| vagrantc | curious how easy swapping out the CPU module on the mnt/reform ... is likely to be ... does it generally involve just unscrewing a few things and popping the module out ... or is it likely to also involve removing the heatsink and reapplying thermal paste and all that "fun" ? | 00:08 |
|---|---|---|
| vagrantc | as ... the risc-v CM4 module just suddenly made things way more interesting :) | 00:08 |
| vagrantc | especially with debian moving riscv64 into an official architecture | 00:10 |
| vagrantc | plenty of ugly warts, still, i'm sure ... but certainly an exciting prospect | 00:11 |
| josch | vagrantc: you have to remove the heatsink to swap out the module | 00:11 |
| josch | when i did that i did not reapply thermal paste | 00:12 |
| josch | i removed the heatsink multiple times to reach the DIP switch | 00:12 |
| josch | switching the module itself is similar to swapping out sodimm ram modules | 00:12 |
| josch | vagrantc: we have another candidate for first building a non-working mainline u-boot which is patched with non-free blobs afterwards: https://source.mnt.re/josch/reform-system-image/-/commit/67eac49f453bc22d0b9435548dfe0084a258f825 | 00:14 |
| vagrantc | oh! | 00:15 |
| vagrantc | josch: ah, yeah, that looks a lot like the existing u-boot-amlogic platform in debian (which really should be moved to contrib) | 00:17 |
| josch | vagrantc: it even has this fun license: https://github.com/LibreELEC/amlogic-boot-fip/blob/master/LICENSE | 00:19 |
| josch | Debian does not happen to have "prior written permission from Amlogic", does it? ;) | 00:20 |
| vagrantc | yeah ... i basically read that as "please do not use anything we make" ... and yet... | 00:23 |
| josch | even my soon-to-be-delivered ls1028a got this blasted non-distributable edp firmware... | 00:29 |
| josch | (yes, i can go rack-reform but i'd like to have a laptop) | 00:30 |
| - stites (QUIT: Ping timeout: 246 seconds) (~stites@c-71-232-14-73.hsd1.ma.comcast.net) | 00:51 | |
| + stites (~stites@130.44.147.204) | 00:52 | |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@40-99-142-46.pool.kielnet.net) | 01:07 | |
| + nsc (~nicolas@40-99-142-46.pool.kielnet.net) | 01:15 | |
| - Boostisbetter (QUIT: Ping timeout: 244 seconds) (4a410829d7@irc.cheogram.com) | 01:28 | |
| - bluerise (QUIT: Ping timeout: 250 seconds) (~bluerise@user/bluerise) | 01:43 | |
| + bluerise (~bluerise@p5b0acd9a.dip0.t-ipconnect.de) | 01:44 | |
| - mtm (QUIT: Ping timeout: 245 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| - jacobk (QUIT: Ping timeout: 246 seconds) (~quassel@utdpat241106.utdallas.edu) | 02:22 | |
| - marty_mcfly88 (QUIT: Ping timeout: 255 seconds) (~marty@2603-6000-8c01-0f3e-452a-9a32-cc5c-9f84.res6.spectrum.com) | 03:17 | |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 04:00 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 04:01 | |
| - nsc (QUIT: Ping timeout: 258 seconds) (~nicolas@40-99-142-46.pool.kielnet.net) | 04:02 | |
| + nsc (~nicolas@152-48-142-46.pool.kielnet.net) | 04:04 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 04:09 | |
| josch | minute: the reform-debian-packages pipeline is now using the build results of the last successful build_patched job if those are newer than the packages in unstable when you create a branch other than main (which is what you do when implementing a new feature) | 04:23 |
| josch | minute: so now a full run of the reform-debian-packages pipeline for your new feature takes ~40 minutes instead of 1:50h | 04:23 |
| josch | because the build_patched job is essentially skipped | 04:24 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 04:39 | |
| josch | minute: why did you remove firmware-realtek from PKGSMISC? if the same packages are installed in all images, it makes it easier to move an existing system from one SoM to the other because less parts have to be changed | 04:57 |
| + jacobk (~quassel@129.110.242.173) | 05:20 | |
| - ggoes (QUIT: Ping timeout: 245 seconds) (~gregf@fsf/staff/ggoes) | 06:23 | |
| + reform6131 (~inhji@ip-095-223-184-042.um35.pools.vodafone-ip.de) | 07:03 | |
| - reform6131 (QUIT: Client Quit) (~inhji@ip-095-223-184-042.um35.pools.vodafone-ip.de) | 07:04 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 07:12 | |
| - flowy (QUIT: Server closed connection) (~flowy@2a01:4f8:c0c:1a8f::1) | 08:04 | |
| + flowy (~flowy@2a01:4f8:c0c:1a8f::1) | 08:04 | |
| josch | minute: and another note of your reform-system-image a311d-v2 branch: you now have an updated list for /etc/initramfs-tools/modules which you write out in mkuserland.sh | 08:15 |
| josch | minute: could you instead use reform-tools to install /usr/share/initramfs-tools/modules.d/reform.conf with the right list as outlined here: https://source.mnt.re/reform/reform-tools/-/merge_requests/42 | 08:16 |
| josch | you could add these changes to your tools-202309 branch of reform-tools | 08:16 |
| josch | alternatively, i can also use your list from mkuserland.sh to finalize that merge request, then merge it and then you rebase your tools-202309 branch on top of it | 08:18 |
| josch | the fewer files reform-system-image puts into /etc the better, because then we can use reform-tools to change them on apt upgrade | 08:19 |
| + mjw (~mjw@gnu.wildebeest.org) | 08:38 | |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 08:47 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@129.110.242.173) | 09:04 | |
| josch | ex--parrot: i created a WIP MR for your lpc poweroff changes: https://source.mnt.re/reform/reform-tools/-/merge_requests/46 | 09:15 |
| josch | ex--parrot: do you have a better value for the git author field for me? :) | 09:15 |
| josch | minute: could you review those changes by ex--parrot? | 09:15 |
| - klardotsh (QUIT: Ping timeout: 240 seconds) (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 09:31 | |
| josch | minute: you might also want to have a look at this for a311d: https://source.mnt.re/reform/reform-tools/-/merge_requests/47 | 09:35 |
| josch | oh shoot... i just roasted something on my mainboard very close to the 3.5 mm jack | 10:06 |
| josch | i heard a small sizzle and then smoke came out from next to the 3.5 mm jack | 10:06 |
| josch | it still turns on... | 10:06 |
| josch | okay, it tried my audio... | 10:08 |
| josch | *fied | 10:08 |
| josch | only a long beeping noise comes from my headphones now | 10:08 |
| andreas-e | Ah, that is terrible! | 10:11 |
| josch | at least the rest of the system still seems to work -- i can plug in a usb sound card | 10:12 |
| - mjw (QUIT: Ping timeout: 255 seconds) (~mjw@gnu.wildebeest.org) | 10:34 | |
| josch | so... the smoke came from around this area https://mister-muffin.de/p/8hgX.png that is where D10, littlefuse SP0503BAHT is located... is that plausible, given that i was messing around with the speaker connectors? | 10:46 |
| * mark_ -> mjw | 11:24 | |
| minute | josch: sorry to hear! and thanks for all the messages, unfortunately i am fighting with a little tech emergency myself atm | 12:26 |
| minute | boundary devices changed something with the imx8mq modules in their latest batch that we have and built into new reforms, and they can't boot from sd card | 12:26 |
| minute | like, at all | 12:27 |
| minute | and the shipped uboot on emmc just says dram_init error :S | 12:27 |
| minute | also mysterious https://github.com/boundarydevices/u-boot/commit/ab98ab7b540e164795f68b2274eedde0707493a0 | 12:27 |
| josch | that sounds way worse than a "little" tech issue o0 | 12:27 |
| minute | yeah, i need to figure out how to make these modules boot, and i'm under a bit of pressure | 12:28 |
| josch | uff... i hope you manage to fix this! | 12:28 |
| minute | also, schematics seem to be gone from boundary's website | 12:30 |
| minute | for nitrogen8m som | 12:30 |
| minute | ok so these have a new version of the imx8mq | 13:02 |
| minute | AB instead of AA | 13:02 |
| grimmware | :< | 13:17 |
| minute | ok so jacqueline's debug board shows that the card enters usb bootloader mode instead of trying to boot from sd card | 13:27 |
| - mtm (QUIT: Ping timeout: 258 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| minute | ok so these imx chips are probably eFUSEd differently | 14:03 |
| minute | instead of sd card they default to usb loader | 14:04 |
| - unixpoet (QUIT: Remote host closed the connection) (~unixpoet@user/unixpoet) | 14:45 | |
| + unixpoet (~unixpoet@user/unixpoet) | 14:46 | |
| minute | i managed to flash our uboot to the emmc of this module using "uuu" and fastboot | 15:14 |
| minute | and via our uboot on emmc i can then auto-boot from sd card or emmc | 15:14 |
| josch | but it cannot load u-boot from sd-card anymore? what does the DIP switch do then? | 15:18 |
| minute | josch: correct. so the dip switch enables USB serial downloader mode. on the revision AA modules, this fails and then falls back to "SD/MMC manufacturing boot" mode | 15:19 |
| minute | josch: but on AB this does not happen for some reason. it always enters USB serial downloader mode | 15:19 |
| minute | one could set the efuses to boot from SD card. but then i think you can never again boot from emmc | 15:19 |
| minute | we should probably offer a very low-cost/minimal usb flashing adapter for the modules | 15:21 |
| + marty_mcfly88 (~marty@2603-6000-8c01-0f3e-3e84-311f-ea58-0631.res6.spectrum.com) | 15:38 | |
| - stites (QUIT: Ping timeout: 246 seconds) (~stites@130.44.147.204) | 15:39 | |
| + stites (~stites@2607:fb90:ad60:f7c7:2331:ab6c:f67e:34fc) | 15:40 | |
| sevan | missed this earlier https://mastodon.f-si.org/@fsi/110957609863026755 | 15:45 |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb90:ad60:f7c7:2331:ab6c:f67e:34fc) | 15:56 | |
| + stites (~stites@2607:fb90:ad60:f7c7:2331:ab6c:f67e:34fc) | 15:57 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:08 | |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb90:ad60:f7c7:2331:ab6c:f67e:34fc) | 16:13 | |
| + stites (~stites@155.33.134.55) | 16:13 | |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@155.33.134.55) | 16:17 | |
| + stites (~stites@155.33.134.55) | 16:17 | |
| - jjbliss (QUIT: Server closed connection) (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 16:19 | |
| + jjbliss (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 16:19 | |
| - XYZ_ (QUIT: Read error: Connection reset by peer) (~XYZ@78-80-106-89.customers.tmcz.cz) | 16:26 | |
| josch | what a bummer... well at least it boots! | 16:38 |
| + XYZ_ (~XYZ@78-80-106-89.customers.tmcz.cz) | 16:42 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 16:52 | |
| - XYZ_ (QUIT: Remote host closed the connection) (~XYZ@78-80-106-89.customers.tmcz.cz) | 17:00 | |
| + XYZ_ (~XYZ@78-80-106-89.customers.tmcz.cz) | 17:04 | |
| - XYZ_ (QUIT: Remote host closed the connection) (~XYZ@78-80-106-89.customers.tmcz.cz) | 17:23 | |
| + XYZ_ (~XYZ@78-80-106-89.customers.tmcz.cz) | 17:26 | |
| minute | josch: yeah, and i'm thankful for the updates you et al put into the uboot stuff so that it auto boots from sd card or usb, this helps now! | 17:27 |
| minute | josch: i moved firmware-realtek only to a311d because some people might argue they don't want unnecessary binary firmware on their imx8mq system (right now imx8mq doesn't need non-free-firmware) | 17:29 |
| sigrid | what's that firmware for? wifi? | 17:31 |
| minute | sigrid: yep, wifi+bt on a311d (realtek 8821cs) | 17:35 |
| josch | minute: but reform-system-image ships non-free-firmware already in the form of /boot/flash.bin | 17:36 |
| minute | josch: well, that is true. | 17:38 |
| - murphnj (QUIT: Remote host closed the connection) (~murph@user/murphnj) | 18:03 | |
| - andreas-e (QUIT: Read error: Connection reset by peer) (~Andreas@2001:861:c4:f2f0::c64) | 18:13 | |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 18:13 | |
| - andreas-e (QUIT: Quit: Leaving) (~Andreas@2001:861:c4:f2f0::c64) | 18:54 | |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 18:55 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:14 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:22 | |
| * mark_ -> mjw | 19:22 | |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:22 | |
| - XYZ_ (QUIT: Remote host closed the connection) (~XYZ@78-80-106-89.customers.tmcz.cz) | 19:24 | |
| + ajr (uid609314@user/ajr) | 19:26 | |
| + XYZ_ (~XYZ@78-80-106-89.customers.tmcz.cz) | 19:27 | |
| minute | is there a modern alternative for synaptic? it errors out blaming wayland | 19:43 |
| abortretryfail | wifi+bt combined sounds convenient. | 19:44 |
| - stites (QUIT: Ping timeout: 246 seconds) (~stites@155.33.134.55) | 19:52 | |
| + stites (~stites@2607:fb91:dc6:e420:4e89:c540:3dc4:b5df) | 19:53 | |
| - inhji (QUIT: Server closed connection) (~inhji@2a01:4f8:1c0c:42e4::1) | 19:58 | |
| + inhji (~inhji@2a01:4f8:1c0c:42e4::1) | 19:58 | |
| + jacobk (~quassel@129.110.242.224) | 20:00 | |
| - marty_mcfly88 (QUIT: Remote host closed the connection) (~marty@2603-6000-8c01-0f3e-3e84-311f-ea58-0631.res6.spectrum.com) | 20:00 | |
| + marty_mcfly88 (~marty@2603-6000-8c01-0f3e-9174-e6f3-c012-bccc.res6.spectrum.com) | 20:01 | |
| - andreas-e (QUIT: Ping timeout: 246 seconds) (~Andreas@2001:861:c4:f2f0::c64) | 20:01 | |
| + andreas-e (~Andreas@2001:861:c4:f2f0:b0aa:56dd:ae73:8afa) | 20:01 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@129.110.242.224) | 20:11 | |
| josch | minute: like gnome-software? | 20:14 |
| josch | the kde equivalent is apper i think | 20:16 |
| unixpoet | discover, for KDE | 20:31 |
| unixpoet | hm. they both seem to do similar things, I'm not sure why they both exist | 20:33 |
| josch | i thought discover is for hardware identification and not for package installation? | 20:33 |
| unixpoet | it definitely does package stuff | 20:33 |
| josch | oh cool! | 20:33 |
| unixpoet | it integrates with packagekit just like apper/gnome-software, and you can build it with flatpak/appimage/snap support | 20:34 |
| unixpoet | I don't know about apper's wayland support, but I can say that discover definitely works in a wayland-first environment | 20:35 |
| + klardotsh (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 20:36 | |
| josch | and there is plasma-discover (also with packagekit backend) | 20:38 |
| unixpoet | is that not just https://apps.kde.org/discover/ ? | 20:38 |
| unixpoet | yeah, it is, just debian's name for the package | 20:38 |
| josch | ooooh | 20:39 |
| josch | that's why i was confused | 20:39 |
| josch | the package "discover" is really just "hardware identification system" and nothing kde specific :) | 20:39 |
| unixpoet | seems to be a debian-specific package, so that explains your confusion, yes | 20:40 |
| unixpoet | not debian-specific now that I look more closely *shrug* | 20:42 |
| josch | the "discover" package is some ancient thing from 23 years ago -- plasma-discove got added in 2016 and thus got a different name i guess | 20:43 |
| minute | josch: i tried gnome-software (without gnome) and it was just completely empty for some reason | 20:47 |
| - andreas-e (QUIT: Quit: Leaving) (~Andreas@2001:861:c4:f2f0:b0aa:56dd:ae73:8afa) | 20:48 | |
| + jacobk (~quassel@129.110.242.224) | 20:49 | |
| minute | kde's discover is also empty, so i guess i'm missing a package that populates those with debian's packages | 20:56 |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb91:dc6:e420:4e89:c540:3dc4:b5df) | 20:58 | |
| + stites (~stites@130.44.147.204) | 20:58 | |
| minute | ok, apper appears to work | 21:07 |
| minute | apper is cool, but fails to acquire authorization :( i wonder why, because pkexec works for me (pops up my lxpolkit agent) | 21:18 |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@129.110.242.224) | 21:20 | |
| erle | minute what's the synaptic error | 21:39 |
| minute | > Failed to initialize GTK. | 21:41 |
| minute | > Probably you're running Synaptic on Wayland with root permission. | 21:41 |
| minute | > Please restart your session without Wayland, or run Synaptic without root permission | 21:41 |
| minute | interestingly, /usr/sbin/synaptic kind of works | 21:41 |
| minute | erle: the problem is that the app is somehow built around being run as root. but actually they could just run the install action through pkexec instead of the whole GUI | 21:42 |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 21:43 | |
| minute | readme even mentions it https://github.com/mvo5/synaptic#usage | 21:43 |
| erle | oh yeah, i remember synaptic always had different gtk deco params | 21:44 |
| erle | obv because it was using root's settings | 21:44 |
| erle | minute, i don't really understand it though, does your “/usr/sbin/synaptic kind of works” mean that it does indeed use pkexec when it is not started as root? | 21:46 |
| minute | erle: unfortunately not. it just pops up a window at the start that says you won't be able to make changes because you're not root | 21:47 |
| minute | erle: and they ship synaptic-pkexec which is literally just: pkexec "/usr/sbin/synaptic" "$@" | 21:47 |
| erle | that reminds me of the guy who told me that the easiest way to fix a race condition happening in code handling minetest 5.4 and minetest 5.3 was to kick everyone with minetest below 5.5 from the server hehe | 21:47 |
| minute | (i just built synaptic from source, and no, it's still like that) | 21:47 |
| erle | “simplest thing that works” hits again | 21:48 |
| minute | the synaptic source code has a mix of tabs and spaces :( | 21:49 |
| erle | seems the synaptic source code is attacking you from multiple angles | 21:49 |
| erle | (or rather, attacking everyone. how obscene!) | 21:50 |
| erle | i'm curious, will look into it now | 21:50 |
| minute | i wanted to know if it calls apt or dpkg under the hood, but it seems to only rarely do that | 21:51 |
| erle | i thought the installation process actually does that? i mean it definitely does have console output | 21:51 |
| minute | hm yeah | 21:51 |
| minute | if(res == GTK_RESPONSE_YES) | 21:52 |
| minute | vte_terminal_feed_child(VTE_TERMINAL(_term), "y\n",2); | 21:52 |
| erle | minute, looks to me like the interesting code is all in libapt-pkg-dev | 21:57 |
| erle | or, at runtime, libapt-pkg | 21:57 |
| erle | i am investigating the results of “apt-get source libapt-pkg-dev” now | 21:58 |
| erle | uh, this is way too convoluted for me, right now | 22:00 |
| erle | > PackageManagerProgressFactory | 22:04 |
| + ggoes (~gregf@fsf/staff/ggoes) | 22:06 | |
| erle | i probably suck at C++, but … i did not find any single, obvious place in synaptic to insert the command to insert the authorization voodoo ritual dance and have it *only* affect non-GUI code | 22:07 |
| - ggoes (QUIT: Remote host closed the connection) (~gregf@fsf/staff/ggoes) | 22:08 | |
| - stites (QUIT: Ping timeout: 258 seconds) (~stites@130.44.147.204) | 22:08 | |
| Zaba | if it were easy, it would've been weird that nobody had already done it, therefore it would be only fair if it were not easy | 22:09 |
| + ggoes (~gregf@fsf/staff/ggoes) | 22:09 | |
| - ggoes (QUIT: Remote host closed the connection) (~gregf@fsf/staff/ggoes) | 22:10 | |
| erle | weird, but not necessarily true | 22:11 |
| - Boostisbetter (QUIT: Ping timeout: 258 seconds) (4a410829d7@irc.cheogram.com) | 22:11 | |
| + ggoes (~gregf@fsf/staff/ggoes) | 22:11 | |
| - jacobk (QUIT: Ping timeout: 246 seconds) (~quassel@utdpat241106.utdallas.edu) | 22:25 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 22:48 | |
| + jacobk (~quassel@129.110.242.224) | 22:58 | |
| - jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@129.110.242.224) | 23:12 | |
| - XYZ_ (QUIT: Quit: Leaving) (~XYZ@78-80-106-89.customers.tmcz.cz) | 23:17 | |
| + XYZ_ (~XYZ@89-24-41-114.nat.epc.tmcz.cz) | 23:23 | |
| + stites (~stites@2607:fb90:ad62:807c:d15b:ce72:ad88:3242) | 23:36 | |
| ex--parrot | hey josch, let me comment on the issue in gitlab | 23:54 |
| * ex--parrot -> ex-parrot | 23:54 | |
| josch | thank you! | 23:55 |
| ex-parrot | you're doing great work! I'd like to revisit how to hook the poweroff thing properly at some point also | 23:56 |
| ex-parrot | cool, added a comment | 23:56 |
| ex-parrot | I had a somewhat bad time trying to present off my Reform last week at a talk | 23:57 |
| ex-parrot | HDMI output was very flickery, and my USB serial adapter died (that part not the Reform's fault) | 23:57 |
| ex-parrot | but we mostly got there | 23:57 |
| ex-parrot | I was presenting on my work getting mainline Debian up and running on the PocketCHIP, which has got fairly far along and probably I need to document properly... | 23:59 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
