2026-05-26.log

minutelidstah: FWIW i do all soldering through a stereo microscope00:22
minutejosch: hmm right00:22
- mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org)01:54
- voltaire28_ (QUIT: Quit: WeeChat 4.6.3) (~jlafon@28.162.2.93.rev.sfr.net)02:20
- lidstah (QUIT: Remote host closed the connection) (~lidstah@gateway/tor-sasl/lidstah)02:42
+ lidstah (~lidstah@gateway/tor-sasl/lidstah)02:42
- RandyK (QUIT: Remote host closed the connection) (~RandyK@user/randyk)02:59
+ RandyK (~RandyK@user/randyk)02:59
- qbit (QUIT: Remote host closed the connection) (~qbit@user/qbit)03:31
+ qbit (~qbit@user/qbit)03:33
- qbit (QUIT: Remote host closed the connection) (~qbit@user/qbit)03:45
+ qbit (~qbit@user/qbit)03:53
+ cow321 (~deflated8@user/meow/deflated8837)03:56
- kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de)03:57
+ kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de)03:57
- kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de)04:08
+ kop316 (m-6f6zq6@157.90.159.138)04:08
- paperManu (QUIT: Ping timeout: 256 seconds) (~paperManu@204.244.197.237)04:14
- RandyK (QUIT: Remote host closed the connection) (~RandyK@user/randyk)04:44
+ RandyK (~RandyK@user/randyk)04:44
- frickler (QUIT: Ping timeout: 265 seconds) (~jens@user/frickler)05:04
- paperManu_ (QUIT: Ping timeout: 252 seconds) (~paperManu@104.163.176.82)05:15
+ potatoes_ (~quassel@130.44.145.181)06:00
- potatoes (QUIT: Ping timeout: 264 seconds) (~quassel@user/potatoes)06:01
+ frickler (~jens@user/frickler)07:44
- lanodan (QUIT: Quit: WeeChat 4.7.2) (~lanodan@2a01:e0a:d6:9930::35)08:07
+ lanodan (~lanodan@2a01:e0a:d6:9930::35)08:08
- xktr (QUIT: Ping timeout: 248 seconds) (~xktr@user/xktr)09:56
+ xktr (~xktr@user/xktr)09:57
- ephase (QUIT: Ping timeout: 248 seconds) (~ephase@82.66.198.11)10:00
+ mjw (~mjw@gnu.wildebeest.org)10:21
+ bkeys (~bkeys@98.19.131.29)10:44
bkeysminute I owe you money and want to pay it, what is the paypal address I need to send the money to?10:44
bkeys(or however you would like me to pay you, I know paypal takes fees from you so if there is a better way to send for you I will do that also)10:45
Svptried the foolish task of switching debian to openrc, which means i currently have to leave behind reform-tools, process was not at all painless but i did pull through for the most part11:59
- mjw (QUIT: Ping timeout: 259 seconds) (~mjw@gnu.wildebeest.org)12:02
joschSvp: in Debian i'm working together with werdahias who is working on making openrc work on gnu/hurd. If you are facing issues or have patches i can get you into contact12:05
joschSvp: also, if you have patches to make reform-tools work with non-systemd init systems, i'll happily apply them12:06
Svpi currently just started this endeavor so i'll have to first poke around12:07
joschi think that's cool! Feel free to keep me in the loop and tell me what reform-tools can do to support this use-case12:07
Svpbut by... tricking systemd into not doing its prerm routines i did manage to switch a running system. that and updating the initramfs but i did get somewhere12:08
Svpis the only thing holding back reform-tools the reform-hwsetup service or am i missing something else12:12
joschSvp: there is also reform-power-daemon.service and reform-sleep.service12:14
joschare you still using udev? otherwise there is also reform-ttymxc-wakeup.rules12:14
+ ephase (~ephase@82.66.198.11)12:14
Svpyeah i do still have udev right now12:14
* Guest4437 -> mjw12:17
+ paperManu (~paperManu@104.163.176.82)12:23
+ paperManu_ (~paperManu@204.244.197.237)12:23
Svpran dpkg --extract on the reform-tools deb and rebuilt initramfs, now i have sway and labwc work, neat. no sound yet tho, will think about it with the power of lunch12:24
joschSvp: sound requires setup by reform-hw-setup12:24
Svpprobably because pipewire is started through a user service though12:24
orvaDebian unstable has recent enough openrc that it has user services as well, no idea how well integraded those are in though12:32
orvaAt least Gentoo seems to bit underway migrating to those in some places12:33
- paperManu (QUIT: Ping timeout: 241 seconds) (~paperManu@104.163.176.82)14:25
- paperManu_ (QUIT: Ping timeout: 261 seconds) (~paperManu@204.244.197.237)14:27
- frickler (QUIT: Ping timeout: 252 seconds) (~jens@user/frickler)14:28
+ frickler (~jens@user/frickler)14:29
- digitalrane (QUIT: Ping timeout: 243 seconds) (~digitalra@user/digitalrane)14:33
+ paperManu (~paperManu@modemcable141.205-200-24.mc.videotron.ca)14:45
minutebkeys: i'll send you in DM14:55
minutejosch: to merge my qcom-7.0 PR, i think you said this commit shouldn't be merged (?) https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/182/diffs?commit_id=64e815367a103931d7385829becb0abadb552dea14:56
joschminute: yes, otherwise you roll out unreleased reform-tools to everybody14:57
minuteok i'll try to rebase this branch and drop that14:57
joschthank you!14:57
minutegonna try this syntax: git rebase -i 64e81536^14:57
joschminute: the memtest commit can also be dropped14:58
minuteright14:58
joschminute: i usually just do "git rebase -i main"14:58
minuteright14:58
minutei'll try that instead14:58
joschdoing that should also get rid of the two merge commits14:59
joschhuh gitlab says 73 commits now?15:02
joschminute: i don't think your main branch is the same as the main branch on source.mnt.re15:03
joschminute: i think you can fix this by first going back to your original state (git still has it if you didn't run a git cleanup) by doing "git checkout ef8ac0e9" because ef8ac0e9 was the original HEAD before you rebased (unless you had local changes you didn't push to source.mnt.re)15:12
joschminute: if you don't want to fiddle right now, this branch might have the result of the rebase that i think you wanted to do: https://source.mnt.re/josch/reform-debian-packages/-/commits/qcom-7.015:12
minutejosch: gitlab shows 8 commits15:21
minuteit was confused for a while15:21
joschah good :)15:27
joschminute: once it's merged i can copypaste the dtbs to the linux7.1 branch15:28
lidstahat engineering school #2 here, climatisation is out of order :') 38°C in room xD I'm gonna dehydrate ^^'15:32
joscho015:35
orvaIt is good that it is in school, in workplace that would probably break some workplace safety laws x)15:35
joschfor the teacher/lecturer the school *is* the workplace15:36
lidstahorva: I'm wondering if we're not already breaking laws15:37
orvaYes, but sadly normal workplace rules rarely affect teachers :/15:37
lidstahyeah, I've added 2 pauses this afternoon so we can refresh/hydrate15:38
lidstahhowever, middle pause is finished, and half the class is still outside15:38
lidstahmight as well end the day outside15:38
lidstahback to work, laters!15:39
+ digitalrane (~digitalra@user/digitalrane)16:10
joschminute: thank you for merging. I rebased linux7.1 on top and it's building now. Once you find some time could you give https://source.mnt.re/reform/qcacld2/-/merge_requests/17 a review and merge if it's okay?16:29
minuteright16:34
minutejosch: thanks for rebasing 7.1!16:34
joschminute: next would be a new reform-tools release with the new machine.conf files16:35
joschthat release would also include the hooks which generate unified kernel images16:36
joschdid you successfully test that or does it need adjustments?16:36
minutejosch: also, many thanks for fixing up qcacld2 :0 wowie16:37
joschyeah that was a headache this time :(16:37
minutelooks annoying yes16:37
minutei wonder if there was ever any more work on the mainline ath10k sdio version...16:37
minutejosch: i didn't yet test booting the UKIs, because tbh i just fixed up the uboot config to also support extlinux...16:38
joschah :D16:38
minutebut i also have to test the latest barebox later today16:39
joschthe ukis might be problematic because u-boot prefers efi over extlinux, so i wonder if they work "too much but not enough" so that uboot picks them up and then fails to boot instead of falling back to extlinux16:39
minuteailurux has been working on getting it polished for a release for rk3588 pocket. it would be an optional thing though at first.16:39
minutejosch: ok, i can test that soonish i guess16:39
joschminute: do you want system images with that or a deb repo?16:40
minutejosch: you mean for barebox?16:40
joschreform-tools and uki16:40
minutejosch: ah. i think system-image would be most comfy16:40
minutejosch: because uboot is already on the machine and i have working sd card now16:40
joschgreat, creating a system image is easier16:41
minuteat least on the 8550. should work on the 6490 too but i can't test it atm as kcxt has that machine.16:41
joschand then i can also easily pop it in16:41
minutenice16:41
joschokay, waiting for the latest reform-debian-packages build to show up in the MNT repo and then pressing the button for the system image build with pre-release of reform-tools16:42
joschoh thank you for merging qcacld2 just now16:43
minutenp!16:43
joschi'll build a reform-system-any image. It will have all the dtbs in the .autodtb sections16:44
joschno clue if and how the right one will get selected so good luck :D16:44
josch(but they'll be in there — you can check with objdump --section=.dtbauto --section-headers)16:46
minutemhm mhm!16:52
minuteour barebox build has some weird issue now where it doesn't work after a power cycle (only worked once on warm reboot). but i restored uboot on that sd card (which has openbsd 7.9) and on serial i get > Welcome to the OpenBSD/arm64 7.9 installation program.17:54
minuteuff, seems to have to do with the sd card :D18:05
joschhah that was me recently on a311d :)18:06
joschsystem image build is now running here: https://source.mnt.re/josch/reform-system-image/-/jobs/2120618:06
joschthat's for this MR which removes flash-kernel and addks uki images and quasar stuff: https://source.mnt.re/reform/reform-system-image/-/merge_requests/15118:07
joschin the kernel, dana reported that the gnome updater wants to remove flash-kernel — this is wrong and i don't know why it is happening. But with the next reform-tools release at least the system should still be bootable.18:08
joschdracut currently conflicts with flash-kernel so you can have only either of them18:08
+ jackhill (~jackhill@kalessin.dragonsnail.net)18:13
+ mark_ (~mjw@2001:1c06:2486:4600:6fcb:4b0e:fedd:fb4d)18:15
minuteok, can test after my barebox adventure here18:24
+ siviq (~siviq@user/siviq)19:07
- siviq (QUIT: Quit: Client closed) (~siviq@user/siviq)19:15
minuteforgot who are the OpenBSD pros here19:40
minutei have openbsd running on the framebuffer console on barebox, but i can't type into the UKC> prompt... neither on usb nor on serial19:40
minutethe cursor after UKC> looks like it's blinking at mega high frequency19:41
minutei have to use UKC because i have to turn off rkusbdpphy (otherwise openbsd hangs when probing it)19:41
joschaccording to chat logs, the "openbsd pros" are bluerise, digitalrane, lidstah, sl and kfx19:42
minutejosch: thx!19:42
^alex(we're an openbsd amateur when it comes to arm64)19:43
minutereading a bit of https://github.com/openbsd/src/blob/4d2a376e6288cabe98a076522907a9444ef48844/sys/dev/fdt/rkusbdpphy.c#L211 (sorry for git(hub) blasphemy)19:43
^alex(we've never used UKC! that's the sorta situation where we'd probably build a custom kernel against medical advice)19:44
minutemhm. i guess i should just try disabling that device in the dtb19:49
minute^alex: is there a file for putting, say, ukc config entries? or is this an interactive thing only?19:51
minuteor i could try getting a 7.8 image...19:51
^alexminute, no, /etc/boot.conf only goes to the top level bootloader. it looks like you have to use `config -e` directly on a kernel to do premodifications19:57
minute^alex: ok thanks!19:57
minutei got it to work by just disabling the usbdphy0+1 in dtb. usb works anyway lol19:57
minute^alex: do you think i could install openbsd on a second partition of the same disk that i imaged image79.img on? (the disk has 32gb so there's a lot of space)19:59
+ spew (~spew@user/spew)20:02
^alexpossibly, we'd just worry about the loader picking up the right partitions20:04
^alexwould probably be a good use for different boot.conf files in each install20:04
^alexdunno how arm64 partitions it, but on amd64 the entire OpenBSD partition is one entry in the MBR20:05
minute^alex: ah. my goal is, as a openbsd noob, to have some sort of bootable system with X going20:05
minute(or wayland? is that a thing?)20:05
^alexopenbsd ships its own X fork20:06
^alexi think wayland is in the package repo20:06
minuteah ok20:06
^alexbut our main openbsd machines literally don't have any VGA hardware lol20:06
minutemhm, network doesn't seem to work... and it doesn't find the firmware for the intel wifi20:10
^alexyeah, the firmwares aren't on the base image, you might have to ferry them over manually.20:12
^alexjust bringup problems20:12
Svpboth netbsd and openbsd use their own in-house flavor of X11 and neither is Xorg20:13
^alexoh? is that new? i think when we used it back in college, netbsd shipped Xorg20:14
^alexbut that was around version 620:14
Svpdont they ship X11R620:14
^alexSvp, https://www.netbsd.org/docs/guide/en/chap-x.html says they ship Xorg on most platforms20:15
+ bob234 (~bob234@81.209.189.43)20:15
Svpi shouldnt operate a computer when tired, sorry lol20:16
bob234Hey, can I download the most recent reform2 dsi dtb somewhere?20:16
minutebob234: you could extract it from the boot partition of the system image20:19
minutebob234: https://mnt.re/system-image/20:19
- mark_ (QUIT: Ping timeout: 245 seconds) (~mjw@2001:1c06:2486:4600:6fcb:4b0e:fedd:fb4d)20:22
bob234minute thx! I will try20:23
bob234oh and btw, whats the most current version?20:24
bob234I mean did it change recently? I think I have one from 10th of December here20:25
bob234(2025) :D20:25
minutebob234:  Apr 22, 202620:26
minute^alex: Svp: aha, disk problem solved, it recognizes an usb stick as sd2 and can partition it20:26
bob234minute thank you so much!20:27
bob234I'm trying to get OpenBSD to recognize some more of the HW. Booting works fine.20:28
+ elb` (~elb@2600:4041:6682:ea00:9bb2:9760:63ba:c33d)20:28
- elb (QUIT: Ping timeout: 246 seconds) (~elb@2600:4041:6682:ea00:f85c:aa14:1e07:cade)20:30
minutebob234: heh, i'm also working on openbsd at the moment (trying it out)20:38
minute"Relinking to create unique kernel..."20:38
minutethat seems to take a while20:38
bob234I've just seen your toots. I'm on a normal Reform, not a Pocket tho20:39
minutebob234: hdmi or dsi display?20:39
^alexminute, yeah, they relink the kernel in a random order and it does take a bit of time lol20:40
bob234dsi which would need the mali driver (so screen is black)20:40
minutewait, mali...20:40
minutenot on rk3588?20:40
minutethe only system we have with mali display engine is the ls1028a20:40
bob234I'm on rk358820:40
bob234oh20:40
bob234wait20:40
minutethat one uses a mali gpu20:40
minuteonly for 3d20:40
minutenot for 2d20:40
bob234whats the one for rk3588 called again?20:41
minutedepends :D20:41
minutevop2 mostly20:41
minutebut if you go via dsi you also need a little driver for the ti sn65dsi86 dsi->edp converter20:41
bob234dmesg is of absolutely no help here: "gpu" at mainbus0 not configured :D20:41
minuteyeah the gpu is not relevant for display20:42
bob234yes "ti,sn65dsi86" at iic3 addr 0x2c not configured I already noticed20:42
bob234I'm not sure that driver will be so little20:44
minutevery little20:46
minuteit can be replaced by a little script that pokes some i2c values20:46
minutebut we should put that in barebox, so that you can just use the EFI GOP20:46
minutebecause openbsd doesn't have the dsi phy drivers etc as well20:46
minutehmmm the installed openbsd system says > init: can't open /dev/console: Device not configured20:47
- Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66)20:47
minutepossibly i have to disable rkdrm?20:56
+ mark_ (~mjw@gnu.wildebeest.org)20:57
bob234hm, I can not mount the reform-system-rk3588-dsi.img on debian20:58
- sterni (QUIT: ) (~quassel@user/sterni)20:58
+ sterni (~quassel@user/sterni)20:59
minutebob234: how did you try to mount it?20:59
minuteok, with boot bsd.rd i can get to a shell21:00
bob234nvm, worked now. Linux is always and adventure for me :D21:03
bob234I'm so used to the BSDs that I need to look up every command the differs (on bad days)21:03
minutesame, but the other way around21:10
minuteok i managed to mount my disk in the bsd.rd21:11
minuteand now doing config -e -o bsd.new bsd21:11
bob234oh nice, with the newer dtb the rng gets recognized21:11
bob234old: "rng" at mainbus0 not configured21:12
bob234new: "rkrng0 at mainbus0"21:12
minuteneat21:12
minutei'm doing ukc> disable rkdrm*21:12
bob234turns out I somehow gave uboot the right commands to boot the newer dts21:13
bob234it even does it automatically when placed in dtbs/rockchip/file.dts very good.21:15
minutei think that did it21:17
minute> starting network21:17
minute> reordering: ld.so libc libcrypto ...21:17
minutenice, sshd up21:17
minute"You have new mail."21:19
minute:021:19
minutenice, logged in via ssh > Welcome to OpenBSD: The proactively secure Unix-like operating system.21:20
bob234good job! btw, I downloaded the linux dts/dtsi from here and noticed mnt-reform is in there.21:20
bob234https://git.kernel.org/pub/scm/linux/kernel/git/devicetree/devicetree-rebasing.git/tree/src/arm64/rockchip21:20
minuteyeah, some reforms are upstreamed21:26
minutebut not necessarily up to date21:26
minutei have X11 running, but xrandr -o doesn't work for rotation. will try to configure wsfb rotation21:27
- paperManu (QUIT: Ping timeout: 265 seconds) (~paperManu@modemcable141.205-200-24.mc.videotron.ca)21:40
bob234Option "Rotate" "CW" in xorg.conf, right?21:45
^alexi have nothing to add, i just like saying "Jean-Baptiste. Emmanuel. Xorg." 21:46
bob234Big Badaboom.21:46
^alexKernel Dallas Multicrash21:46
bob234Multipass?21:47
^alexMulticrash21:47
minutebob234: yes, CCW worked, for wsfb driver21:54
minutenow only the scale is very very smol (high dpi screen)21:54
+ paperManu (~paperManu@bras-base-mtrlpq0776w-grc-29-65-94-79-216.dsl.bell.ca)22:06
minutehello from openbsd on the pocket reform!22:25
minute(in firefox)22:25
rick_woaaahhh nice!!22:25
+ Gooberpatrol66 (~Gooberpat@user/gooberpatrol66)22:32
joschminute: this is what happens when trying to boot the image from the link above on rk3588-dsi classic reform: https://mister-muffin.de/p/Zqly.txt22:35
joschi think this is not using extlinux.conf but uses efi because of Kernel command line: root=LABEL=reformsdroot rw console=ttyS2,150000022:36
joschthe "append" line of extlinux.conf has a bunch more stuff22:36
joschbacklight is on most of the time but nothing is shown for most of the boot-up process22:36
minutejosch: yes that's using EFI22:44
minutejosch: [    0.000000] efi: EFI v2.10 by Das U-Boot22:44
minute[    0.000000] efi: RTPROP=0xeceb1040 SMBIOS 3.0=0xedf43000 INITRD=0xece88040 RNG=0xece87040 MEMRESERVE=0xece8604022:44
minutethat's the proper proof :D22:44
joschlooks like it magically picks up the right dtb even22:46
minuteoh nice, wifi works22:47
+ minutebsd (~root@p509936c9.dip0.t-ipconnect.de)22:49
+ bkeys1 (~Thunderbi@2600:1005:b1c6:c4b3:2bba:3adb:5660:d40b)22:54
minutenot sure if cpufreq works yet, htop always shows 1200 mhz :D22:54
minuteoh! with `apm -H` it goes to 1800mhz22:57
minutebsdi think battery will last longer on openbsd because the gpu isn't used, which is quite power hungry23:20
- bkeys1 (QUIT: Read error: Connection reset by peer) (~Thunderbi@2600:1005:b1c6:c4b3:2bba:3adb:5660:d40b)23:27
- paperManu (QUIT: Ping timeout: 252 seconds) (~paperManu@bras-base-mtrlpq0776w-grc-29-65-94-79-216.dsl.bell.ca)23:39
+ paperManu (~paperManu@204.244.197.237)23:48
+ paperManu_ (~paperManu@104.163.176.82)23:49
- minutebsd (QUIT: Quit: leaving) (~root@p509936c9.dip0.t-ipconnect.de)23:57

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!