#
…

2023-10-06

00:14dcow has joined #asahi-dev
00:22dcow has quit [Ping timeout: 480 seconds]
00:55dcow has joined #asahi-dev
01:06dcow has quit [Ping timeout: 480 seconds]
01:10eiln has quit [Ping timeout: 480 seconds]
01:34dcow has joined #asahi-dev
01:42dcow has quit [Ping timeout: 480 seconds]
01:55amarioguy has quit [Remote host closed the connection]
02:00gabuscus has quit []
02:08dcow has joined #asahi-dev
02:16dcow has quit [Ping timeout: 480 seconds]
02:38gabuscus has joined #asahi-dev
02:40dcow has joined #asahi-dev
02:41crabbedhaloablut has joined #asahi-dev
02:48dcow has quit [Ping timeout: 480 seconds]
02:58dcow has joined #asahi-dev
03:01skipwich has quit [Quit: DISCONNECT]
03:02skipwich has joined #asahi-dev
03:08dcow has quit [Ping timeout: 480 seconds]
03:19jeisom has quit [Quit: Leaving]
03:47tristan2 has joined #asahi-dev
03:49tristan2_ has quit [Ping timeout: 480 seconds]
05:47systwi has quit [Ping timeout: 480 seconds]
06:30dcow has joined #asahi-dev
06:38dcow has quit [Ping timeout: 480 seconds]
06:43slp_ has joined #asahi-dev
06:49jannegleio: https://github.com/AsahiLinux/m1n1/pull/338 makes 1st stage display init reliable for me
06:52tbmjanneg: it says 25 but the mdelay is 20
06:53jannegtbm: forgot to git add, fixed. thanks
06:54kaazoo has joined #asahi-dev
06:59jannegeiln: calibration upload works with the 13.5 isp firmware so this seems to be a 12.4 incompatibility
07:01dcow has joined #asahi-dev
07:02jannegthe 720p sensor is even with calibration data very noisy. could that be caused by a wrong config?
07:05leiojanneg: thanks, I'll test it out when I get around to m1n1 updates overall, hopefully weekend
07:05kaazooHI, in case anybody here uses ZFS: developers added support for BTI (Branch Target Identification) which is set via CONFIG_ARM64_BTI: https://github.com/openzfs/zfs/pull/15339
07:05kaazooEarlier, ZFS kernel module loading resulted in internal error: https://github.com/openzfs/zfs/issues/14862
07:08Yannik_Sc has joined #asahi-dev
07:09leioon that topic, I'd be curious what packages overall need BTI disabling, besides chromium, to include these in ::gentoo out of the box
07:09dcow has quit [Ping timeout: 480 seconds]
07:30dcow has joined #asahi-dev
07:40dcow has quit [Ping timeout: 480 seconds]
07:44Yannik_Sc has quit [Ping timeout: 480 seconds]
07:46chadmednothing i use regularly so far
07:46chadmedthe fix for chromium is simple: stop using it
07:48leioI saw it done in webkitgtk fedora package too iirc, but I have other issues with it
07:50dcow has joined #asahi-dev
07:57Yannik_Sc has joined #asahi-dev
08:02tenkuu has quit [Ping timeout: 480 seconds]
08:05dcow has quit [Ping timeout: 480 seconds]
08:07leitao has joined #asahi-dev
08:08leitao has quit []
08:20jannegI doubt you've configured toolchain/portage configured to use BTI in the first place contrary to what fedora does
08:21chadmedleio took my make.conf which does explicitly enable -mbranch-protection=standard
08:27leitao has joined #asahi-dev
08:30Yannik_Sc has quit [Ping timeout: 480 seconds]
08:31leiochadmed: note that strip-flags used by brittle packages used to filter that out, but since yesterday it doesn't
08:36leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
08:36dcow has joined #asahi-dev
08:40eiln has joined #asahi-dev
08:43eilnsounds like m2 SET_FILE_LOAD abi changed between 12.4 & 13.5? maybe that's where iova became u32 -> u64? if so we'll need the dcp os version checks
08:43eilnjanneg: if it's "noisy" as in grainy it's very likely a temporal/motion filter issue (constants are off, graph order isn't right). lina should have an m2 fork of the tracer, which I do not see in any branch
08:43eilnchadmed: agreed
08:44dcow has quit [Ping timeout: 480 seconds]
08:46leiosomeone needs to tell slack to add huddle support when used with firefox.
08:49psykoseapparently it's only disabled on firefox+linux
08:53leioright, time to user agent spoof then :)
08:53leio(which is bad in way, as they'll think they have no users on firefox+linux or something); anyhow, getting off-topic
08:56leitao has joined #asahi-dev
09:02cy8aerI just played around with the camera and want to present my results. first of all: The device works. ffplay/mpv /dev/video0 works for my desktop user. The problems I have now: Firefox says the device does not have the right resolution when I open a jitsi session. And cheese only works if run as root. Permissions of the camera device is the same like the permissions of a standard logitech camera: crw-rw----+ 1 root video, the desktop user belongs to...
09:02cy8aer... the video group. Are there any hints or is there something still missing in the kernel?
09:03cy8aermbp14/m1pro
09:04linaeiln: Good thing I added the version to the m1n1 side ^^
09:04linaShould be easy to fix then!
09:05linaAnd that makes sense since 12.4 didn't support any machines where high IOVAs are used
09:05bgb has quit [Quit: WeeChat 4.0.5]
09:06linacy8aer: We support a lot of resolutions on those machines, including vertical ones... maybe Firefox/jitsi is unhappy with that?
09:06linaCan you try some webcam test sites?
09:06cy8aerlina ok, I'll do some webcam test sites...
09:06linaDepending on what Firefox is doing exactly, we might be able to change the logic in the driver to prefer some resolutions, or hide the weird vertical ones behind a module argument if they cause too much trouble
09:07linaI know for one ffmpeg is broken with vertical resolutions (I have a patch for that)
09:07linaGStreamer works though
09:07mpscy8aer: guvcview have few resolutions to test
09:08mpsI tested some last night
09:08linaIf too much stuff is broken with extra resolutions we can just limit the 1080p cameras to 1920x1080p only for now, that's a simple device tree change.
09:09linaAlso I didn't technically test any of this on m1pro/max, I just assumed the resolutions are the same as on m2pro/max since they use the same sensor... but it could be wrong.
09:09lina1080p should definitely be correct though
09:09cy8aerI now tried de.webcamtests.com and that works fine Resolution 1920x1080 FHD. There the browser does not make problems. I will now play around with guvcview
09:11systwi has joined #asahi-dev
09:11dcow has joined #asahi-dev
09:14leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
09:19leitao has joined #asahi-dev
09:21dcow has quit [Ping timeout: 480 seconds]
09:27chadmed has quit [Remote host closed the connection]
09:33chadmed has joined #asahi-dev
09:34cy8aerOk, guvcview works with every tested resolution. But with jitsi I do not know what there is happening. I tried it with chromium and there the video is shown in diagonal stripes. Probably they want 720p
09:35cy8aer(found that somewhere in the page source)
09:36cy8aer1280x720 is not offered in guvcview
09:43cy8aerOk, I got jitsi running, but: I am hosting my own jitsi server. And by default there is defined: prefered 720p, min 240p, max 720p! I set max resolution to 1080p and it worked. This may be a solution for me, but because max is 720p by default there should be 720p offered somehow.
09:45Yannik_Sc has joined #asahi-dev
09:47dcow has joined #asahi-dev
09:51Yannik_Sc has quit [Remote host closed the connection]
09:55Yannik_Sc has joined #asahi-dev
09:56dcow has quit [Ping timeout: 480 seconds]
09:58jannegeiln: yes, very grainy
10:01janneglina: can you push the m2 tracer to a branch? I have a partial port with fixed mailbox/doorbell offsets but that wasn't enough
10:02jannegdo we know if the imac uses the same sensor as the m1 pro/max macbooks? it has a 1080p sensor
10:10Yannik_Sc has quit [Ping timeout: 480 seconds]
10:19dcow has joined #asahi-dev
10:28dcow has quit [Ping timeout: 480 seconds]
10:29eilnjanneg: I'd be really surprised if it wasn't. do any of us have an imac to test? I've never even seen one in the wild
10:41jannegmarcan has or has access to one and I can get access to one as well. I can't remember seeing an user with an imac. the telemetry might tell us how many users with imacs we have
10:54dcow has joined #asahi-dev
11:00bgb has joined #asahi-dev
11:02dcow has quit [Ping timeout: 480 seconds]
11:06Yannik_Sc has joined #asahi-dev
11:12eilndidn't the imac have some weirdo audio codec? on second thought it might be a different sensor
11:13chadmedeiln: both
11:13chadmedi think we have a driver for the codec now though
11:27leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
11:28pjakobsson has joined #asahi-dev
11:33dcow has joined #asahi-dev
11:37tirr has quit [Quit: ZNC 1.8.2 - https://znc.in]
11:37tirr has joined #asahi-dev
11:41dcow has quit [Ping timeout: 480 seconds]
12:08dcow has joined #asahi-dev
12:09linapushed the tracer/etc changes to isp/t602x-b
12:09linajanneg: ^
12:11jjanzic has quit [Ping timeout: 480 seconds]
12:12leitao has joined #asahi-dev
12:14Yannik_Sc has quit [Remote host closed the connection]
12:20jjanzic has joined #asahi-dev
12:23dcow has quit [Ping timeout: 480 seconds]
12:27sefidel has quit [Remote host closed the connection]
12:28sefidel has joined #asahi-dev
12:35midou has joined #asahi-dev
12:49DannyB has joined #asahi-dev
12:52hightower3 has quit [Ping timeout: 480 seconds]
12:53dcow has joined #asahi-dev
12:59cuiltb^ has quit [Ping timeout: 480 seconds]
13:05dcow has quit [Ping timeout: 480 seconds]
13:06leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
13:13leitao has joined #asahi-dev
13:23AnuthaDev has joined #asahi-dev
13:24jeisom has joined #asahi-dev
13:29leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
13:31dcow has joined #asahi-dev
13:32leitao has joined #asahi-dev
13:37leitao has quit []
13:39leitao has joined #asahi-dev
13:46dcow has quit [Ping timeout: 480 seconds]
13:58hex-spell has joined #asahi-dev
14:01slp_ is now known as slp
14:10dcow has joined #asahi-dev
14:23dcow has quit [Ping timeout: 480 seconds]
14:46dcow has joined #asahi-dev
14:54dcow has quit [Ping timeout: 480 seconds]
14:58cylm has joined #asahi-dev
15:13cuiltb^ has joined #asahi-dev
15:17sam_[09:20:07] <janneg> I doubt you've configured toolchain/portage configured to use BTI in the first place contrary to what fedora does
15:17sam_i'd be surprised if fedora is really doing that much magic to do it
15:17sam_if there's much we need to do, please let me know
15:17sam_but what leio seemed to do from chadmed's make.conf seemed sufficient and t otake effect
15:20dcow has joined #asahi-dev
15:22leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
15:24leitao has joined #asahi-dev
15:28dcow has quit [Ping timeout: 480 seconds]
15:30cuiltb^ has quit [Ping timeout: 480 seconds]
15:30cuiltb^ has joined #asahi-dev
15:42lena6 has joined #asahi-dev
15:45DannyB has quit [Ping timeout: 480 seconds]
15:52jacksonchen666 has joined #asahi-dev
15:52dcow has joined #asahi-dev
15:58jacksonchen666 is now known as Guest2389
15:58jacksonchen666 has joined #asahi-dev
15:59leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
15:59tenkuu has joined #asahi-dev
16:03Guest2389 has quit [Ping timeout: 480 seconds]
16:03leitao has joined #asahi-dev
16:04dcow has quit [Ping timeout: 480 seconds]
16:08cuiltb^ has quit [Read error: Connection reset by peer]
16:08cuiltb^ has joined #asahi-dev
16:18leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
16:22dcow has joined #asahi-dev
16:25leitao has joined #asahi-dev
16:30dcow has quit [Ping timeout: 480 seconds]
16:36nela has quit [Ping timeout: 480 seconds]
16:52dcow has joined #asahi-dev
16:58leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
16:59compassion1785 has quit [Quit: lounge quit]
17:00dcow has quit [Ping timeout: 480 seconds]
17:07rootbeerdan has joined #asahi-dev
17:08dcow has joined #asahi-dev
17:15compassion1785 has joined #asahi-dev
17:20dcow has quit [Ping timeout: 480 seconds]
17:44kettenisnote that chromium works fine on OpenBSD with BTI
17:45kettenisso the problem is either in Linux-specific code or in some system library
17:46sam_oh interesting, thanks
17:48dcow has joined #asahi-dev
17:51nela has joined #asahi-dev
17:52mpsit works also on alpine with BTI enabled in kernel
17:53psykosethe actual chromium build doesn't have bti on chromium
17:53psykoselast i checked anyway
17:54mpsah, I didn't looked, I don't use chromium normaly, just sometimes if site doesn't work with firefox
18:12dcow has quit [Ping timeout: 480 seconds]
18:13dcow has joined #asahi-dev
18:22dcow has quit [Ping timeout: 480 seconds]
18:31dcow has joined #asahi-dev
18:34leiothis is with -mbranch-protection=standard passed to chromium build via CXXFLAGS.
18:35psykoseyou also want CFLAGS
18:35psykosetho i thought bti is =bti
18:35psykoseanyway, offtopic
18:39dcow has quit [Ping timeout: 480 seconds]
18:48dcow has joined #asahi-dev
18:50Amber_Harmonia has joined #asahi-dev
18:52flokli has quit [Quit: WeeChat 4.0.2]
18:52flokli has joined #asahi-dev
18:58cyrinux has quit [Quit: bye]
18:59cyrinux has joined #asahi-dev
18:59tenkuu has quit [Read error: No route to host]
18:59tenkuu has joined #asahi-dev
19:01tenkuu_ has joined #asahi-dev
19:05tenkuu has quit [Read error: Connection reset by peer]
19:07jannegm2 with 12.4 firmware is fixed by using the 32-bit CISP_CMD_CH_SET_FILE_LOAD command
19:09dcow has quit [Ping timeout: 480 seconds]
19:12dcow has joined #asahi-dev
19:17PaulFertser_ is now known as PaulFertser
19:20dcow has quit [Ping timeout: 480 seconds]
19:26dcow has joined #asahi-dev
19:38dcow has quit [Ping timeout: 480 seconds]
20:04luix has joined #asahi-dev
20:05maximbaz has quit [Quit: bye]
20:05midou has quit [Ping timeout: 480 seconds]
20:07maximbaz has joined #asahi-dev
20:11midou has joined #asahi-dev
20:16luix has quit [Remote host closed the connection]
20:21dcow has joined #asahi-dev
20:23jannegis it intentional that the webcam image is not mirrored by default? or are the applications supposed to handle this?
20:24jannegeiln: image is better with the same temporal filter settings as macos, i.e. '.temporal_filter = 0,'
20:25janneglina: does macos use 1 on the m2 mb air?
20:26kettenisyay, managed to get my m2 pro mini running again with 13.5 OS firmware
20:27kettenisnew m1n1 mostly fails to detect my monitor properly; but I've had issues with it before
20:27kettenisI think it doesn't support the standard VESA mode
20:29jannegkettenis: uncommenting '#define RTKIT_SYSLOG' in config.h makes dcp chatty
20:29dcow has quit [Ping timeout: 480 seconds]
20:31kettenisthanks, I'll give that a go later this weekend
20:33kettenisit took me far too much time I didn't get any kernel output because I left the usb keyboard plugged in...
20:35jeisom has quit [Quit: Leaving]
20:40dcow has joined #asahi-dev
20:46hightower2 has joined #asahi-dev
20:48dcow has quit [Ping timeout: 480 seconds]
21:15dcow has joined #asahi-dev
21:17ChaosPrincess has quit [Quit: ChaosPrincess]
21:25dcow has quit [Ping timeout: 480 seconds]
21:26ChaosPri1 has joined #asahi-dev
21:27ChaosPri1 has quit []
21:27ChaosPrincess has joined #asahi-dev
21:49dcow has joined #asahi-dev
21:57dcow has quit [Ping timeout: 480 seconds]
22:05dcow has joined #asahi-dev
22:13dcow has quit [Ping timeout: 480 seconds]
22:26eilnjanneg: my other machines aren't mirrored either. on macos the preview is mirrored by software but the actual capture isn't (which is weird)
22:34eilnv4l2 loopback with -vflip is always a solution but
22:34jannegI hope you mean by software either the scaler or the gpu
22:35jannegfor macos
22:36eilnI don't recall ever seeing logs about scaler/gpu in Core Media. let me quickly check scaler
22:44eilnjanneg: looks like a memory copy with scaler. we could probably do m2m by making up our own DART endpoint
22:46janneghere is the fw version check. much code for a single check: https://github.com/AsahiLinux/linux/pull/203
22:46eilnohh it uses scaler for mirroring + window resizing. that makes sense why only the preview is mirrored
22:46mpsffplay -vf hflip /dev/video0
22:53crabbedhaloablut has quit []
23:17eilnok I think my certificate has worked out
23:17eilnavd notes for those interested https://eiln.net/avd-notes.html
23:25DannyB has joined #asahi-dev
23:28dcow has joined #asahi-dev
23:36dcow has quit [Ping timeout: 480 seconds]
23:45AnuthaDev has quit []
23:50tenkuu_ has quit [Ping timeout: 480 seconds]
23:52tenkuu has joined #asahi-dev
23:52DannyB has quit [Ping timeout: 480 seconds]
…