2023-07-17.log

- mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org)00:06
- Ar|stote|is (QUIT: Ping timeout: 240 seconds) (~linx@149-210-4-73.mobile.nym.cosmote.net)00:14
+ Ar|stote|is (~linx@149-210-4-147.mobile.nym.cosmote.net)00:19
+ jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net)00:21
- Ar|stote|is (QUIT: Ping timeout: 272 seconds) (~linx@149-210-4-147.mobile.nym.cosmote.net)00:42
+ Ar|stote|is (~linx@149-210-5-215.mobile.nym.cosmote.net)00:46
- jacobk (QUIT: Ping timeout: 264 seconds) (~quassel@47-186-122-163.dlls.tx.frontiernet.net)01:17
- mtm (QUIT: Ping timeout: 258 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)02:03
+ jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net)02:20
+ klardotsh (~klardotsh@98.97.113.255)03:15
- klardotsh (QUIT: Quit: zzz) (~klardotsh@98.97.113.255)04:05
+ klardotsh (~klardotsh@98.97.113.255)04:06
+ mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)04:10
- yewscion (QUIT: Remote host closed the connection) (~yewscion@2601:547:a01:2900:5c15:78d2:b913:fc63)04:24
+ yewscion (~yewscion@2601:547:a01:2900:5c15:78d2:b913:fc63)04:25
- yewscion (QUIT: Remote host closed the connection) (~yewscion@2601:547:a01:2900:5c15:78d2:b913:fc63)04:26
+ yewscion (~yewscion@2601:547:a01:2900:5c15:78d2:b913:fc63)04:27
- jacobk (QUIT: Ping timeout: 264 seconds) (~quassel@47-186-122-163.dlls.tx.frontiernet.net)04:40
joschhttps://source.mnt.re/reform/reform-debian-packages/-/jobs/1593/artifacts/raw/repo/pool/main/m/mesa/mesa-vulkan-drivers_23.1.3-1+reform1_arm64.deb05:27
joschminute: maybe try that .deb -- it contains ./usr/lib/aarch64-linux-gnu/libvulkan_panfrost.so which is what you need, right?05:27
+ jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net)06:03
- XYZ_ (QUIT: Ping timeout: 272 seconds) (~XYZ@78-80-114-190.customers.tmcz.cz)06:33
+ XYZ_ (~XYZ@37-48-48-159.nat.epc.tmcz.cz)06:54
- jacobk (QUIT: Ping timeout: 246 seconds) (~quassel@47-186-122-163.dlls.tx.frontiernet.net)07:01
- klardotsh (QUIT: Ping timeout: 272 seconds) (~klardotsh@98.97.113.255)07:08
joschso, cross compiling anything using llvm in Debian is broken due to the llvm maintainer ignoring the issue: https://bugs.debian.org/89727507:50
joschso if we build a patched mesa, it can only be done via emulated native build which take ~1 hour on the CI07:50
joschI filed https://salsa.debian.org/xorg-team/lib/mesa/-/merge_requests/28 so maybe this gets accepted quickly07:51
joschminute: tell me if that diff is not what you had in mind. Thanks!07:51
- XYZ_ (QUIT: Ping timeout: 272 seconds) (~XYZ@37-48-48-159.nat.epc.tmcz.cz)08:29
+ XYZ_ (~XYZ@37-48-34-18.nat.epc.tmcz.cz)08:42
minutejosch: amazing, thank you very much! i will try it out asap09:44
violetminute: the cm4 adapter thingy only lets u have the internal display driven thru the hdmi to eDP adapter thingy right?09:46
violetlike the soquartz which just has DP on those HDMI lines wouldnt be able to do the internal display cause itd be tryna send displayport to the hdmi side of the hdmi-to-displayport converter09:47
minuteviolet: no. with rpi cm4, one hdmi is used for that, correct. and the other hdmi is going to the outside world. for bpi cm4, mipi-dsi is used just like with the im8xmq09:48
violetim asking about the soquartz though which is cm4 module formfactor09:48
minuteviolet: well, if it has DP, one could make an internal adapter for that too (passive probably)09:49
minuteviolet: why do you prefer the soquartz? do you want to run BSD?09:49
violet8 gigs of ram09:49
violet(limited availability but you can get the 8 gig ones sometimes)09:49
minuteah09:49
minutedoes it also have mipi dsi?09:49
violethmm lemme check09:49
minuteif that is the case, then you can either use that (if it works) or, one could make a FPC adapter cable for the DP output it has09:50
+ mjw (~mjw@gnu.wildebeest.org)09:50
violetthe page says MIPI-DSI 2/4 lanes  09:51
violetand the mipi-dsi is in mainline!!!!!!09:51
minutegood good09:51
violetthe eDP is only in the rockchip kernel not mainline so mipi dsi would be better09:52
amospalla[m]Regarding TUI programs with low requirements, there is a good list of TUI programs I take a look when looking for something https://github.com/rothgar/awesome-tuis09:52
amospalla[m]That list is still being mantained.09:53
minutejosch: hmm, it installs, but i'm struggling to get it picked up, lets see09:54
violetok well i am gonna try to get the soquartz working with the cm4 adapter when it gets here and ill let u know if it works :)09:54
violetthe soquartz hdmi i think is on the pins that you have going to the external so it might even have dual display output09:55
minuteviolet: nice09:56
amospalla[m]I remembered the conversation about spotify some days and this list.09:58
vkoskivamospalla[m]: Thanks for the link09:58
amospalla[m]I hope it may be useful, and I'm not spamming hehe.09:58
amospalla[m]Also, a program not in that list, that I discovered some months ago, that I find incredible. It does what a keyboard firmware like QMK does but in software (manage layouts, layers, macros and so on). https://github.com/jtroo/kanata.10:03
amospalla[m]Before that I had to program a QMK on a keyboard of mine, another QMK on  a usb-to-usb gadget for a non-programmable keyboard, and use an unfinished remapper program (I wrote but didn't finish) on laptops. With kanata I can manage everything on a single place.10:05
ex-parrotoooh this looks great10:06
amospalla[m]Anyway, and related to this, the Reforms have programmable firmware on its input device, are these using some kind of framework like QMK ?10:06
ex-parrotI use a QMK keyboard every day but I dread the next time I have to reflash it10:06
amospalla[m]This program is a gem for me, I've got one single configuration file to rule all my usb keyboards on my desktop and laptops.10:08
amospalla[m]And it can reload on the fly your changes. With this I don't care if a laptop has a crappy keyboard layout. There are more daemons like this, another one is kmonad, but I prefer this one.10:10
minutejosch: i asked in the panfrost channel on oftc and they gave me the magic trick10:14
violetoh yeah also we got hdmi output to the 1440p display but it only does at 1080p which makes sense i think thats all the iMX8MQ can do(?)10:14
violetbut it works!10:14
minuteviolet: imx8mq can do 4K theoretically10:14
minuteviolet: but i think right now the colors are messed up (?)10:15
minutejosch: one has to set the env var PAN_I_WANT_A_BROKEN_VULKAN_DRIVER10:15
violetye it just doesnt even show the option for stuff above 1080. maybe thats a driver thing where the clocks arent in there for it idk10:15
violeti dont have any 4K to test10:15
minute4k is very slow on imx8mq anyway10:19
violetye i dont have a use for 4k so not gonna end up with one10:19
violettoo many pixels for me!10:19
minutehehe 10:24
- mjw (QUIT: Quit: Leaving) (~mjw@gnu.wildebeest.org)10:52
violetoh cool the soquartz also does the thing the banana pi does of having wifi on board 10:54
violetwith an antenna connector10:54
violetand then it has 1x PCIe gen2 lane 10:54
minutenice10:55
minutei do have soquartz somewhere in the office10:55
minutebut i can't stomach another bringup at the moment :D10:55
violetweaker GPU, i think it has 1/3 the execution units10:55
minutejosch: vulkan works! idk how constrained it is though10:55
violetyeah i have a lot of experience with the quartz64 cause ive been using the Model-A SBC-formfactor board as one of my main devices for about half a year now10:56
violetso i know how to get u-boot and the kernel happy and whatnot10:56
violetI cant promise how much i could contribute to the mnt repos around it since im just like, not really familiar with debian. but i can for sure write about whatever i do get working10:57
minuteviolet: nice, i mean i can def. help you debug/get it up and running10:58
violetawesome!10:58
violetthe kernel/device trees are all mainline aside from stuff that only works in the downstream (NPU, for example). and the GPU is a G52 like the banana pi just fewer cores. so if we did try to add support into the mnt image I think the MNT userspace would be compatible and the kernel would just need the right device trees/modules built, no patches needed. the only finnicky thing is that mainline u-boot 11:07
violetand arm trusted firmware doesnt work right now, thats been a whole saga. apparently they're getting *somewhere* with it which is more progress than there was last time i looked into it11:07
violetso basically i just need to learn how to run the script to build the kernel and add my own config lines in to get the debian package to have the right stuff in it hahaha. but yeah we will see when the hardware is here11:08
minutei recently learned how to do that, so i could write a little guide11:13
* mark_ -> mjw11:20
joschPAN_I_WANT_A_BROKEN_VULKAN_DRIVER... oh no XD[6~[6~11:45
joschminute: okay, but then i'll not build patched mesa in the mnt.re CI because it takes too long for questionable value, right?11:46
- XYZ_ (QUIT: Ping timeout: 260 seconds) (~XYZ@37-48-34-18.nat.epc.tmcz.cz)11:46
joschminute: currently, global environment variables are shipped in /etc/profile.d/reform.sh by reform-tools. That file should probably have some shell conditionals added to export the right variables depending on the value in /proc/device-tree/model.11:56
joschDoing so might also help bring-up on new platforms as any unknown platform would start with no custom environment variables exported.11:57
josch(as they might mess things up)11:57
josch(PAN_I_WANT_A_BROKEN_VULKAN_DRIVER doesn't sound very trustworthy :D)11:57
+ XYZ_ (~XYZ@37-48-48-59.nat.epc.tmcz.cz)11:59
josch"WARNING: panvk is not a conformant vulkan implementation, pass PAN_I_WANT_A_BROKEN_VULKAN_DRIVER=1 if you know what you're doing." :D12:02
joschwell, i certainly do not :)12:02
- XYZ_ (QUIT: Ping timeout: 250 seconds) (~XYZ@37-48-48-59.nat.epc.tmcz.cz)12:37
minutejosch: yeah, we don't want this to be enabled by default. just wanted to bring it up12:45
+ XYZ_ (~XYZ@37-48-48-59.nat.epc.tmcz.cz)13:17
- iank (QUIT: Quit: ZNC 1.8.2+deb2+b1 - https://znc.in) (~iank@fsf/staff/iank)13:25
+ iank (~iank@fsf/staff/iank)13:26
- mtm (QUIT: Ping timeout: 245 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)14:04
- buckket (QUIT: Quit: buckket) (~buckket@vps.buckket.org)14:43
+ buckket (~buckket@vps.buckket.org)14:44
+ mark_ (~mjw@gnu.wildebeest.org)14:55
+ vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50)15:41
- mark_ (QUIT: Ping timeout: 245 seconds) (~mjw@gnu.wildebeest.org)15:56
+ mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net)16:10
+ jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net)18:02
- vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50)18:07
- jacobk (QUIT: Ping timeout: 246 seconds) (~quassel@47-186-122-163.dlls.tx.frontiernet.net)18:18
+ mark_ (~mjw@gnu.wildebeest.org)18:42
- mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae)18:43
* mark_ -> mjw18:43
+ mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae)18:43
+ ajr (uid609314@user/ajr)18:48
+ klardotsh (~klardotsh@98.97.114.50)19:54
- S0rin (QUIT: Ping timeout: 245 seconds) (~S0rin@user/s0rin)21:12
+ S0rin (~S0rin@user/s0rin)21:14
+ jacobk (~quassel@47-186-122-163.dlls.tx.frontiernet.net)21:57
+ vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50)22:07
- jacobk (QUIT: Read error: Connection reset by peer) (~quassel@47-186-122-163.dlls.tx.frontiernet.net)22:36

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!