| vagrantc | minute: whatever is in upstream u-boot (or the dtb from the kernel based on patches from reform-debian-packages with the 6.18.7 pull request merged) | 00:04 |
|---|---|---|
| minute | vagrantc: ah, no idea what is there | 00:04 |
| vagrantc | let's see when it was last updated, it is probably from whatever is in upstream linux | 00:05 |
| + mjv (~matt@user/reform11910) | 00:05 | |
| vagrantc | 2f3766949bbea7aa5a472157561d387fd94205d2 Subtree merge tag 'v6.15-dts' of dts repo [1] into dts/upstream ... so looks like whatever was in v6.15 ~may 2025 | 00:06 |
| vagrantc | minute: i'm not sure which adapter it is, whatever came with the rk3588 module upgrades before you started shipping the dsi adapter | 00:07 |
| vagrantc | says MNT RHDP on the adapter | 00:09 |
| vagrantc | MNTRHDPXXX10R01 | 00:10 |
| minute | there is no separate dsi adapter, because that one is integrated on the motherboard | 00:10 |
| minute | vagrantc: ok so you have the hdmi adapter | 00:10 |
| vagrantc | ok :) | 00:10 |
| vagrantc | ~before you started not shipping any adapter~ :) | 00:11 |
| minute | vagrantc: right :D | 00:11 |
| vagrantc | anyways, was pretty cool to see upstream u-boot boot at all :) | 00:11 |
| minute | yeah | 00:12 |
| minute | interesting, i had a new bug where usb2 would always stop after a minute or so (keyboard+trackball) on pocket with barebox boot... but it looks like it was just extra kernel options that i had for debugging: clk_ignore_unused pd_ignore_unused regulator_ignore_unused | 00:14 |
| josch | gordon1: i didn't hear back from the trackstick people at sprintek. I wrote them again a few days ago, still waiting for a reply... | 00:20 |
| josch | minute: are you using any bluetooth audio? What is your solution on how to get bluetooth with rk3588 classic reform? | 00:22 |
| josch | inb4: not a problem with reform next ;) | 00:23 |
| josch | (nor with pocket reform) | 00:24 |
| minute | josch: not using bluetooth on classic | 00:30 |
| minute | josch: the best solution is probably that mPCIe card by realleif | 00:30 |
| josch | yeah, i keep thinking of that one | 00:30 |
| minute | josch: maybe ping him how that's going! | 00:30 |
| josch | unfortunately they said that their design uses non-free components, so it cannot be reliesed under a foss license | 00:31 |
| minute | hmm weird | 00:31 |
| minute | i think that's not how OHL works | 00:31 |
| josch | no, not in hardware componets | 00:32 |
| minute | schematics are not copyrightable in germany anyway afaik | 00:32 |
| josch | the design is apparently derived from something non-free | 00:32 |
| josch | wait, i have leif's messages on fedi | 00:32 |
| minute | i doubt there's any "schaffenshöhe" in whatever was the original, but maybe i'm overlooking sth | 00:33 |
| josch | oh it was a private message | 00:33 |
| minute | maybe the drawing of the card or sth | 00:33 |
| josch | but leif hinted they will do a next version | 00:33 |
| minute | ok | 00:33 |
| josch | apparently they think the license of the footprints is the problem | 00:34 |
| minute | nice | 00:34 |
| minute | nice @ next version i mean | 00:34 |
| minute | ah, footprints schmootprints | 00:34 |
| minute | i'm not sure those are defensible for copyright either, because 1. there are not many ways of how to make a footprint, 2. they are trivial shapes and neither brands nor artwork (not recognizable ownership) 3. a technical necessity | 00:36 |
| minute | i wanna see the court case where they're like, you pirated our rectangle and some circles | 00:37 |
| minute | sorry, i meant schöpfungshöhe https://de.wikipedia.org/wiki/Sch%C3%B6pfungsh%C3%B6he | 00:37 |
| josch | yeah, we have the same discussion for software too -- sometimes i do a drive-by commit somewhere and then the org wants me to sign a CLA and i'm like "people, it's a 2 line change, i cannot possibly claim copyright for this..." | 00:38 |
| minute | hehe https://de.wikipedia.org/wiki/Kleine_M%C3%BCnze | 00:39 |
| minute | > So sind auch einfache Werke durch das Urheberrecht schützbar, die über ein schwaches Maß an individueller, schöpferischer und gestalterischer Ausdruckskraft verfügen. Maßgeblich ist, dass eine schöpferisch wertvolle und daher schutzwürdige Errungenschaft erschaffen wurde | 00:40 |
| josch | i know why law is not my hobby XD | 00:41 |
| josch | but TIL :) | 00:41 |
| minute | :D | 00:43 |
| josch | but yes, i think the leif solution is really nice: it would give me more usb while at the same time i can plug my intel ax210 into the classic | 00:46 |
| josch | and some free usb ports into which i can plug other stuff | 00:46 |
| minute | yeah | 00:46 |
| kop316 | josch: In regards to the request for the dmesg for crashing wifi, I put it here https://paste.debian.net/hidden/07d77b68 (sorry for the double notification, I wasn't sure if my first one went through) | 00:54 |
| josch | mmc1: Card stuck being busy! __mmc_poll_for_busy -- this is the slowest sd-card effer... at least at some point rk3588 has everything in ram... | 00:54 |
| josch | kop316: the dmesg is interesting for minute not for me -- i'm not a kernel person at all :) | 00:54 |
| kop316 | ohhh sorry! I was responding to you asking for it | 00:55 |
| josch | kop316: sorry for the confusion -- me asking was because i know that those who can help you might want to see it :) | 00:58 |
| kop316 | josch: fair enough, thank you! | 00:58 |
| + siviq (~siviq@user/siviq) | 01:05 | |
| elb | I had to boot my Pocket maybe 8 times to get it running to give lecture today | 01:11 |
| elb | I got a photo of the on-screen messages, but I _really_ need to get it on a serial adapter | 01:11 |
| minute | iwlwifi pretty borked here on 6.18.5 on rk3588, not sure if barebox related or new kernel thing? | 01:18 |
| minute | elb: sucks, sorry to hear! if you wanna share the photo i can take a look | 01:19 |
| minute | also, why does networkmanager.service have a 10 minute timeout... | 01:20 |
| elb | minute: I was talking to josch about this the other day, I think this is 6.17 -> 6.18 related | 01:20 |
| elb | it's a new symptom, where it finds the nvme, asks for the password, and then ... fails to find any data it wants to use after that | 01:21 |
| minute | ugh ok | 01:21 |
| minute | i wonder if rk3588 is also blessed with recent dw-pcie issues/changes, like a311d | 01:21 |
| josch | i took time tomorrow to investigate the a311d issue -- with rk3588 here it should be quick to do some kernel builds | 01:23 |
| minute | josch: cool, thank you! | 01:23 |
| elb | the messages it prints are a bunch of 'rockchip-pm-domain fd8d8000.power-management:power-controller: sync_state() pending due to fb000000.gpu' (and then due to .video-codec at various fbxxxx addresses, .rga, .i2s) | 01:24 |
| kop316 | minute before the latest update it was fine for me. not sure if thats useful to know? | 01:24 |
| minute | elb: ah, those messages are harmless | 01:25 |
| elb | and a couple of times, but not always, 'pcieport 0004:40:00.0: bridge configuration invalid ([bus 00-00]), reconfiguring' followed by 'pci_bus 0004:41: busn_res: [bus 41] end is updated to 41' | 01:25 |
| elb | and that address I'm 99.44% sure is the NVMe | 01:25 |
| minute | kop316: ah i have the same issue as you, now | 01:25 |
| minute | elb: that message is also "normal" | 01:25 |
| elb | minute: I thought they might be because I also see them at other times | 01:26 |
| elb | but in this case they're emitted _as soon_ as I hit enter on the cryptsetup password prompt | 01:26 |
| minute | not very helpfully worded messages, i think | 01:26 |
| minute | mhm | 01:26 |
| elb | if those are harmless, there's nothing else on-screen that's relevant | 01:26 |
| kop316 | minute understood. if there's anything i can do to help let me know! | 01:26 |
| elb | I'll need to get a debug dump in some fashion that has scrollback | 01:26 |
| minute | elb: "tio" is really useful for that over serial | 01:27 |
| elb | yeah I have to crack it open and put serial on it | 01:27 |
| minute | kop316: yeah, i'll need to investigate/fix this | 01:27 |
| elb | I couldn't do that prior to a few weeks ago because m yserial adapter couldn't sustain 1.5Mbs or whatever it is | 01:27 |
| elb | but I got a new one that allegedly can | 01:27 |
| minute | elb: my need for a nice not-needing-to-crack-it-open uart solution is constantly increasing | 01:27 |
| elb | I haven't even plugged it in yet 😭 | 01:28 |
| elb | yeah ... I woudl love to have some kind of tiny coaxial port someplace like GPSes and stuff used to have | 01:28 |
| elb | like a little 3/32" trs | 01:28 |
| minute | elb: yeah | 01:30 |
| elb | port space on the pocket makes me think a lot ;-) | 01:31 |
| minute | kop316: i'm now putting "blacklist iwlmvm" in /usr/lib/modprobe.d/reform.conf here so my system isn't very very stuck/slow after booting due to iwlwifi error spam | 01:31 |
| minute | elb: i was thinking (temporary bodge) to make a cable that would put rx/tx on the headset board -> headphone/mic pins | 01:32 |
| kop316 | minutegood to know, ill add that | 01:32 |
| minute | kop316: and then i'm doing update-initramfs -u | 01:32 |
| minute | i also put pcie_aspm=off in my kernel options now, maybe it'll help | 01:32 |
| minute | josch: btw regarding a311d, maybe that issue is fixed in 6.19... i remember you said it was in experimental? | 01:34 |
| elb | minute: I've thought about trying to put a floating panel mount jack above the board, but I briefly looked to see if I could find one with a small enough circular footprint and didn't see one | 01:34 |
| elb | you can get board mount jacks that are barely bigger than the plug itself, but I guess panel mount is passe | 01:35 |
| minute | kop316: hmm weird, that "blacklist" entry didn't work | 01:35 |
| minute | kop316: but on this boot, i'm not getting the errors | 01:35 |
| minute | i.e. wifi works now | 01:36 |
| minute | not sure if because of pcie_aspm=off | 01:36 |
| minute | elb: ah hm ;/ | 01:36 |
| minute | josch: funfact: barebox can edit files on fat partitions. but not on ext4. (thinking back to our efi partitioning chats) | 01:38 |
| minute | i.e. with the built in text editor i can edit blspec files. that alone makes me want to make an extra fat partition :D | 01:39 |
| josch | is there some sort of boot menu that lets you iterate through the available spec files? I didn't see one in the last build of barebox i had running here | 01:41 |
| josch | i think the fat vs. ext4 | 01:41 |
| josch | topic is different now because back then i thought u-boot would only read from the first partition | 01:41 |
| josch | now we are not talking about u-boot anymore | 01:41 |
| josch | and even u-boot looks at more than the first partition | 01:41 |
| minute | might be total placebo but now my GPLL is at 1.188GHz instead of 1.68GHz and the pocket doesn't get as hot (?) | 01:44 |
| minute | josch: yes, there's an interactive boot menu | 01:45 |
| josch | nice | 01:45 |
| minute | (called menutree) | 01:45 |
| minute | josch: it discovers all bspec files on all devices, and probably also other things like EFI (but haven't tested EFI support yet) | 01:45 |
| josch | maybe my situation was that there was only a single valid choice and thus it didn't give me an interactive selection | 01:47 |
| elb | yeah this is 8mm OD, that's the absolute smallest I'm turning up: https://www.digikey.com/en/products/detail/schurter-inc/4831-2300/1731047 | 01:47 |
| josch | or do i have tointerrupt boot and ask for it? | 01:47 |
| - xktr (QUIT: Ping timeout: 244 seconds) (~xktr@user/xktr) | 01:47 | |
| minute | josch: you press "m" to stop auto boot | 01:48 |
| elb | to be fair, I'm not sure they can really make it that much smaller, looking at it | 01:48 |
| minute | josch: it depends on how it is configured i guess | 01:48 |
| minute | josch: i made a little video to show you, uploading | 01:48 |
| minute | josch: https://mastodon.social/@mntmn/115986965428592424 | 01:51 |
| josch | minute: thank you! are your display fixes what you recently pushed into the ci-202601 branch? | 01:52 |
| minute | josch: yes, it should all work now. also consider the README-mnt, i had the wrong tf-a in there before (bl31) | 01:53 |
| minute | josch: don't rely on it though and have another microsd with uboot or another computer ready in case of issues :D also, probably only works with display v2 | 01:53 |
| minute | josch: not sure if you have that? | 01:54 |
| josch | on my todo list tomorrow after a311d :D | 01:54 |
| + xktr (~xktr@user/xktr) | 01:54 | |
| minute | ok, feierabend time | 01:54 |
| josch | _o/ | 01:55 |
| bremner | for some reason feirabend always reminds me of Feuerzangenbowle. Probably some English / German punning, among other problems. | 01:56 |
| - siviq (QUIT: Quit: Client closed) (~siviq@user/siviq) | 01:58 | |
| - chomwitt (QUIT: Ping timeout: 246 seconds) (~chomwitt@2a02:85f:9a3c:a200:1ac0:4dff:fedb:a3f1) | 02:21 | |
| kop316 | minuteif i may ask, is there a good README for.editting kernel config options? searching online is very unhelpful | 02:23 |
| - paperManu_ (QUIT: Ping timeout: 260 seconds) (~paperManu@146.71.9.156) | 02:26 | |
| - xktr (QUIT: Ping timeout: 245 seconds) (~xktr@user/xktr) | 02:31 | |
| + xktr (~xktr@user/xktr) | 02:45 | |
| - paperManu (QUIT: Ping timeout: 240 seconds) (~paperManu@146.71.9.156) | 02:45 | |
| minute | kop316: IIRC the cleanest place is a config file of the package u-boot-menu | 02:53 |
| minute | kop316: whose name i have to research (maybe josch knows by heart) | 02:53 |
| minute | kop316: another temporary option is to put it in your /boot/extlinux/extlinux.conf (if you use extlinx/u-boot-menu, which is default by now, but not sure about your setup) | 02:54 |
| kop316 | thanks! i it looks like the extlinux.conf works, but u-boot-updaye overrides it. that may be the best solution though, as hopefilly it won't be needed in the future | 02:55 |
| minute | yep! | 02:58 |
| minute | kop316: does that option fix iwlwifi for you? | 02:58 |
| - pomel0 (QUIT: Ping timeout: 252 seconds) (~pomel0@user/pomel0) | 03:04 | |
| kop316 | minute: its been working for 5 minutes, much longer thann before! | 03:04 |
| minute | kop316: :D | 03:20 |
| minute | "you're absolutely right!" https://github.com/armbian/build/pull/9285#issuecomment-3788489509 | 03:32 |
| ^alex | slop! slop! slop! | 03:42 |
| minute | yeah... | 04:07 |
| - xktr (QUIT: Ping timeout: 260 seconds) (~xktr@user/xktr) | 04:29 | |
| + xktr (~xktr@user/xktr) | 04:31 | |
| + pomel0 (~pomel0@user/pomel0) | 04:31 | |
| - pomel0 (QUIT: Ping timeout: 264 seconds) (~pomel0@user/pomel0) | 04:36 | |
| + pomel0 (~pomel0@user/pomel0) | 04:42 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 04:46 | |
| + pomel0 (~pomel0@user/pomel0) | 04:47 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 04:49 | |
| + pomel0 (~pomel0@user/pomel0) | 04:50 | |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 04:54 | |
| - xktr (QUIT: Ping timeout: 244 seconds) (~xktr@user/xktr) | 04:57 | |
| + pomel0 (~pomel0@user/pomel0) | 05:01 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 05:06 | |
| + pomel0 (~pomel0@user/pomel0) | 05:06 | |
| + xktr (~xktr@user/xktr) | 05:10 | |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 05:10 | |
| + pomel0 (~pomel0@user/pomel0) | 05:15 | |
| - xktr (QUIT: Ping timeout: 265 seconds) (~xktr@user/xktr) | 05:16 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 05:31 | |
| + pomel0 (~pomel0@user/pomel0) | 05:31 | |
| - pomel0 (QUIT: Ping timeout: 245 seconds) (~pomel0@user/pomel0) | 05:40 | |
| + pomel0 (~pomel0@user/pomel0) | 06:10 | |
| + xktr (~xktr@user/xktr) | 06:13 | |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 06:17 | |
| bremner | my ix-industrial adapter (to cat5/6) is so stiff I worry about damage to the connector. Does anyone have experience with adapters (rather than male-male cables) that work well? | 06:50 |
| bremner | I guess I could pick up another cable, it's just more weight for travel | 06:54 |
| josch | bremner: I have the same problem with hdmi cables | 07:15 |
| josch | minute, kop316: I don't know the location by heart. What I do is that because I know the file is shipped by reform-tools and because I know it's for u-boot-menu I run: dpkg -L reform-tools | grep menu | 07:16 |
| josch | and then I get: /usr/share/u-boot-menu/conf.d/reform.conf | 07:16 |
| josch | And when I read that file there is a comment on top which t ells me what to do: edit /etc/default/u-boot | 07:17 |
| josch | or /etc/u-boot-menu/conf.d/*.conf | 07:17 |
| bremner | is there a way to disable nvme completely by e.g. blacklisting a module? I guess I could physically pull the thing, but I'm more of a software person | 07:26 |
| bremner | I know (softof) about the about the sysfs files one can hit | 07:26 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 07:48 | |
| + arjan (~arjan@user/arjan) | 07:57 | |
| arjan | anyone at fosdem who feels like helping me troubleshoot the display v2 of my pocket not not working? | 07:59 |
| arjan | ops double negation was unintended | 08:00 |
| bremner | followup question about removing the SSD: do I really need to disconnect the batteries first? that sounds more hazardous than removing the SSD | 08:01 |
| josch | bremner: I insert and remove SSD without disconnecting the batteries first all the time. On the other hand, I also recently fried some components on my motherboard 3.0 because of an unintentional short... | 09:14 |
| bremner | kindof mixed message ;) | 09:35 |
| josch | intentional :) | 09:35 |
| rwa_ | oh, i broke one of the usb-a ports on my classic :( | 09:46 |
| josch | rwa_: ouch! how did you manage that? | 09:48 |
| rwa_ | clumsy handling while a stick was attached i guess | 09:54 |
| rwa_ | https://rw-net.de/reform_usb_port.jpg | 09:57 |
| gordon1 | been there, done that (not with reform tho) | 10:24 |
| gordon1 | josch: so trackpoint people don't want your money? are there successful stories of people ordering from them on the internet& | 10:25 |
| gordon1 | ? | 10:25 |
| josch | gordon1: maybe they lost track of my mail. Yes, custom keyboard people seem to buy their tracksticks there. | 10:26 |
| josch | I'll ping them again if needed. | 10:27 |
| josch | just wanted to give you an update because it has been a few days :) | 10:27 |
| gordon1 | no worries | 10:31 |
| gordon1 | josch: i'm slowly getting to the finish line with ordering usb hub, unfortunately there is this pesky reform motherboard in the way, so i guess JST usb connectors will be on the underside of the board https://tinystash.undef.im/il/5Abi326h1q4hnW8BUhkpdbi3zcxL8onHH1bL4SEuDhw2jTmPwyFCRfd333ZUwwj6FzRWrSrd7n8uq3BTEqDxqhQ8.png | 10:34 |
| gordon1 | j5 and j6 will plug into the jst usb near the usb ports | 10:35 |
| gordon1 | directly | 10:35 |
| gordon1 | j8 for (optional) ws-lora-rp2040 https://tinystash.undef.im/il/3nr5mNN1jBbbxLaRfJsHqpvpurW95y9TvVenAYbCp4hvXJWm85rrybNT8954SkoECXZbRmsagMQT4iMzPrQdVSbE.png | 10:36 |
| gordon1 | and probably some sort of a 3d-printed spacer on the other side so it wont stress those connectors | 10:37 |
| - cli (QUIT: Ping timeout: 265 seconds) (~m-vsauiy@user/cli) | 10:38 | |
| + cli (~m-vsauiy@user/cli) | 10:40 | |
| - cli (QUIT: Ping timeout: 264 seconds) (~m-vsauiy@user/cli) | 11:14 | |
| + cli (~m-vsauiy@user/cli) | 11:18 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 12:02 | |
| - xktr (QUIT: Ping timeout: 256 seconds) (~xktr@user/xktr) | 12:04 | |
| + xktr (~xktr@user/xktr) | 12:05 | |
| - murphnj (QUIT: Ping timeout: 260 seconds) (~murph@user/murphnj) | 12:07 | |
| - voltaire28_ (QUIT: Ping timeout: 240 seconds) (~jlafon@28.162.2.93.rev.sfr.net) | 12:33 | |
| gordon1 | so i pulled cells out of my reform classic, charged them precisely with external lifepo4 charger, put them back, then, with reform ON i measured them in-situ and the voltage on the cells agree within 0.02V, however if i read cell voltages from LPC - last cell is really low | 12:38 |
| gordon1 | $ cat /sys/bus/spi/drivers/reform2_lpc/spi1.0/cells | 12:38 |
| gordon1 | 3.262 3.282 3.282 3.273 3.266 3.279 3.280 3.210 | 12:38 |
| gordon1 | ideas? | 12:38 |
| gordon1 | is it just an artifact that last cell is being on the ground return path or something? | 12:39 |
| gordon1 | also LPC voltages seem to be ~0.03V lower than real cell voltages | 12:41 |
| gordon1 | no, actually 0.05 | 12:41 |
| gordon1 | i measured on average 3.32 | 12:41 |
| gordon1 | 3.31 maybe | 12:41 |
| gordon1 | can't do math in my head | 12:41 |
| gordon1 | will this result in overcharging the last cell? | 13:07 |
| minute | gordon1: multiple factors involved, like the cell protectors, cables etc | 13:07 |
| minute | gordon1: this little difference isn't a problem i would say, and lifepo4 cells are quite tolerant also | 13:07 |
| gordon1 | ok, i'll check in few weeks if that has any consequences, 0.07V seems small | 13:09 |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 13:30 | |
| - buckket (QUIT: Quit: buckket) (~buckket@vps.buckket.org) | 13:52 | |
| + buckket (~buckket@vps.buckket.org) | 13:53 | |
| josch | gordon1: maybe the unpopulated underside of your usb hub is a good flat area to stick one or two antennas onto? | 14:12 |
| gordon1 | josch: i need to make sure to skip ground pour then | 14:25 |
| + murphnj (~murph@user/murphnj) | 14:29 | |
| + paperManu (~paperManu@146.71.9.156) | 14:39 | |
| - voltaire28_ (QUIT: Ping timeout: 244 seconds) (~jlafon@28.162.2.93.rev.sfr.net) | 14:46 | |
| - xktr (QUIT: Ping timeout: 256 seconds) (~xktr@user/xktr) | 14:48 | |
| + pomel0 (~pomel0@user/pomel0) | 14:53 | |
| + xktr (~xktr@user/xktr) | 14:55 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 14:57 | |
| + pomel0 (~pomel0@user/pomel0) | 14:57 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 14:59 | |
| + pomel0 (~pomel0@user/pomel0) | 14:59 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 15:01 | |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 15:08 | |
| + pomel0 (~pomel0@user/pomel0) | 15:08 | |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 15:19 | |
| + pomel0 (~pomel0@user/pomel0) | 15:41 | |
| josch | minute: barebox rocks -- I just installed kernel 6.19 and it's the first time booting it so maybe it's borked but you know what? With barebox I can now select to boot *either* 6.18 or 6.19 \o/ | 16:05 |
| josch | (also turns out i have the right display) | 16:05 |
| josch | Linux m 6.19-rc6-mnt-reform-arm64 #1 SMP PREEMPT Debian 6.19~rc6-1~exp1+reform20260131T122352Z (20 aarch64 GNU/Linux | 16:07 |
| josch | somebody has to do the HDMI testing -- i don't have any hardware for that | 16:07 |
| - mjv (QUIT: Remote host closed the connection) (~matt@user/reform11910) | 16:19 | |
| josch | okay, so according to the boot menu, barebox sees both the bootspec entry for 6.18 and 6.19 but when i do not use the boot menu, then it automatically boots 6.18 an not 6.19 -- i wonder why it prefers the name which orders lower | 16:19 |
| elb | is it possible to use an encrypted reform system image on micro SD? | 16:47 |
| elb | let me rephrase that; I'm absolutely sure it is | 16:48 |
| elb | is it achievable with the provided scripts :-) | 16:48 |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@134.22.115.162) | 16:49 | |
| elb | it looks to me like the answer is "not quite?" | 16:57 |
| elb | specifically I would have to configure eMMC to use a /boot off eMMC and then load the entire SD card as an encrypted volume; what I actually want is for it to behave "as usual" where sticking th SD card in causes it to load /boot off the SD and then mount _another partition_ on the SD as the encrypted root | 16:58 |
| - voltaire28_ (QUIT: Ping timeout: 264 seconds) (~jlafon@28.162.2.93.rev.sfr.net) | 17:04 | |
| + spew (~spew@user/spew) | 17:05 | |
| + siviq (~siviq@user/siviq) | 17:08 | |
| elb | (just FYI the current PR rk3588 system image is not a multiple of 8 MiB uncompressed, so the operator handbook dd command hangs at the end) | 17:10 |
| + paperManu_ (~paperManu@146.71.9.156) | 17:10 | |
| + spew_ (~spew@user/spew) | 17:11 | |
| - spew (QUIT: Ping timeout: 260 seconds) (~spew@user/spew) | 17:11 | |
| gordon1 | josch: so it the barebox ready for classic or just for pocket? | 17:12 |
| elb | it actually doesn't seem to be a convenient multiple of any given size | 17:12 |
| - spew_ (QUIT: Client Quit) (~spew@user/spew) | 17:13 | |
| josch | gordon1: just for pocket for now | 17:31 |
| gordon1 | aww, oh well | 17:31 |
| josch | gordon1: but don't ask me i know it from m.inute and i have been known to misremember :) | 17:31 |
| gordon1 | i'll wait till it is tester or i find a way to do maskrom on rk3588 | 17:31 |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 17:39 | |
| josch | elb: yes, it's possible but maybe a bit hacky depending what you want and i'm not yet sure I quite understand. Do you want to migrate an existing system? Do you want to do a new installation? If you open a forum thread with all the details, then i can reply once i have time again. Family now. | 17:42 |
| gordon1 | *tested | 17:45 |
| josch | gordon1: why maskrom? | 17:45 |
| gordon1 | josch: if i break the bootloader on the emmc this is the only way to recover it, isn't it? | 17:48 |
| josch | gordon1: yes, so don't use emmc | 17:49 |
| gordon1 | i mean not completely break, but if it sorta works but does not boot anything | 17:49 |
| gordon1 | but i want to have it in my emmc | 17:49 |
| josch | when i experiment with u-boot i have emmc zeroed and have u-boot on sd-card exclusively | 17:49 |
| minute | elb: are you sure it "hangs" or it just needs a long time in the end committing everything that it has only committed to cache before? i don't think that there is a block size multiple limitation in dd | 17:49 |
| josch | then first test in on sd-card and only once you verified that it works, then flash it to emmc | 17:49 |
| minute | elb: there's an option for dd to write directly without using a write cach | 17:49 |
| minute | cache | 17:49 |
| gordon1 | it's too risky for the only workhorse device i have | 17:49 |
| + mark_ (mjw@2001:67c:1810:f055:ae2d:9c17:a900:20c7) | 17:49 | |
| minute | elb: but i would use bmaptool in any case because it's faster | 17:49 |
| gordon1 | unless i have backup plan for maskrom | 17:50 |
| minute | gordon1: the way i develop is to zero the bootloader on emmc and have bootloader on microsd, or have 1 uboot microsd and 1 barebox microsd. but alas, there's not internal display support yet for barebox and i wouldn't declare it ready for testing anyway. first i need to create separate dtbs for the devices etc | 17:51 |
| gordon1 | right, i see | 17:51 |
| elb | josch: it's not urgent, my general goal is to be able to smack an as-fully-as-possible encrypted SD card into the sd card slot and run an entirely different image than what might be on emmc or nvme | 17:51 |
| + wielaard (mjw@151.216.149.88) | 17:52 | |
| elb | minute: good question | 17:52 |
| elb | re: caching | 17:52 |
| minute | elb: in summary, the "hang" is normal behavior and you just need to wait several minutes or so | 17:52 |
| - mark_ (QUIT: Ping timeout: 265 seconds) (mjw@2001:67c:1810:f055:ae2d:9c17:a900:20c7) | 17:56 | |
| elb | understood | 18:00 |
| elb | I have reasons for all of these questions, but irc is not the place for them, unfortunately | 18:00 |
| - f_ (QUIT: Ping timeout: 240 seconds) (16abab341f@postmarketOS/funderscore) | 18:04 | |
| - voltaire28_ (QUIT: Quit: WeeChat 4.7.2) (~jlafon@28.162.2.93.rev.sfr.net) | 18:06 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 18:06 | |
| - voltaire28_ (QUIT: Client Quit) (~jlafon@28.162.2.93.rev.sfr.net) | 18:07 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 18:08 | |
| + mjv (~matt@user/reform11910) | 18:16 | |
| - siviq (QUIT: Quit: Client closed) (~siviq@user/siviq) | 18:18 | |
| + siviq (~siviq@user/siviq) | 18:23 | |
| mjv | hi, i have an rk 3588 pocket reform with a crucial 1tb nvme. i've updated to 2026-01-28 uboot, system controller 20251030. i am having to try restarting 5-10 times to get to the login prompt. i had read something about nvme not starting soon enough, and some recent updates to fix some uboot issues. i do see some errors for power manger power controller sync state pending. is it a bad nvme? taking to long to wake? what can i do to help diagnose and fix? (i'm | 18:29 |
| - paperManu_ (QUIT: Ping timeout: 260 seconds) (~paperManu@146.71.9.156) | 18:33 | |
| siviq | Hi, can you take a look at MNT Community, there has been reports for similar issues. https://community.mnt.re/t/pocket-reform-wont-boot-after-apt-upgrade/4148/27 | 18:50 |
| - pomel0 (QUIT: Ping timeout: 244 seconds) (~pomel0@user/pomel0) | 18:59 | |
| + pomel0 (~pomel0@user/pomel0) | 19:00 | |
| + habasud (~Srain@user/habasud) | 19:01 | |
| - pomel0 (QUIT: Ping timeout: 245 seconds) (~pomel0@user/pomel0) | 19:06 | |
| + paperManu_ (~paperManu@142.169.16.126) | 19:07 | |
| minute | mjv: i think the problem is mostly with uboot, i would say downgrade it | 19:08 |
| minute | but hmm 01-28 sounds recent | 19:09 |
| minute | like, 3 days ago :D so i would suspect that josch fixed some stuff there | 19:09 |
| minute | mjv: try adding pcie_aspm=off to your kernel options (for example in /boot/extlinux/extlinux.conf) | 19:09 |
| mjv | minute: i was following along with some recent stuff josch did in the thread siviq mentioned to fix that boot issue so recently upgraded to see if that does the job. thanks much i'll try adding that to kernel options | 19:11 |
| - wielaard (QUIT: Ping timeout: 265 seconds) (mjw@151.216.149.88) | 19:18 | |
| mjv | minute: getting oriented, not familiar with updating kernel options. file says don't update directly. safely ignore this or is there another place to add that option? | 19:31 |
| josch | mjv: which file? | 19:32 |
| + f_ (16abab341f@postmarketOS/funderscore) | 19:35 | |
| - voltaire28_ (QUIT: Quit: WeeChat 4.7.2) (~jlafon@28.162.2.93.rev.sfr.net) | 19:35 | |
| mjv | josch: minute pointed me to /boot/extlinux/extlinux.conf | 19:36 |
| josch | mjv: you are allowed to edit that file by hand. But it will be overwritten the next time u-boot-menu is executed, so your changes will not be permanent. | 19:37 |
| mjv | josch: assume i'd add the param to the append line(s)? assuming that fixes it will need to understand how to persist changes but that can come later | 19:42 |
| amk | when barebox is ready will it just be a matter of apt upgrade on the pocket? excited to try it out :) | 19:51 |
| - habasud (QUIT: Remote host closed the connection) (~Srain@user/habasud) | 19:52 | |
| mjv | in any case, i need to drop, will get up to speed and see if that does the trick. thanks again for all the help | 19:53 |
| - paperManu_ (QUIT: Ping timeout: 264 seconds) (~paperManu@142.169.16.126) | 19:54 | |
| josch | mjv: yes, you add it to the append line of the default kernel | 19:55 |
| josch | amk: No. The bootloader is never updated using apt. The bootloader is not a "file" and can thus not be managed by the package manager. | 19:55 |
| josch | upgrading your bootloader should be a concious decision | 19:55 |
| mjv | josch: thanks! | 19:55 |
| josch | and this is also why when you currently try to flash a new bootloader on rk3588, there is a big scary warning | 19:56 |
| - mjv (QUIT: Remote host closed the connection) (~matt@user/reform11910) | 19:56 | |
| josch | if you don't need to, there should not be a need to upgrade it | 19:56 |
| amk | ah cool, makes sense | 19:56 |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 20:00 | |
| - voltaire28_ (QUIT: Ping timeout: 265 seconds) (~jlafon@28.162.2.93.rev.sfr.net) | 20:06 | |
| + pomel0 (~pomel0@user/pomel0) | 20:42 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 20:59 | |
| - siviq (QUIT: Ping timeout: 272 seconds) (~siviq@user/siviq) | 21:06 | |
| + paperManu_ (~paperManu@146.71.9.156) | 21:09 | |
| josch | minute: good news: taifast ssd shows up again with kernel 6.19 on a311d | 21:19 |
| + siviq (~siviq@user/siviq) | 21:24 | |
| josch | rwa_, kop316: if you like you could also experiment and try out if 6.19~rc6 magically fixes your pci problems on rk3588 by flashing this system image onto an sd-card and booting from it: https://source.mnt.re/josch/reform-system-image/-/jobs/18098 | 21:34 |
| kop316 | josch: sure, I can try later today! | 21:59 |
| - pomel0 (QUIT: Ping timeout: 265 seconds) (~pomel0@user/pomel0) | 22:00 | |
| rwa_ | should be u | 22:04 |
| rwa_ | should be tested with updated u-boot? | 22:05 |
| josch | rwa_: don't touch u-boot yet and see what you get. You can get your u-boot version in the running system by looking at /proc/device-tree/chosen/u-boot,version or by running reform-check | 22:10 |
| + pomel0 (~pomel0@user/pomel0) | 22:11 | |
| - pomel0 (QUIT: Ping timeout: 240 seconds) (~pomel0@user/pomel0) | 22:15 | |
| + pomel0 (~pomel0@user/pomel0) | 22:17 | |
| - pomel0 (QUIT: Ping timeout: 260 seconds) (~pomel0@user/pomel0) | 22:40 | |
| + pomel0 (~pomel0@user/pomel0) | 22:46 | |
| rwa_ | seems to work fine, nvme is detected | 23:05 |
| kop316 | josch: just tested, the wifi hasn't crashed after 4 minutes, so I htink it is good | 23:06 |
| rwa_ | but wifi is missing for whatever reason | 23:06 |
| josch | okay, so 6.19 does not solve rk3588 wifi reliably | 23:07 |
| kop316 | josch: you mean for lwiwifi or the other one? | 23:10 |
| josch | we have the iwlwifi crash (and apparently it hasn't crasched for you -- yet) and we have the pci problems where linux doesn't even see the card | 23:11 |
| josch | those are on rk3588. And then we have the a311d issue with nvme | 23:11 |
| josch | but so far only with the taifast ssd | 23:11 |
| kop316 | ahh gotcha. im leaving it running, but it hasnt crashed still. I think if it hasnt crashed yet it will not crash on iwlwifi | 23:13 |
| + wielaard (~mjw@213.246.238.181) | 23:49 | |
| - voltaire28_ (QUIT: Quit: WeeChat 4.7.2) (~jlafon@28.162.2.93.rev.sfr.net) | 23:56 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
