| + S0rin (~S0rin@user/s0rin) | 00:14 | |
| - murdock (QUIT: Read error: Connection reset by peer) (~murdock@2600:380:bc5a:19d6:44de:36e0:2028:dc39) | 00:14 | |
| + murdock (~murdock@dhcp-209-99-207-141.greenmountainaccess.net) | 00:15 | |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 00:17 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:00 | |
| - sevan (QUIT: Ping timeout: 255 seconds) (~sevan@user/venture37) | 01:11 | |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 01:23 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@utdpat241106.utdallas.edu) | 01:31 | |
| + mrdaught (~mrdaught@184.147.249.228) | 01:35 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@184.147.249.228) | 01:38 | |
| - sevan (QUIT: Ping timeout: 264 seconds) (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 02:00 | |
| + jacobk (~quassel@utdpat242031.utdallas.edu) | 02:04 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@utdpat242031.utdallas.edu) | 02:17 | |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 02:21 | |
| - S0rin (QUIT: Ping timeout: 268 seconds) (~S0rin@user/s0rin) | 02:26 | |
| + S0rin (~S0rin@user/s0rin) | 02:51 | |
| - S0rin (QUIT: Ping timeout: 256 seconds) (~S0rin@user/s0rin) | 02:56 | |
| - bleb (QUIT: Ping timeout: 256 seconds) (~cm@user/bleb) | 03:15 | |
| + bleb (~cm@user/bleb) | 03:21 | |
| - digitalrayne (QUIT: Quit: じゃまたね) (~digitalra@vps-446f4f39.vps.ovh.ca) | 03:46 | |
| + digitalrayne (~digitalra@vps-446f4f39.vps.ovh.ca) | 03:46 | |
| + S0rin (~S0rin@user/s0rin) | 03:51 | |
| - S0rin (QUIT: Ping timeout: 260 seconds) (~S0rin@user/s0rin) | 03:57 | |
| - nsc (QUIT: Ping timeout: 272 seconds) (~nicolas@44-49-142-46.pool.kielnet.net) | 04:01 | |
| + nsc (~nicolas@159-96-142-46.pool.kielnet.net) | 04:03 | |
| + S0rin (~S0rin@user/s0rin) | 04:22 | |
| - S0rin (QUIT: Ping timeout: 246 seconds) (~S0rin@user/s0rin) | 04:27 | |
| + S0rin (~S0rin@user/s0rin) | 04:52 | |
| - S0rin (QUIT: Ping timeout: 268 seconds) (~S0rin@user/s0rin) | 05:28 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.12.179) | 05:32 | |
| + Ar|stote|is (~linx@149.210.12.179) | 05:32 | |
| - Ar|stote|is (QUIT: Client Quit) (~linx@149.210.12.179) | 05:32 | |
| + Ar|stote|is (~linx@149.210.12.179) | 05:41 | |
| - Ar|stote|is (QUIT: Client Quit) (~linx@149.210.12.179) | 05:41 | |
| + Ar|stote|is (~linx@149.210.12.179) | 05:42 | |
| - Ar|stote|is (QUIT: Client Quit) (~linx@149.210.12.179) | 05:44 | |
| + S0rin (~S0rin@user/s0rin) | 05:52 | |
| + Ar|stote|is (~linx@149.210.12.179) | 05:52 | |
| - S0rin (QUIT: Ping timeout: 246 seconds) (~S0rin@user/s0rin) | 05:58 | |
| + mrdaught (~mrdaught@bras-base-toroon01zb3-grc-78-184-147-249-228.dsl.bell.ca) | 06:13 | |
| + S0rin (~S0rin@user/s0rin) | 06:23 | |
| - S0rin (QUIT: Ping timeout: 260 seconds) (~S0rin@user/s0rin) | 06:28 | |
| + jacobk (~quassel@64.189.201.150) | 06:47 | |
| + S0rin (~S0rin@user/s0rin) | 07:23 | |
| - chartreuse (QUIT: Read error: Connection reset by peer) (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 07:26 | |
| + chartreuse (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 07:27 | |
| - S0rin (QUIT: Ping timeout: 264 seconds) (~S0rin@user/s0rin) | 07:28 | |
| + chomwitt (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 07:31 | |
| + S0rin (~S0rin@user/s0rin) | 07:53 | |
| - S0rin (QUIT: Ping timeout: 255 seconds) (~S0rin@user/s0rin) | 07:58 | |
| - robin_ (QUIT: Remote host closed the connection) (~robin@user/terpri) | 08:05 | |
| + robin_ (~robin@user/terpri) | 08:05 | |
| josch | minute: linux 6.8-rc7 has some changes that make your device tree fail to compile with: | 08:09 |
|---|---|---|
| josch | Error: /<<PKGBUILDDIR>>/arch/arm64/boot/dts/rockchip/rk3588-mnt-reform2.dts:914.1-15 Label or path usbdp_phy0_dp not found | 08:09 |
| josch | Error: /<<PKGBUILDDIR>>/arch/arm64/boot/dts/rockchip/rk3588-mnt-reform2.dts:918.1-15 Label or path usbdp_phy0_u3 not found | 08:09 |
| josch | Error: /<<PKGBUILDDIR>>/arch/arm64/boot/dts/rockchip/rk3588-mnt-reform2.dts:926.1-15 Label or path usbdp_phy1_u3 not found | 08:09 |
| + polezaivsani (~polezaivs@user/polezaivsani) | 08:12 | |
| josch | i'll try to figure out why but maybe you are faster :) | 08:14 |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 08:15 | |
| josch | ah the collabora patch stack changed rk3588s.dtsi, let me try to see if i can do this... | 08:20 |
| + S0rin (~S0rin@user/s0rin) | 08:23 | |
| josch | since those three were all set to status = "disabled" in the old dtsi and are now missing i guess i can just remove their references as they are all just status = "okay"? | 08:29 |
| - S0rin (QUIT: Ping timeout: 272 seconds) (~S0rin@user/s0rin) | 08:29 | |
| - klardotsh (QUIT: Ping timeout: 256 seconds) (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 08:49 | |
| + S0rin (~S0rin@user/s0rin) | 08:53 | |
| - S0rin (QUIT: Ping timeout: 264 seconds) (~S0rin@user/s0rin) | 08:59 | |
| + chomwitt (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 09:21 | |
| + S0rin (~S0rin@user/s0rin) | 09:35 | |
| - S0rin (QUIT: Ping timeout: 268 seconds) (~S0rin@user/s0rin) | 09:40 | |
| - vkoskiv (QUIT: Ping timeout: 268 seconds) (~vkoskiv@87-100-167-110.bb.dnainternet.fi) | 10:08 | |
| + vkoskiv (~vkoskiv@87-100-167-110.bb.dnainternet.fi) | 10:10 | |
| - chomwitt (QUIT: Ping timeout: 240 seconds) (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 10:24 | |
| + mjw (~mjw@gnu.wildebeest.org) | 10:46 | |
| - murdock (QUIT: Ping timeout: 268 seconds) (~murdock@dhcp-209-99-207-141.greenmountainaccess.net) | 12:18 | |
| + murdock (~murdock@dhcp-209-99-207-141.greenmountainaccess.net) | 12:19 | |
| josch | minute: I saw that yesterday the collabora folks pushed commits into a branch called rk3588-test-cpufreq, so in the hopes that maybe this fixes something for you, I built that thing. This is now rebased on linux 6.8-rc7 and 28 of the 70 patches are already marked as merged: https://mister-muffin.de/reform/linux6.8/ | 12:52 |
| josch | that should be linux-image-6.8-rc7+unreleased-arm64_6.8~rc7-1_arm64.deb | 12:53 |
| minute | oh wow @ 70 patches | 12:53 |
| minute | mega awesome | 12:53 |
| josch | https://gitlab.collabora.com/hardware-enablement/rockchip-3588/linux/-/issues/7 has more details on the cpufreq integration | 13:00 |
| josch | uh apparently there is a fully working and FOSS BL31 also available. Collabora has the patches applied to their fork of TF-A: https://gitlab.collabora.com/hardware-enablement/rockchip-3588/trusted-firmware-a | 13:04 |
| josch | minute: would you like this to be built on source.mnt.re so that we can ship our self-built blobs with the final image? | 13:04 |
| - polezaivsani (QUIT: Remote host closed the connection) (~polezaivs@user/polezaivsani) | 13:35 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@bras-base-toroon01zb3-grc-78-184-147-249-228.dsl.bell.ca) | 13:38 | |
| - murdock (QUIT: Ping timeout: 264 seconds) (~murdock@dhcp-209-99-207-141.greenmountainaccess.net) | 13:38 | |
| + murdock (~murdock@2600:380:bc55:2d77:f1b4:36e:63fd:ccf7) | 13:38 | |
| + mrdaught (~mrdaught@bras-base-toroon01zb3-grc-78-184-147-249-228.dsl.bell.ca) | 13:39 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@bras-base-toroon01zb3-grc-78-184-147-249-228.dsl.bell.ca) | 13:41 | |
| - murdock (QUIT: Ping timeout: 256 seconds) (~murdock@2600:380:bc55:2d77:f1b4:36e:63fd:ccf7) | 13:42 | |
| + murdock (~murdock@2600:380:8c24:96d6:6d9c:602b:22a3:7f2c) | 13:42 | |
| + chomwitt (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 13:43 | |
| + mrdaught (~mrdaught@184.147.249.228) | 13:52 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@184.147.249.228) | 13:56 | |
| * mjw -> Guest1069 | 13:56 | |
| - Guest1069 (QUIT: Killed (osmium.libera.chat (Nickname regained by services))) (~mjw@gnu.wildebeest.org) | 13:56 | |
| * Guest5500 -> mjw | 13:56 | |
| + Guest1069 (~mjw@gnu.wildebeest.org) | 13:57 | |
| + mrdaught (~mrdaught@184.147.249.228) | 14:31 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@184.147.249.228) | 14:31 | |
| + mrdaught (~mrdaught@172.58.111.136) | 14:42 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.136) | 14:43 | |
| + mrdaught (~mrdaught@172.58.111.136) | 15:30 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.136) | 15:32 | |
| + mrdaught (~mrdaught@172.58.111.136) | 15:44 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.136) | 15:46 | |
| + mrdaught (~mrdaught@172.58.111.136) | 15:52 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.136) | 15:52 | |
| + mrdaught (~mrdaught@172.58.111.136) | 16:04 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.136) | 16:04 | |
| minute | josch: before i test your rk3588 stuff, one question, we're working here on the setup wizard. we would like to ship "ly" as the login manager. did you already make an effort to package it for debian? | 16:06 |
| - mjw (QUIT: Quit: Leaving) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 16:21 | |
| + mjw (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 16:21 | |
| - mjw (QUIT: Remote host closed the connection) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 16:23 | |
| + mjw (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 16:23 | |
| minute | ok, done with the meeting, will install the new rk3588 kernel | 16:35 |
| sigrid | minute: does uboot have lcd output on rk3588 as well? | 16:44 |
| + polezaivsani (~polezaivs@user/polezaivsani) | 16:47 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@64.189.201.150) | 16:59 | |
| minute | sigrid: if uboot supports hdmi then yes, but i don't know that yet--so far i use the weird uboot on emmc that came with the icore, haven't tested a cleanly built one yet | 17:00 |
| minute | ok with the latest debian kernel i don't get anymore hdmi output, but i might have messed sth up, lets see | 17:03 |
| minute | josch: the system is waiting a looong time during boot until it says > Timed out for waiting the udev queue being empty. | 17:13 |
| - pandora (QUIT: Quit: Connection closed for inactivity) (uid585533@2a03:5180:f:3::8:ef3d) | 17:22 | |
| josch | minute: no, i have not looked deeper into ly | 17:44 |
| josch | minute: funny, so there is a regression. Would you like me to rebase the new patch queue on 6.8-rc1 to see if the problem is the patch queue or the kernel itself? | 17:46 |
| + jacobk (~quassel@utdpat242053.utdallas.edu) | 17:53 | |
| minute | josch: yes please :3 | 18:04 |
| minute | in the meantime i'll try to figure out how to interact with the system | 18:05 |
| minute | as i don't get a console on serial anymore, it stops after > Begin: Running /scripts/init-bottom ... done. | 18:05 |
| minute | so maybe it switches over to an invisible framebuffer | 18:05 |
| josch | okay, i'll not do this now because i wanted to watch some videos on my reform this evening but i can run it over night and have it ready tomorrow morning | 18:10 |
| minute | sure thing! i'll tinker a bit to see if i find what's the problem and/or install the old package | 18:11 |
| minute | josch: when you have time, one Q: i would like to ship a /etc/systemd/system/getty@.service.d/override.conf with reform-system-image that autologins as root. this file will be deleted after completing the setup wizard. i think it's best to generate this in reform-system-image, would you agree? (because it is temporary) it's 3 lines of text. | 18:21 |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@utdpat242053.utdallas.edu) | 18:21 | |
| minute | perhaps i can even make it start the setup wizard directly without using agetty, then we need even less temporary configuration | 18:22 |
| minute | otherwise, i would execute the setup wizard in the same way that reform-root-help is currently started | 18:25 |
| jackhill | oooh, the RK3588 is in the shop. I wonder if I should buy one. | 18:25 |
| minute | jackhill: yep it is :3 | 18:25 |
| jackhill | I never know if I should treat myself or finish working on some of my ohter projects first :) | 18:27 |
| jackhill | how does power draw compare to i.MX8MQ? | 18:27 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 18:40 | |
| + mrdaught (~mrdaught@172.58.109.148) | 18:47 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.109.148) | 18:51 | |
| + mrdaught (~mrdaught@172.58.109.148) | 18:56 | |
| - pinoaffe (PART: !!unknown attribute: msg!!) (~pinoaffep@2a01:4f9:c010:a00b:1337:1337:11be:10) | 19:00 | |
| minute | josch: using my monolithic kernel, but leaving the initramfs from your build in place, the display comes on again and i get initramfs console with break=premount. but i couldn't get to this console with your new kernel | 19:00 |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.109.148) | 19:01 | |
| minute | jackhill: i can't really give good power numbers atm because one also has to consider the hdmi adapter's power draw, and it might not be necessary once DP drivers are available. it also depends a lot on the system load. i.e. if you use gpu and all 8 cores the draw can get up to like 0.5 A | 19:02 |
| minute | i believe that once frequency scaling and the NI clock stuff is more mature the idle power consumption should be quite low | 19:02 |
| minute | while i'm at it, i will do a `7z b` on my monolithic kernel and then compare to josch's debian kernel | 19:05 |
| minute | total score on 7z b is 14987 | 19:07 |
| minute | on imx8mq it is 4178 | 19:08 |
| minute | so speedup almost 3.59x | 19:09 |
| minute | a311d score is 8981 | 19:09 |
| josch | on a311d it is 8981 | 19:09 |
| josch | ah :) | 19:09 |
| minute | hehe yeah | 19:09 |
| minute | so 1.67x speedup approx | 19:09 |
| josch | that should give you kernel build times of approx 2 hours compared to my 3:20 | 19:10 |
| minute | but it's not operating at full speed yet | 19:10 |
| minute | the big cores are only at 2.2ghz and small cores at 1.8ghz | 19:11 |
| minute | i think because the pvt stuff is not in the collabora patches yet? | 19:11 |
| josch | no clue, i didn't read the patches but just built the freqscaling branch :) | 19:11 |
| minute | nice | 19:11 |
| minute | so it can normally operate at 2.4ghz/2.2ghz | 19:12 |
| minute | if i'm not mistaken | 19:12 |
| minute | oh also i don't have the heatsink on at the moment... :D | 19:13 |
| josch | so it might scale down by itself? | 19:14 |
| minute | not sure | 19:16 |
| josch | at least there is one mostly working version of the kernel. With that we can bisect to find out what broke it. I actually expected the issue you are having right now with the first try. :D | 19:17 |
| minute | ohh | 19:17 |
| minute | trying your rc1 again now | 19:17 |
| minute | yeah that one is fine | 19:17 |
| josch | perfect | 19:17 |
| minute | i wonder if my dtb became incompatible with your rc7? | 19:17 |
| josch | rc1-1 or rc1-2? | 19:17 |
| minute | josch: linux-image-6.8-rc1+unreleased-arm64_6.8~rc1-1_arm64.deb | 19:18 |
| josch | yes, the collabora patches do change rk3588 dtsi files, so that seems to be a possibility | 19:18 |
| minute | ok... can i find a dtb in there somewhere? | 19:18 |
| josch | minute: i also had rc1-2 which enabled more config options | 19:18 |
| minute | oh i can download that | 19:18 |
| josch | you want the dts or the dtb? | 19:19 |
| josch | the dtb is at ./usr/lib/linux-image-6.8-rc1+unreleased-arm64/rockchip/rk3588-mnt-reform2.dtb | 19:20 |
| josch | minute: i saw your localectl post on mastodon -- what do you want to do? is that for the first-boot stuff? | 19:22 |
| minute | josch: your rc1-2 works | 19:29 |
| josch | okay, then it's not about my changes to the kernel config | 19:29 |
| josch | (i marked some settings as "m" instead of "y") | 19:30 |
| minute | ohh interesting | 19:30 |
| minute | i get more info now when trying to modprobe rockchip-cpufreq | 19:30 |
| josch | rc1-2 also has the scaling setting turned on which rc1-1 does not | 19:30 |
| josch | exactly | 19:30 |
| minute | > [ 53.247273] cpu cpu0: _parse_named_prop: Invalid number of elements in microvolt property (6) with supplies (1) | 19:30 |
| minute | > [ 53.249124] cpu cpu0: _of_add_opp_table_v2: Failed to add OPP, -22 | 19:30 |
| josch | the name of the config setting is somewhere up above in the backlog | 19:30 |
| minute | so that's the cause of the -22 | 19:30 |
| josch | CONFIG_ARM_SCMI_CPUFREQ=m | 19:31 |
| minute | ok cool so i need to tweak the OPP table in my homebrew dts | 19:31 |
| minute | josch: we want to do this: https://source.mnt.re/reform/mnt-reform-setup-wizard/-/issues/1 | 19:31 |
| minute | (regarding localectl) | 19:31 |
| minute | we found out that we have to edit /etc/default/keyboard | 19:32 |
| minute | so that's fine | 19:32 |
| josch | minute: feel free to look at my first-boot merge request, it already has some code doing edits to /etc/default/keyboard | 19:33 |
| minute | ah nice, i hope so... just need to find that again | 19:33 |
| josch | but it's not querying x11 xml files | 19:33 |
| josch | https://source.mnt.re/reform/reform-system-image/-/merge_requests/43 | 19:33 |
| josch | but this is using the initramfs which is not the route that you want to go :) | 19:34 |
| minute | yeah! but it is otherwise close to how we want to do it, so that's cool | 19:35 |
| josch | but maybe you can steal some code like partition resizing and stuff | 19:35 |
| josch | <= afk dinner | 19:35 |
| minute | enjoy dinner! | 19:36 |
| josch | thank you! :) | 19:36 |
| - polezaivsani (QUIT: Remote host closed the connection) (~polezaivs@user/polezaivsani) | 19:42 | |
| + klardotsh (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 19:47 | |
| minute | josch: when you have time, where can i see the exact patchstack you applied? my guess is that the rockchip-cpufreq driver in that stack is perhaps too old, not too new? | 19:47 |
| josch | minute: you can obtain the sources by running "dget -u https://mister-muffin.de/reform/linux6.8/linux_6.8~rc1-2.dsc" or choose any other .dsc from that directory | 19:49 |
| minute | ah | 19:49 |
| josch | dget will download the dsc itself, the orig tarball, the debian patches and unpack it | 19:49 |
| minute | thanks! | 19:49 |
| josch | i wrote these steps already yesterday i think? there are in the backlog somewhere | 19:49 |
| josch | and then the collabora patches are in ./debian/patches/rk3588 in the unpacked source package | 19:49 |
| minute | hmmmmmmmm but maybe i was wrong, these opps have only 3 microvolt entries https://gitlab.collabora.com/hardware-enablement/rockchip-3588/linux/-/commit/50cef69e777ac927308f3f357cb7e6ec62e0a19b | 19:50 |
| minute | trying that now | 19:54 |
| minute | that table also has the 2.4ghz opp | 19:54 |
| minute | yessss that was the issue | 19:54 |
| minute | ok josch i was mistaken, i have that new opp table now in my dtb and i get frequencies again in htop | 19:55 |
| minute | now i really need a heatsink... | 19:56 |
| + mrdaught (~mrdaught@172.58.109.148) | 19:58 | |
| minute | josch: the only question now is why isn't rockchip-cpufreq being loaded automatically | 19:58 |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.109.148) | 20:00 | |
| minute | 7z b score now 16751 | 20:01 |
| minute | so almost exactly 4x the imx8mq score | 20:03 |
| ehenter | Congratz for making RCORE module reality! I got 3136 points on LS1028A (or did I do something wrong?) so it seems quite a big upgrade over that as well :O | 20:40 |
| josch | ehenter: your ls1028a score is realistic. Compiling linux on imx8mq on all its four slow cores is quicker than compiling it on ls1028a on its two fast cores. | 20:41 |
| ehenter | Btw, have you benchmarked on how much faster it is on imx8mq compared to ls1028a? | 20:47 |
| + mrdaught (~mrdaught@172.58.109.148) | 20:50 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.109.148) | 20:54 | |
| erle | i have rendered build instructions for the LEGO MNT reform keyboard holder. i will put this on my website, but if anyone thinks i should adjust the instructions, please tell! isometric: https://mister-muffin.de/p/myNa.pdf not isometric: https://mister-muffin.de/p/-E_t.pdf | 20:54 |
| + jacobk (~quassel@64.189.201.150) | 21:05 | |
| minute | ehenter: yes, it's much faster in multicore than ls1028a (8 vs 2 cores). but ls1028a is the only module that is completely open hardware. | 21:07 |
| ehenter | That is why I have one and I'm very happily daily-driving it :) | 21:13 |
| ehenter | Honestly I only hope to have a bit more performance when using web browser... | 21:14 |
| minute | ehenter: are you using firefox or chromium? | 21:15 |
| ehenter | Firefox | 21:17 |
| ehenter | Is there noticeable difference between those two? | 21:18 |
| minute | ehenter: firefox is def. the better overall choice but chromium might give you more perf | 21:18 |
| ehenter | Ah | 21:18 |
| minute | it's usually faster on arm | 21:18 |
| ehenter | Okay, I should try it to see if it brings big enough difference to give reason tolerate google | 21:20 |
| ehenter | Now Debian unstable seems still to be in bad shape so I do not want to try breaking my system | 21:22 |
| minute | yeah good call | 21:27 |
| vkoskiv | Can upstream debian be installed on the Reform yet? Someone asked me that at instanssi and I wasn't sure | 21:41 |
| minute | i'm not sure myself cc vagrantc | 21:43 |
| josch | ehenter: would you prefer installing Debian bookworm on your reform? | 21:44 |
| ehenter | Well I can deal with unstable but I'm currently waiting for new SSD and I'm considering if I install stable release on it | 21:46 |
| josch | vkoskiv: yes, with just vanilla debian on ls1028a you get a serial terminal | 21:46 |
| minute | and on imx8m? | 21:47 |
| josch | i didn't try | 21:48 |
| josch | i'd expect that something will not work on imx8mq because there are a few quite critical patches that are still not upstreamed | 21:48 |
| josch | on the other hand i remember that somebody got debian installer to work with the vanilla kernel via uart? | 21:48 |
| ehenter | But wasn't it possible to load required edp fw with uboot on ls1028a and having functional display output that way, or did I misunderstood something? | 21:51 |
| josch | ehenter: the edp firmware is embedded in the u-boot binary shipped with MNT, you don't have to manually put it there | 21:52 |
| josch | oh wait, it was the other thing | 21:53 |
| - mjw (QUIT: Killed (osmium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 21:53 | |
| * Guest1069 -> mjw | 21:53 | |
| josch | it's in /boot | 21:53 |
| + Guest114 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 21:53 | |
| josch | ls1028a-mhdpfw.bin is put into /boot on MNT system images | 21:54 |
| ehenter | Yeah, but does the kernel need that as well? Or are there other kernel level changes needed on vanilla debian kernel? Sorry for my ignorance on this but I'm not too familiar how bootloader and kernel initializes stuff on boot :D | 21:57 |
| josch | i used to know but forgot all the details :D | 21:58 |
| josch | ah we have an initramfs hook which copies /boot/ls1028a-mhdpfw.bin into /lib/firmware | 21:59 |
| josch | i guess there it is picked up by the respective kernel module | 21:59 |
| josch | minute: while refreshing my memory i just saw that initramfs-tools/scripts/reform copies this long list of modules in the initramfs -- but it should only do this for imx8mq and on none of the other platforms, correct? | 22:00 |
| ehenter | But is it possible to use vanilla debian and just add that fw there manually and make it work that way? | 22:01 |
| ehenter | Oh, device tree is also pathed in? | 22:02 |
| josch | there are a few more tweaks which are done by reform-tools | 22:02 |
| ehenter | Ok, so not yet at least :D | 22:02 |
| josch | ehenter: we ship a custom dts but what vanilla linux has might be good enough | 22:03 |
| josch | if it has the mnt-reform bits -- i'm not sure | 22:03 |
| josch | ehenter: why don't you just try it out and see what happens? | 22:03 |
| josch | ehenter: essenitally, the reform-tools package documents all the changes between vanilla debian and what the reform requires | 22:04 |
| josch | except for kernel stuff that is | 22:04 |
| josch | ehenter: what is your motivation behind running vanilla debian? | 22:08 |
| ehenter | Sorry to not mention but nothing really. I was just curious how far I could go with vanilla kernel and debian. I tried it once but got black screen and no idea how to move forward | 22:09 |
| ehenter | I looked at upstream kernel and some reform bits are missing from device tree at least | 22:10 |
| josch | yes, the diff can be found in the reform-debian-packages repo | 22:10 |
| josch | there are a couple of reform specific patches in that repo which need somebody to do the work of submitting them to upstream linux and work with kernel maintainers to get them mainlined | 22:11 |
| ehenter | Hmm, but what kernel non-backport is from reform.debian.net? Vanilla or patched stable one? | 22:15 |
| josch | ehenter: the kernels on reform.debian.net are all from debian plus the reform patches | 22:17 |
| josch | ehenter: this is the same that is done for the kernels on MNT repos | 22:17 |
| josch | the difference is the debian kernel version they are applied to | 22:17 |
| josch | the same scripts are generating the content for reform.debian.net and the MNT repos | 22:17 |
| josch | the difference is that for reform.debian.net, DIST is set to bookworm or bookworm-backports an for the MNT repos, the DIST variable is set to unstable | 22:18 |
| minute | josch: i think so @ modules... need to check the list if there's only imx8mq stuff | 22:18 |
| josch | reset_imx7 mux_mmio fixed i2c-imx fan53555 i2c_mux_pca954x pwm_imx27 pwm_bl panel_edp ti_sn65dsi86 phy-fsl-imx8-mipi-dphy nwl-dsi mxsfb | 22:20 |
| josch | that's all imx stuff i think | 22:20 |
| josch | maybe you need to check whether those are required for imx8m plus? | 22:20 |
| ehenter | josch: Ah ok, thanks! For some reason I was under impression that those were vanilla debian with some userspace tweaks but I have misunderstood something at some point. Sorry! | 22:24 |
| minute | josch: hmm if a bunch of these are missing then the display can't work on a311d and rpi-cm4 and imx8mplus too | 22:26 |
| josch | oh okay! | 22:26 |
| josch | well, there is not much benefit in *not* including extra modules even when they are not needed i guess | 22:26 |
| minute | josch: its strange to me that common stuff like "fixed" or "pwm_bl" or "mux_mmio" or "fan53555" etc wouldn't be there by default though | 22:27 |
| josch | so unless somebody has too much free-time and wants to get the exact list per platform... :) | 22:27 |
| minute | josch: yeah | 22:27 |
| minute | yeah not worth i think | 22:27 |
| josch | i'll leave a code-comment for future-me | 22:27 |
| josch | minute: well, if you have too much free time (haha) you could check what happens if you delete some of these modules from the initramfs on imx8mq | 22:32 |
| minute | i don't really have any imx8mq reform ready for such tests atm | 22:33 |
| minute | the only one in use here atm is the silver reform that we use to print all shipping labels | 22:33 |
| ehenter | I do have one *real* question: how do I read battery status? I have reform2-lpc module loaded but I don't know where to e.g. read battery %? | 22:34 |
| minute | but that can't be hacked on :D | 22:34 |
| josch | hrm... i can try on my reform what happens :) | 22:35 |
| josch | maybe modules get loaded automatically | 22:35 |
| josch | but only tomorrow when i have a second computer at hand :) | 22:36 |
| + mrdaught (~mrdaught@172.58.111.135) | 22:50 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.135) | 22:51 | |
| - chomwitt (QUIT: Ping timeout: 256 seconds) (~chomwitt@2a02:587:7a18:fc00:1ac0:4dff:fedb:a3f1) | 22:52 | |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 22:52 | |
| - sevan (QUIT: Changing host) (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 22:54 | |
| + sevan (~sevan@user/venture37) | 22:54 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@64.189.201.150) | 22:56 | |
| josch | ehenter: if you have reform2-lpc loaded, then you should have this: /sys/class/power_supply/8xlifepo4/charge_now | 23:00 |
| ehenter | I do not have that. When I lsmod, I see reform2_lpc so I guess it's loaded? | 23:03 |
| minute | hmm, looks like spi is still disabled in our ls1028a dts | 23:05 |
| minute | will have to look at that next week | 23:05 |
| minute | so far then you can only get the battery status via the oled display @ ls1028a | 23:05 |
| ehenter | Yeah, I've been relying on that every time when I'm on battery | 23:06 |
| ehenter | But it would be nice to get some other feedback when battery is running low | 23:07 |
| ehenter | But please ping me if I can try some stuff on my reform! And diffs are also ok, I can rebuild new kernel if needed :) | 23:08 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 23:32 | |
| - murdock (QUIT: Read error: Connection reset by peer) (~murdock@2600:380:8c24:96d6:6d9c:602b:22a3:7f2c) | 23:33 | |
| + murdock (~murdock@dhcp-209-99-207-141.greenmountainaccess.net) | 23:33 | |
| + mrdaught (~mrdaught@172.58.111.135) | 23:45 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.135) | 23:49 | |
| + mrdaught (~mrdaught@172.58.111.135) | 23:53 | |
| - mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@172.58.111.135) | 23:54 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!