| - plomtest (QUIT: Read error: Connection reset by peer) (~plom@user/plomtest) | 00:37 | |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 00:50 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 01:17 | |
| - AnimaInvicta (PART: !!unknown attribute: msg!!) (~AnimaInvi@88-120-179-216.subs.proxad.net) | 01:18 | |
| - Gooberpatrol66 (QUIT: Client Quit) (~Gooberpat@user/gooberpatrol66) | 01:21 | |
| + kfx (~kfx@wopr.sciops.net) | 01:27 | |
| kfx | I'm trying to track down a bug. I have the latest mnt reform keyboard v4 firmware flashed. When I power on the machine, the backlight comes on, but it turns back off sometime during u-boot, before the kernel is loaded. I can turn it back on in the os with an init script, but in the meantime to type my LUKS passphrase, I have to manually turn the backlight back on. | 01:29 |
|---|---|---|
| kfx | is this bug in the keyboard firmware or is it in u-boot? | 01:29 |
| kfx | this is with the rk3588 core. | 01:29 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 01:37 | |
| - Guest5591 (QUIT: Ping timeout: 244 seconds) (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 01:39 | |
| gordon1 | kfx: i noticed similar behaviour but only if i try to make backlight to be as dim as possible, the trick for me is to set the level slightly above the absolute minimum | 01:47 |
| gordon1 | just keep in mind that backlight adjustment on keyboard is "analog" of sort, i.e. it does not adjust it by increments but by the timing on how long you press the button for (or maybe some combination of both? didn't actually check the code) | 01:49 |
| + Guest5591 (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 01:52 | |
| kfx | gordon1: I altered the firmware to put it on full-blast at init, but at some point between that and u-boot loading the kernel it turns back off. turning it back on works, but I don't want it to turn off at all | 02:03 |
| kfx | I suspect u-boot is doing usb shenanigans that cause this but u-boot is a labyrinthian nightmare world so I can't figure out where | 02:04 |
| gordon1 | right, that doesn't look like problem i encountered then | 02:04 |
| + robin (~robin@user/terpri) | 02:15 | |
| - paperManu (QUIT: Quit: WeeChat 4.1.1) (~paperManu@64.58.44.160) | 02:23 | |
| + paperManu (~paperManu@64.58.44.160) | 02:39 | |
| - paperManu (QUIT: Ping timeout: 264 seconds) (~paperManu@64.58.44.160) | 02:51 | |
| - mjw (QUIT: Ping timeout: 244 seconds) (~mjw@195.23.216.83) | 03:23 | |
| - nsc (QUIT: Ping timeout: 256 seconds) (~nicolas@247-96-142-46.pool.kielnet.net) | 03:50 | |
| + nsc (~nicolas@82-98-142-46.pool.kielnet.net) | 03:51 | |
| - timide (QUIT: K-Lined) (~timide@user/timide) | 04:19 | |
| + jogu (~jogu@user/jogu) | 04:55 | |
| + plomtest (~plom@user/plomtest) | 11:35 | |
| minute | kfx: keyboard backlight? | 12:26 |
| gordon1 | oh, it's probably screen backlight | 12:27 |
| gordon1 | that makes more sense | 12:27 |
| gordon1 | no? | 12:28 |
| minute | the part about uboot and usb suggests keyboard though... but uboot can't mess with the backlight of the keyboard. | 12:32 |
| minute | kfx: what kind of init script are you using and how exactly did you change the firmware? (.diff?) | 12:33 |
| + paperManu (~paperManu@64.58.44.160) | 12:40 | |
| sigrid | minute: something of note - if I have the acryl lid screwed fully, the top row of the keyboard stops working. likewise, ethernet keeps breaking | 13:41 |
| sigrid | this is classic reform with rk3588 module. I think the thing is way too thick to git | 13:41 |
| sigrid | *fit | 13:41 |
| sigrid | I had to remove the two screws right where the heatsink is. I can see the lid is pushed out by the heatsink | 13:43 |
| gordon1 | sigrid: i definitely have a gap between heatsink and the lid, judging by cookie crumbs stuck between, do you have latter copper heatsink plate or some other version? | 13:49 |
| sigrid | it's the copper one | 13:50 |
| gordon1 | i also remember when i did my thermal experiments, i fitted something incorrectly and lid screws were very tight so i unscrewed and re-seated the heatsink and it helped | 13:50 |
| josch | here is somebody on the forum who lost USB with the heatsink on: https://community.mnt.re/t/new-heatsink-mysteriously-interferes-with-usb/3781 | 13:50 |
| gordon1 | i have no proof that it was interference between lid and heatsink but i think i noticed a bulge | 13:50 |
| josch | and keyboard bulging here: https://community.mnt.re/t/keyboard-bulges-out-with-rk3588/3473 | 13:51 |
| josch | i have rk3588 classic reform with copper heatsink here as well but no bulging | 13:51 |
| gordon1 | oh wow, that's really interesting | 13:51 |
| josch | i think there was also a re-design of the copper heatsink without fins but i don't know if that gets shipped yet | 13:52 |
| gordon1 | maybe my hypothesis about DMA issue caused by mechanical stress in connectors isn't completely unfounded after all (btw it didn't reappear for a long while after my last heatsink fiddling) | 13:53 |
| gordon1 | >the heatsink of the RK3588 module sits flush with the aluminum case’s lid. | 13:53 |
| gordon1 | mine btw has ~0.2mm or so gap that is noticeable where heatsink and the case meet | 13:54 |
| gordon1 | a little step | 13:54 |
| gordon1 | cookie crumbs don't lie | 13:55 |
| - paperManu (QUIT: Ping timeout: 248 seconds) (~paperManu@64.58.44.160) | 14:13 | |
| + mjw (~mjw@89.205.242.167) | 14:18 | |
| + paperManu (~paperManu@modemcable141.205-200-24.mc.videotron.ca) | 14:46 | |
| - mjw (QUIT: Ping timeout: 240 seconds) (~mjw@89.205.242.167) | 15:10 | |
| ch | filed this https://gitlab.com/fwupd/lvfs-website/-/issues/1037 | 15:15 |
| + mjw (~mjw@213.58.148.170) | 15:57 | |
| - mjw (QUIT: Quit: Leaving) (~mjw@213.58.148.170) | 16:12 | |
| + mjw (~mjw@213.58.148.170) | 16:12 | |
| + wielaard (~mjw@213.58.148.170) | 16:17 | |
| - mjw (QUIT: Ping timeout: 264 seconds) (~mjw@213.58.148.170) | 16:19 | |
| - xktr (QUIT: Ping timeout: 244 seconds) (~xktr@user/xktr) | 16:20 | |
| + xktr (~xktr@user/xktr) | 16:21 | |
| - wielaard (QUIT: Ping timeout: 240 seconds) (~mjw@213.58.148.170) | 16:21 | |
| - paperManu (QUIT: Ping timeout: 240 seconds) (~paperManu@modemcable141.205-200-24.mc.videotron.ca) | 17:05 | |
| + paperManu (~paperManu@modemcable141.205-200-24.mc.videotron.ca) | 17:07 | |
| minute | sigrid: which version of the copper you got? | 17:23 |
| minute | sigrid: this one or more complex bigger one? https://shop.mntre.com/spree/products/593/large/reform-rcore-v2-module_heatsink.jpg?1752073321 | 17:23 |
| sigrid | more complex one, minute | 17:53 |
| gordon1 | aah, i see, i have v2 it seems | 18:08 |
| minute | sigrid: ok we should send you the simple one | 18:09 |
| minute | sigrid: it's a bit thinner and should solve this issue | 18:09 |
| minute | josch: btw i think i missed it here, what happened when you tried the ssd install thing in the wizard? (i.e. that prompted you to change system image to non-ssd-page wizard) | 18:11 |
| josch | minute: https://community.mnt.re/t/installer-crashing-instead-on-ssd-step/3893 | 18:20 |
| josch | minute: reverting to a known-good state took less time than investigating what broke now (because i remember that we had fixed at least part of what was reported there) and my goal was to get something working out quickly and not delay with however long it takes to debug it | 18:21 |
| josch | now that the default system image works again (see the last message in the thread) we have time to figure out what went wrong | 18:22 |
| minute | josch: ahaa, thanks for taking care of that btw! | 18:28 |
| josch | minute: also, i did not do a simple revert and cherry-picked changes like switching from wayfire to gnome -- details here: https://source.mnt.re/reform/reform-system-image/-/merge_requests/140 | 18:29 |
| josch | err sorry | 18:29 |
| josch | wrong link | 18:29 |
| josch | minute: details here: https://source.mnt.re/reform/mnt-reform-setup-wizard/-/merge_requests/17 | 18:29 |
| + AnimaInvicta (~AnimaInvi@88-120-179-216.subs.proxad.net) | 18:36 | |
| kfx | minute: I just flashed an unmodified keyboard firmware from git tip and I get the same behavior (and yeah it's the keyboard backlight that turns off) | 18:39 |
| minute | kfx: ok can you tell me again about the rest of your system? rk3588 classic reform, mnt debian system image? | 18:44 |
| kfx | I'm running alpine linux on mine, but yeah rk3588 classic reform, original motherboard I got during the first order run, keyboard v4 | 18:45 |
| minute | kfx: does this happen if you run the standard current system image from microSD card? | 18:46 |
| minute | i mean SD card | 18:46 |
| kfx | my init script just echos to hidraw0 to turn on the LEDs | 18:46 |
| kfx | can't remember (been a while since I ran debian) but I'll test that next | 18:46 |
| wickedshell | minute: I'm having fomo woes about the RK upgrade on crowdsupply. Can I ask how much it's going up by on the 1st? I was perfectly happy with the imx8 in my pocket till I started doing some work in rust on it.... Recompiles aren't too bad, but any fresh compile/clippy run is agonizing :P | 18:46 |
| minute | wickedshell: oh yeah, rust is a lot more fun on the rk3588 :D unfortunately idk about the exact price change. probably not _that_ much | 18:47 |
| wickedshell | minute: right? When I was doing ocaml work on it it was totally fine, but then rust is just like "nope, experience pain!". Also a little worried about battery life tradeoffs, but at the same time faster compiles mean actually getting more work done, so maybe I should just accept it :) | 18:48 |
| josch | ah drats, doing a battery life comparison between rk3588 and imx8mq was on my todo list but it's a bit late now to get this done before oct 1st XD | 18:53 |
| ch | pff, more than 24h to go! | 18:54 |
| wickedshell | lol | 18:58 |
| + reform5722 (~rene@5.42.207.7) | 19:11 | |
| * reform5722 -> _rwa | 19:11 | |
| * _rwa -> rwa_ | 19:11 | |
| - rwa_ (QUIT: Client Quit) (~rene@5.42.207.7) | 19:13 | |
| ch | minute: fyi i'm trying to debug the lvfs uploads, so you'll see some notification mails | 19:37 |
| minute | ch: ok cool! | 19:39 |
| - shtrophic (QUIT: Remote host closed the connection) (~m-hrdsqi@user/shtrophic) | 21:15 | |
| + shtrophic (~m-hrdsqi@user/shtrophic) | 21:17 | |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 22:13 | |
| - natalie (QUIT: Ping timeout: 245 seconds) (~natalie@user/natalie) | 22:31 | |
| + natalie (~natalie@user/natalie) | 22:32 | |
| - helgoman__ (QUIT: Quit: Konversation terminated!) (~helgoman@178.38.121.62) | 22:35 | |
| kfx | minute: replicated the behavior with the system image on an sd card: http://sciops.net/downloads/keyboard.mp4 | 22:55 |
| kfx | minute: the above video is booting alpine from nmve, power off, then inserting the sd card and booting from that. you can see the keyboard comes alight, then extinguishes before the kernel takes over | 22:56 |
| - paperManu (QUIT: Ping timeout: 264 seconds) (~paperManu@modemcable141.205-200-24.mc.videotron.ca) | 23:13 | |
| + paperManu (~paperManu@64.58.44.160) | 23:22 | |
| + mjw (~mjw@gnu.wildebeest.org) | 23:26 | |
| - paperManu (QUIT: Quit: WeeChat 4.1.1) (~paperManu@64.58.44.160) | 23:27 | |
| + paperManu (~paperManu@64.58.44.160) | 23:36 | |
| + dormito (~dormito@user/dormito) | 23:43 | |
| dormito | regarding the rk3588 module: is there a on-module boot storage device, or does it boot purely off the carrier? | 23:57 |
| dormito | (to clarify, I'm refering to where Arm trusted firmware/uboot/etc are stored, not the linux kernel/initramfs) | 23:58 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
