| ephase | good night! | 00:12 |
|---|---|---|
| - ephase (QUIT: Quit: WeeChat 4.0.5) (~ephase@2a01:e0a:2a:5300:f8d4:1fb6:b769:4362) | 00:12 | |
| - stites (QUIT: Ping timeout: 260 seconds) (~stites@130.44.147.204) | 00:29 | |
| + stites (~stites@2607:fb91:dc3:1267:8194:4057:3670:88d7) | 00:29 | |
| - mjw (QUIT: Ping timeout: 255 seconds) (~mjw@gnu.wildebeest.org) | 00:33 | |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb91:dc3:1267:8194:4057:3670:88d7) | 01:05 | |
| + stites (~stites@130.44.147.204) | 01:05 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:18 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 01:55 | |
| - GNUmoon (QUIT: Client Quit) (~GNUmoon@gateway/tor-sasl/gnumoon) | 01:57 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:02 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 02:19 | |
| - eery (QUIT: Ping timeout: 240 seconds) (~eery@139.55.3.210) | 02:53 | |
| + eery (~eery@139.55.3.210) | 02:54 | |
| - eery (QUIT: Client Quit) (~eery@139.55.3.210) | 02:58 | |
| + eery (~eery@172.97.103.152) | 02:58 | |
| + bkeys1 (~Thunderbi@134.22.115.162) | 03:09 | |
| - bkeys (QUIT: Ping timeout: 272 seconds) (~Thunderbi@199.226-24.cm.ptn.tftn.dynamic.friendlycity.net) | 03:13 | |
| * bkeys1 -> bkeys | 03:13 | |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@201-98-142-46.pool.kielnet.net) | 04:01 | |
| + nsc (~nicolas@129-97-142-46.pool.kielnet.net) | 04:02 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 04:08 | |
| - jacobk (QUIT: Ping timeout: 272 seconds) (~quassel@utdpat241106.utdallas.edu) | 04:50 | |
| + jacobk (~quassel@64.189.201.150) | 05:11 | |
| - klardotsh (QUIT: Ping timeout: 245 seconds) (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 08:25 | |
| vkoskiv | Any recommendations for a good bluetooth management UI? AirPods work great with the A311D Reform, but since switching to linux, I've just been using bluetoothctl | 08:33 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:37 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:38 | |
| josch | vkoskiv: blueman and its applet? | 08:54 |
| vkoskiv | I think I tried that at some point, but moved back to using bluetoothctl. But now I forget why, so I'll give it another go. | 09:00 |
| vkoskiv | Maybe I was just going through a TUI phase :] | 09:01 |
| minute | vkoskiv: blueman https://mastodon.social/@mntmn/111182380093922973 | 09:24 |
| vkoskiv | Ah, nice, that looks very nice | 09:24 |
| vkoskiv | Did I just say nice twice? I guess I haven't woken up fully yet | 09:25 |
| - mark_ (QUIT: Ping timeout: 258 seconds) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 09:39 | |
| minute | double nice | 10:01 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:54 | |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 11:47 | |
| - mark_ (QUIT: Quit: Leaving) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 12:12 | |
| josch | 20 files changed, 347 insertions(+), 207 deletions(-) | 12:17 |
| josch | anybody willing to thrash their system with another update of reform-tools? | 12:18 |
| josch | i'm going to put my ls1028a back in now to do tests that do not affect my working system | 12:18 |
| josch | these are mostly changes that make reform-tools work on the other SoMs | 12:18 |
| vkoskiv | I'll give it a go after work | 12:19 |
| josch | it would be sweet to have an a311d user test this :) | 12:19 |
| josch | because half the scripts have a "don't touch emmc" in them for a311d ;) | 12:19 |
| vkoskiv | Is it in the repo, or do I build from your branch? | 12:20 |
| josch | it is very much not in the repo because there will be bugs! | 12:20 |
| josch | i'm only about to test everything now | 12:20 |
| josch | if you are only testing things later, i'm going to test first, weed out some of the bugs and then create a MR | 12:21 |
| minute | josch: solution for flicker is here https://mastodon.social/@mntmn/111204608407114964 | 12:37 |
| josch | oh lol | 12:38 |
| josch | and i hoped i could already get rid of it by removing my wifi card, for example :D | 12:39 |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 12:41 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae))) (~mjw@gnu.wildebeest.org) | 12:42 | |
| minute | josch: we will send you a ferrite, just need to decide the best location & attach it (probably with tape) | 12:42 |
| * mark_ -> mjw | 12:42 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 12:42 | |
| josch | minute: it's not urgent because i'm only doing tests right now and need to switch to imx during the semester aryways because of hdmi output | 12:43 |
| josch | minute: maybe just put it into the box when i order my second mainboard | 12:43 |
| minute | josch: ok! | 12:54 |
| + reform1962 (~josch@ip1f12fe3b.dynamic.kabel-deutschland.de) | 13:13 | |
| reform1962 | hah... of course the disk is not named | 13:13 |
| reform1962 | dev/nvme on ls1028a | 13:14 |
| reform1962 | but yes i have a /dev/sda | 13:14 |
| - marty_mcfly88 (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@2603-6000-8c01-0f3e-7dac-14c7-19c1-fd5c.res6.spectrum.com) | 13:19 | |
| + marty_mcfly88 (~quassel@2603-6000-8c01-0f3e-7dac-14c7-19c1-fd5c.res6.spectrum.com) | 13:19 | |
| - marty_mcfly88 (QUIT: Client Quit) (~quassel@2603-6000-8c01-0f3e-7dac-14c7-19c1-fd5c.res6.spectrum.com) | 13:21 | |
| + marty_mcfly88 (~quassel@2603-6000-8c01-0f3e-7dac-14c7-19c1-fd5c.res6.spectrum.com) | 13:21 | |
| - GNUmoon (QUIT: Ping timeout: 252 seconds) (~GNUmoon@gateway/tor-sasl/gnumoon) | 13:21 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 13:33 | |
| - mtm (QUIT: Ping timeout: 240 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:02 | |
| - reform1962 (QUIT: Remote host closed the connection) (~josch@ip1f12fe3b.dynamic.kabel-deutschland.de) | 14:12 | |
| josch | great, ls1028a boots just fine with /boot and / on emmc :) | 14:24 |
| josch | installing the heatsink was much easier this time but it's also obvious that the holes now have quite some wear and one can see the thread that ate into the hole borders | 14:26 |
| josch | at least the screws were stronger than the pcb and not tor way round | 14:26 |
| josch | now i'm only afraid to braking the fragile display cable | 14:26 |
| josch | i think the next operation i'll be doing exclusively via serial | 14:26 |
| josch | minute: when was the last time you tried reform-flash-rescue on imx8mq? The LABEL=reformsd* change might've broken reform-boot-config as findmint --evaluate comes back empty. Maybe I got this because there were still bugs in my tools (i'm fixing those right now) but having another person confirm whether it works for them or not would be helpful to figuring out what i just had happen on ls1028a. | 14:37 |
| minute | sorry, we have some issues with some of the ls1028a modules here (won't boot from sd) and i'm scrambling to understand what's up. | 14:52 |
| josch | oh dear :( | 14:53 |
| josch | then i'll just be content that mine boots just fine XD | 14:54 |
| - jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@64.189.201.150) | 14:56 | |
| minute | yeah! | 14:59 |
| josch | well then, once you have a lot of free time again: https://source.mnt.re/reform/reform-tools/-/merge_requests/58/diffs | 15:02 |
| josch | i'll try again myself as well but not very soon because it takes a long time to swap modules around | 15:03 |
| josch | vkoskiv: that MR contains the stuff you can try out if you like, thank you! | 15:03 |
| josch | worst bugs already fixed :) | 15:03 |
| + jacobk (~quassel@129.110.242.224) | 15:09 | |
| minute | 40% failure rate so far with ls1028a modules | 15:41 |
| minute | i hope i can still find out where this comes from | 15:41 |
| vkoskiv | Ouch. Any clues yet? | 15:42 |
| [tj] | ouch, I'm sorry to hear that | 15:42 |
| minute | vkoskiv: nope | 15:42 |
| josch | wow that sucks so much :( | 15:52 |
| josch | will you get any money back in case it turns out to be a manufacturing problem? | 15:53 |
| q66 | minute: failure in what way? | 15:53 |
| q66 | the hardware dying? | 15:53 |
| + murphnj (~murph@user/murphnj) | 15:54 | |
| minute | josch: doubtful | 15:57 |
| minute | q66: no, not booting or rarely booting | 15:57 |
| minute | sometimes only 100khz clock on sd card, not going to high speed | 15:57 |
| q66 | ah | 15:58 |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@129.110.242.224) | 16:00 | |
| q66 | sounds like something on the module then | 16:02 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:08 | |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 16:12 | |
| minute | q66: yip | 16:12 |
| q66 | with a bit of luck you'll solve that with a resistor somewhere or something :p | 16:14 |
| vkoskiv | josch: Maybe bug, reform-check from ~a week ago suggested I delete /etc/flash-kernel/preboot.d/00reform2_preboot, but running it again after deleting notifies that it's missing | 17:07 |
| vkoskiv | Is the intent of 'consider deleting X in favour of Y' to replace X with Y? | 17:08 |
| josch | vkoskiv: i don't see where it warns if /etc/flash-kernel/preboot.d/00reform2_preboot does *not* exist -- what does it print? | 17:35 |
| josch | vkoskiv: yes | 17:35 |
| vkoskiv | missing c /etc/flash-kernel/preboot.d/00reform2_preboot | 17:38 |
| vkoskiv | Last line in the printout is that | 17:38 |
| vkoskiv | So I should copy the file from /usr/share to replace that file? | 17:39 |
| vkoskiv | unrelated: With my kbd backlight fix, the PWM frequency is now actually in the human hearing range at 3906 hertz, I can hear a faint tone from the backlight chip with the brightness cranked up! :D | 17:40 |
| vkoskiv | I'll see if I can push up the frequency higher, without going over the 20kHz limit | 17:40 |
| vkoskiv | Or lower, too, as the ti chip isn't very acoustically performant at low frequencies, I bet. | 17:41 |
| vkoskiv | And, I also noticed that the lowest brightness isn't actually quite fully off, I'll fix that too. | 17:43 |
| vkoskiv | Only noticed these in a dark and quiet room | 17:43 |
| vkoskiv | I'll kick it down to a /64 prescale, that should take it down to ~488Hz | 17:46 |
| - stites (QUIT: Ping timeout: 260 seconds) (~stites@130.44.147.204) | 17:59 | |
| + stites (~stites@2607:fb91:dcf:5265:c7a5:8dfc:6762:14b0) | 18:00 | |
| - stites (QUIT: Read error: Connection reset by peer) (~stites@2607:fb91:dcf:5265:c7a5:8dfc:6762:14b0) | 18:24 | |
| + stites (~stites@130.44.147.204) | 18:25 | |
| vkoskiv | That makes it blink, for reasons I do not understand. Valid PWM at a valid frequency, should work. | 18:47 |
| vkoskiv | ti spec for the backlight chip does mention a pull-down resistor of 800kohm for the PWM line, I don't see that in the schematic, but I have no idea if that would matter | 19:05 |
| vkoskiv | I know I'm putting in pwm at valid frequencies, but at 976 and 7812 hertz it blinks, when I think it shouldn't | 19:05 |
| vkoskiv | Correction, at 7812 it works, but there is one bigger jump between 2 and 3 on the brightness level. | 19:06 |
| vkoskiv | So my current setting is the best so far, but that 3kHz whine is irritating, now that I noticed it | 19:06 |
| vkoskiv | I couldn't find a prescale+mode permutation where it works better than this, so far. | 19:07 |
| + ajr (uid609314@user/ajr) | 19:14 | |
| sigrid | could have static setting for each brightness level | 19:14 |
| sigrid | and choose those by trial | 19:14 |
| minute | yeah | 19:14 |
| + klardotsh (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 19:17 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:27 | |
| * mark_ -> mjw | 19:28 | |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:28 | |
| josch | vkoskiv: you found a bug, thank you! | 19:28 |
| josch | vkoskiv: what do you get when you run this: dpkg --verify reform-tools | 19:31 |
| Boostisbetter | well it looks like my Reform after a recent update was hit with the failed lcd_controller connect issue. The Reform boots, but acts real weird, and can't hold a wifi connection. Why oh why did I update? | 19:51 |
| minute | i don't seem to be able to figure out why some of the ls1028a can't reliably boot. so i think i'm giving up on these. lets hope that enough modules of the 36 remaining from the batch will work so we can ship 'em, and then that's the end of this path | 20:27 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.32.204) | 20:28 | |
| Boostisbetter | minute: I am on the normal imx8 | 20:29 |
| Boostisbetter | if you were meaning me as well | 20:29 |
| minute | Boostisbetter: i was talking about something completely different. | 20:29 |
| + Ar|stote|is (~linx@149.210.32.204) | 20:29 | |
| minute | yield issues with ls1028a | 20:30 |
| minute | Boostisbetter: so you can't boot? how do you know about any lcd controller issue then? via serial? | 20:31 |
| minute | Boostisbetter: or does the display work and is just stuck? | 20:31 |
| - jacobk (QUIT: Ping timeout: 272 seconds) (~quassel@utdpat241106.utdallas.edu) | 20:37 | |
| - ndufresne (QUIT: Quit: The Lounge - https://thelounge.chat) (~ndufresne@apple.collaboradmins.com) | 20:45 | |
| Boostisbetter | minute: the system boots and I can get into sway. Both wifi will not connect, or if it does it disconnects almost immediately. I have visual glitching that is significant. | 20:47 |
| + ndufresne (~ndufresne@apple.collaboradmins.com) | 20:47 | |
| Boostisbetter | the lcd controller error is what I see on boot. Maybe it is a recent change meant to work with it, but my oled is still on old firmware. | 20:48 |
| minute | hmm, these are a bunch of unrelated things. | 20:49 |
| minute | linux does not know anything about the oled, for example. | 20:49 |
| Boostisbetter | was a real bummer, because before I updated, I was on my 100+ resume from suspend, and the system had been running perfectly. | 20:49 |
| minute | what kind of glitching do you get? can you make screenshots/video? | 20:49 |
| minute | Boostisbetter: and can you share the output of `cat /proc/cmdline` ? | 20:50 |
| Boostisbetter | i will try to in about 5 minutes | 20:51 |
| + jacobk (~quassel@utdpat241040.utdallas.edu) | 20:52 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@utdpat241040.utdallas.edu) | 20:58 | |
| + jacobk (~quassel@utdpat241040.utdallas.edu) | 20:59 | |
| Boostisbetter | https://upload.im.boosterfive.com:5281/upload/wrQwsovMZwOf5ylNfESLYCw-/85vSuTmtSVGS6ZJPK8yrYw.jpg | 21:02 |
| Boostisbetter | that was during boot | 21:02 |
| Boostisbetter | The product of cat is root=/dev/nvme0n1p1 ro quiet | 21:03 |
| Boostisbetter | now that I am logged in, the biggest issue is huge graphical glitches on sway, such that the system doesn't seem to be responsive. I mean I can change workspaces but things don't seem to work on them. | 21:09 |
| - jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@utdpat241040.utdallas.edu) | 21:10 | |
| Boostisbetter | VS Codium, for examples, open and it is just dark grey background and nothing every shows. However, when moving the cursor the cursor icons changes as if over text. So I think things are working. The display is just really hosed. | 21:11 |
| + jacobk (~quassel@utdpat241040.utdallas.edu) | 21:12 | |
| - Boostisbetter (QUIT: Ping timeout: 252 seconds) (4a410829d7@irc.cheogram.com) | 21:32 | |
| - ethulhu (QUIT: Excess Flood) (ethulhu@nora.ethulhu.co.uk) | 21:40 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 21:44 | |
| minute | Boostisbetter: your cmdline is missing important things, that's the problem | 21:44 |
| minute | Boostisbetter: can you pastebin the contents of your /boot/boot.scr please? this should shed some light on where this came from | 21:46 |
| - jacobk (QUIT: Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@utdpat241040.utdallas.edu) | 21:47 | |
| + ethulhu (ethulhu@nora.ethulhu.co.uk) | 21:50 | |
| - Boostisbetter (QUIT: Ping timeout: 240 seconds) (4a410829d7@irc.cheogram.com) | 21:50 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 21:59 | |
| Boostisbetter | https://pastebin.com/cYkJwmbw | 21:59 |
| Boostisbetter | also there a bak file from today there, should I just restore that? | 21:59 |
| vkoskiv | josch: with that dpkg command I get: | 22:08 |
| vkoskiv | ??5?????? /var/lib/alsa/asound.state | 22:08 |
| vkoskiv | missing c /etc/flash-kernel/preboot.d/00reform2_preboot | 22:08 |
| minute | Boostisbetter: hmm, i don't see how the cmdline you wrote earlier could be produced by this boot.scr | 22:11 |
| vkoskiv | sigrid: I think I'll just have to read the ti chip doc more closely to understand why it's not behaving the way I expect | 22:12 |
| vkoskiv | It's not very long anyway | 22:12 |
| minute | Boostisbetter: for example, "root=nvme0n1p1" does not appear in this file | 22:12 |
| Boostisbetter | The system was shutdown with a power out as it hung. | 22:12 |
| Boostisbetter | Is there a command I can run to get the right cmdline to populate? | 22:13 |
| vkoskiv | minute: if you happen to remember, is the max brightness of the keyboard roughly the max current the backlight chip can supply? I haven't checked if the keyboard blinking is just exceeding the max current, or something wrong with the pwm signal | 22:13 |
| minute | Boostisbetter: are you booting via sd card? | 22:13 |
| minute | Boostisbetter: normally this boot.scr should do the right thing! so something is set up differently than expected around your uboot or boot partition | 22:13 |
| minute | Boostisbetter: maybe there's extlinux.conf on your /boot ? | 22:14 |
| sigrid | vkoskiv: you have an oscilloscope or some kind of logic analyzer? | 22:14 |
| sigrid | it could help | 22:14 |
| minute | vkoskiv: sorry, idk right now | 22:14 |
| Boostisbetter | minute: yes booting through the sd card, but then running off the nvme | 22:14 |
| vkoskiv | I've been wanting to buy a scope for some time, this project is *not* helping :D | 22:14 |
| vkoskiv | minute: yeah, no worry, I'll read the doc and find the answer | 22:14 |
| minute | Boostisbetter: yep, ok, but something is non-standard with your uboot or /boot. can you list the contents of /boot (first partition of the sd card)? is there extlinux.conf by any chance? | 22:15 |
| vkoskiv | I've been reading a lot of technical docs lately, for various diferent things. | 22:15 |
| minute | Boostisbetter: or perhaps you have some kind of custom uboot build? | 22:15 |
| vkoskiv | sigrid: Would certainly help if I could actually see the pwm signal. But I have a pretty strong confirmation that at least the frequency is what I expect - I can hear the chip whine, and it matches the predicted frequency well. | 22:16 |
| minute | Boostisbetter: 1. check if there is extlinux.conf on your /boot, if so, move it aside 2. try to restore to a known good uboot by running `reform-flash-uboot sd` as root | 22:17 |
| vkoskiv | It's surprising to me that the supplier sent modules with a 40% defect rate, and you can't get reimbursement/replacements? Or would it just take months? | 22:18 |
| Boostisbetter | minute: the boot folder on the sd card has nothing in it. | 22:18 |
| vkoskiv | Either way, sucky situation, I really hope you find some reliable fix | 22:18 |
| minute | vkoskiv: there's no supplier. i'm kind of the supplier | 22:18 |
| minute | Boostisbetter: not the folder. the first partition | 22:18 |
| vkoskiv | But some company supplied them to you, no? | 22:18 |
| minute | vkoskiv: well, multiple companies are involved. | 22:19 |
| vkoskiv | Well, either way, you know the situation the best | 22:19 |
| Boostisbetter | minute: but on the nvme in boot there is a extlinux directory | 22:19 |
| minute | vkoskiv: there's a company who soldered the modules | 22:19 |
| vkoskiv | Ah, so they are custom made | 22:19 |
| minute | vkoskiv: yes | 22:19 |
| Boostisbetter | on weird thing is that I have a 1tb encrypted and free thing that is mountable. In it there is the pocket firmware. | 22:20 |
| Boostisbetter | 1_1.bin | 22:20 |
| minute | Boostisbetter: i can't follow. | 22:21 |
| minute | Boostisbetter: do you have your reform booted at the moment and the first partition of sd card is mounted as /boot? | 22:21 |
| Boostisbetter | Yes it is booted. By mounted, I have thunar open and I have clicked on the sd card | 22:22 |
| Boostisbetter | I am able to look at the contents of the sd card through there. | 22:22 |
| minute | Boostisbetter: sorry, that way i can't understand what is going on. can you use a terminal please? and type lsblk | 22:23 |
| minute | Boostisbetter: do you see the device mmcblk1p0 and is it mounted at /boot ? | 22:23 |
| Boostisbetter | https://upload.im.boosterfive.com:5281/upload/KZHMYHnkEQy6MSUgEWLTNz_R/Ae31EwdQQJiDz3f7YXRcmQ.jpg | 22:23 |
| minute | Boostisbetter: i can't see what i need to see there, sorry | 22:23 |
| Boostisbetter | https://upload.im.boosterfive.com:5281/upload/tFMXm-9NeWq1pnnK7lrebEBY/Yem9NJwfTeGxTTNdreILzA.jpg | 22:24 |
| Boostisbetter | so no, I don't see that the sd card is mounted in the way you would like. | 22:25 |
| Boostisbetter | however, the MNTREFORMROOT is mounted under /media/committed88/ | 22:25 |
| minute | Boostisbetter: but the photo shows that the sd card is mounted on /boot | 22:25 |
| minute | Boostisbetter: the line with mmcblk1p1 shows /boot | 22:26 |
| Boostisbetter | ok, you just mentioned mmcblk1p0 | 22:26 |
| Boostisbetter | threw me off | 22:26 |
| minute | Boostisbetter: indeed! sorry for that | 22:26 |
| Boostisbetter | ok, well in that case, I have the extlinux directory in /boot with the extlinux.conf file in it | 22:27 |
| Boostisbetter | so am I at the point now where I can use the reform tool? | 22:27 |
| minute | Boostisbetter: delete that directory | 22:27 |
| minute | rm -rf /boot/extlinux | 22:28 |
| Boostisbetter | done | 22:28 |
| Boostisbetter | although without the f flag, as i did it before seeing that | 22:29 |
| minute | Boostisbetter: ok, please try to reboot! | 22:29 |
| Boostisbetter | ok no need to force it with the tool, just see if it works better on reboot? | 22:29 |
| minute | yeah, lets try that first | 22:30 |
| josch | Boostisbetter: if your extlinux.conf is causing you trouble, you also need to uninstall u-boot-menu or otherwise it will be regenerated on the next kernel upgrade | 22:32 |
| minute | ah, u-boot-menu is where that comes from, i didn't have that memorized | 22:32 |
| josch | vkoskiv: thank you, what you see there is a leftover config file from a very old reform-tools version | 22:32 |
| Boostisbetter | I was really liking being able to select the kernel on boot | 22:33 |
| minute | ah, that's the custom thing you have that i'm not familiar with | 22:33 |
| Boostisbetter | but so far the reboot fixed it. Didn't get to select the kernel, but it booted normal and wifi is working. No visual glitches either. | 22:33 |
| josch | vkoskiv: essentially, this can be fixed by reform-tools itself deleting that file in /etc because it got moved to /usr | 22:33 |
| josch | Boostisbetter: if you want to use u-boot-menu, make sure that /boot/extlinux/extlinux.conf contains the correct "append" values | 22:34 |
| josch | Boostisbetter: do you have a /etc/default/u-boot? | 22:34 |
| Boostisbetter | josch: actually now that 6.5 has fixed suspended for real, I guess I don't really need it anymore. | 22:34 |
| josch | okay | 22:35 |
| Boostisbetter | josch: but to answer your question, yes there is a u-boot file in there | 22:35 |
| josch | what does it contain? | 22:35 |
| josch | it should contain this line: | 22:36 |
| josch | U_BOOT_PARAMETERS="ro no_console_suspend cma=512M pci=nomsi console=ttymxc0,115200 console=tty1" | 22:36 |
| minute | josch: how could we make the update process robust against this problem? or maybe it's enough to mention that currently we're incompatible with having u-boot-menu installed? | 22:36 |
| josch | it was never supported | 22:37 |
| josch | because there is no common cmdline | 22:37 |
| minute | hmm, but Boostisbetter was booting with it before | 22:37 |
| josch | and u-boot-menu ignores ${bootargs} from u-boot | 22:37 |
| Boostisbetter | minute: thank you very much, especially so late, for helping me rescue my Reform. | 22:37 |
| minute | and certainly had a good cmdline | 22:37 |
| minute | Boostisbetter: you're welcome, i'm happy it is fixed | 22:37 |
| josch | indeed | 22:37 |
| Boostisbetter | josch: that line is absolutely not there. I think an update replaced the files and removed those options and this is why u-boot was an issue. | 22:37 |
| josch | it's a file in /etc | 22:38 |
| Boostisbetter | because before the update it worked great. | 22:38 |
| josch | reform-tools stopped messing with files in /etc a long while ago | 22:38 |
| josch | precisely for this reason | 22:38 |
| josch | vkoskiv is seeing some leftovers still | 22:38 |
| Boostisbetter | although sometimes the screen would flicker until the kernel started tro boot. | 22:38 |
| josch | Boostisbetter: if you want u-boot-menu back, add that line to your /etc/default/u-boot and make sure that the generated /boot/extlinux/extlinux.conf contains those values in "append" | 22:39 |
| Boostisbetter | I will keep that in mind in case we have a kernel regression affecting suspend again, but for now I'll just not use u-boot. | 22:41 |
| Boostisbetter | Should I uninstall it from the system, or without the folder in boot it wont get started anyway? | 22:41 |
| minute | i don't know anything about u-boot-menu, but maybe josch can answer this | 22:47 |
| Boostisbetter | no worries, it's late, and not urgent. My reform is up and running and suspending. I will probably not fresh boot for quite some time. | 22:52 |
| josch | Boostisbetter: if you have u-boot-menu installed, it will regenerate /boot/extlinux/extlinux.conf next time you install or upgrade a kernel | 22:53 |
| josch | Boostisbetter: if you want that, make sure your /etc/default/u-boot is set up correctly | 22:53 |
| josch | and to make extra sure, inspect /boot/extlinux/extlinux.conf after it got generated | 22:53 |
| josch | if you do not want /boot/extlinux/extlinux.conf to get generated, remove u-boot-menu | 22:53 |
| abortretryfail | Boostisbetter: did you get bit by the u-boot-menu making your Reform unbootable after an update too? | 23:08 |
| abortretryfail | I might take another crack at that, since I personally prefer the extlinux.conf being a normal text file versus the boot.scr that's a binary thing you can't really edit if things need fixing. | 23:09 |
| josch | theoretically you could also not install u-boot-menu and hand-edit extlinux.conf yourself adding the entries you like | 23:10 |
| josch | i really like the convenience to select the kernel on boot which allows me to install even wonky kernels because i know i can easily recover | 23:11 |
| josch | it just frustrates me a lot how it ignores ${bootargs} that makes it a show-stopper on the reform | 23:11 |
| abortretryfail | Yeah. Tbh what I really want is the ability to edit the kernel cmdline on the fly from the bootloader the way I can do with syslinux and grub. | 23:11 |
| josch | unless of course we accept that each user crafts their own /etc/default/u-boot | 23:11 |
| josch | abortretryfail: if you do not use u-boot-menu but boot.scr you can do that by entering the u-boot console and setting bootargs | 23:12 |
| abortretryfail | Ignores ${bootargs}? | 23:12 |
| josch | yes, the kernel will be booted with exactly the cmdline that is written in the "append" line in extlinux.conf | 23:12 |
| josch | independent on what u-boot itself sets for ${bootargs} | 23:13 |
| abortretryfail | i don't even think it reads boot.scr at all if it finds extlinux.conf | 23:13 |
| abortretryfail | been meaning to test that too, but i have had a busy weekend and not much time for tinkering | 23:14 |
| josch | yes, it first tries extlinux.conf and if that does not exist, falls back to boot.scr | 23:14 |
| abortretryfail | what can or can't be done from the u-boot prompt is kind of unclear to the uninitiated tbh. | 23:15 |
| abortretryfail | the help scrolls off the screen and the docs are very much geared towards embedded use | 23:15 |
| abortretryfail | I'm not even totally new to this (used it a bunch with openwrt over the years) and I still dont have a clue how to boot a kernel unless it was preconfigured. | 23:16 |
| abortretryfail | Memory addresses to load the kernel and initrd to? Beats me. No clue. This thing's bricked. \o/ | 23:16 |
| abortretryfail | So yeah, big value in having a rescue memory option. :D | 23:17 |
| abortretryfail | s/memory/menu/ | 23:17 |
| josch | one sec... the internet says that extlinux.conf "append" line supports variables | 23:18 |
| josch | can we maybe make this work after all? | 23:18 |
| abortretryfail | ooh neat | 23:18 |
| josch | i'll try out tomorrow | 23:18 |
| josch | then successful booting "just" relies on u-boot being recent enough such that it exports the important ${bootargs} | 23:19 |
| josch | ah and of course the most recent reform-boundary-uboot doesn't do that | 23:22 |
| josch | have to create a new tag for that | 23:22 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
