-dev
22:17raveling has joined #asahi-dev
22:17pharonix71 has joined #asahi-dev
22:18bps has quit [Remote host closed the connection]
22:22jannauis there anything mtp related which could be global system fw dependent? it fails here with:
22:22jannau24e400000.mtp: RTKit buffer request outside SRAM region: [mem 0x24eca8000-0x24eca8fff]
22:23jannaubut I think the only thing I changed since yesterday is the 13.2 -> 13.3 macOS update
22:25raveling has quit [Ping timeout: 480 seconds]
22:32jannaunot global firmware. still works when booted through the 12.3 stub. It's broken when booted via the hypervisor from the macOS 13.3 install
22:35raveling has joined #asahi-dev
22:43raveling has quit [Ping timeout: 480 seconds]
22:49chadmed_ has joined #asahi-dev
22:55raveling has joined #asahi-dev
23:04raveling has quit [Ping timeout: 480 seconds]
23:21A_L_I_C_E has quit [Remote host closed the connection]
23:52amw has quit [Ping timeout: 480 seconds]
…