| josch | reform28775: just make sure you are on the sysimage-v3 branch -- if it's compiling stuff then you are on the wrong branch | 00:00 |
|---|---|---|
| reform28775 | Oh then I'm definitely on the wrong one | 00:00 |
| reform28775 | This does clones and builds a kernel, then uboot, and then a bunch of chrooted apt stuff | 00:01 |
| josch | reform28775: yeah, that's wrong. kernel, u-boot and some more packages are now built in other repos on source.mnt.re and the result is published on https://mntre.com/reform-debian-repo which in turn is used by the sysimage-v3 mkimage script so that this script only has to install stuff but nothing is compiled by it. | 00:14 |
| reform28775 | Okay I will continue to bumble around in Git, thanks for you help | 00:21 |
| reform28775 | Ah finally, I see how it is different! I'll try it on my build box | 00:25 |
| - reform28775 (QUIT: Quit: reform28775) (~trimbeej@2601:545:8202:19d0::b4dd) | 00:26 | |
| - rnjqvy5[m] (PART: !!unknown attribute: msg!!) (~rnjqvy5mi@2001:470:69fc:105::2:526) | 00:27 | |
| vagrantc | josch: ah, hdmi output is working in u-boot ? | 01:11 |
| + reform23110 (~trimbeej@2601:545:8202:19d0::b4dd) | 02:24 | |
| - reform23110 (QUIT: Client Quit) (~trimbeej@2601:545:8202:19d0::b4dd) | 02:25 | |
| - Christoph_ (QUIT: Remote host closed the connection) (~Christoph@p4fe73e0d.dip0.t-ipconnect.de) | 03:49 | |
| - Guest5330 (QUIT: Ping timeout: 256 seconds) (~nicolas@150-48-142-46.pool.kielnet.net) | 03:59 | |
| + nsc (~nicolas@218-49-142-46.pool.kielnet.net) | 04:00 | |
| * nsc -> Guest974 | 04:01 | |
| - sts-q (QUIT: Ping timeout: 250 seconds) (~sts-q@2a01:41e1:2e90:1b00:fc51:e3cc:8102:f600) | 04:34 | |
| + sts-q (~sts-q@2a01:41e1:2ec0:8800:96d:481f:3874:2dda) | 04:37 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 05:33 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149-210-16-210.mobile.nym.cosmote.net) | 06:01 | |
| + Ar|stote|is (~linx@149-210-16-210.mobile.nym.cosmote.net) | 06:01 | |
| - sbp (QUIT: *.net *.split) (~sbp@apache/doge/sbp) | 06:25 | |
| - leonardo (QUIT: *.net *.split) (~leonardo@user/leonardo) | 06:25 | |
| - bleb (QUIT: *.net *.split) (~cm@user/bleb) | 06:25 | |
| - vkoskiv_ (QUIT: *.net *.split) (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 06:25 | |
| + sbp (~sbp@2001:19f0:6800:1102:5400:ff:fe11:39a1) | 06:25 | |
| + vkoskiv (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 06:25 | |
| + bleb (~cm@user/bleb) | 06:25 | |
| + leonardo (~leonardo@user/leonardo) | 06:25 | |
| - lastebil (QUIT: *.net *.split) (truck@shell.suomiscene.fi) | 06:28 | |
| + lastebil (truck@shell.suomiscene.fi) | 06:28 | |
| - hl (QUIT: Remote host closed the connection) (~hl@user/hl) | 08:57 | |
| + hl (~hl@user/hl) | 08:57 | |
| - joeyh (QUIT: Ping timeout: 260 seconds) (joeyh@kitenet.net) | 09:30 | |
| - sbp (QUIT: Changing host) (~sbp@2001:19f0:6800:1102:5400:ff:fe11:39a1) | 09:47 | |
| + sbp (~sbp@apache/doge/sbp) | 09:47 | |
| - ex-parrot (QUIT: Ping timeout: 248 seconds) (~fincham@user/ex-parrot) | 10:29 | |
| + ex-parrot (~fincham@user/ex-parrot) | 10:30 | |
| + Christoph_ (~Christoph@p54bf648a.dip0.t-ipconnect.de) | 11:28 | |
| + MajorBiscuit (~MajorBisc@145.94.235.5) | 11:43 | |
| + reform8044 (~cliffya@180-150-53-95.b49635.mel.static.aussiebb.net) | 13:15 | |
| - reform8044 (QUIT: Client Quit) (~cliffya@180-150-53-95.b49635.mel.static.aussiebb.net) | 13:19 | |
| dustfing` | My mnt-reform arrived :-) I assembled it yesterday and I love it. The only surprise was that I did not have a track ball. I was certain that I ordered a track ball, but when I reviewed my order history, there was no trackball on order. I ordered my laptop last August and it arrived less than a week ago. If I were to order a track ball today, now long would it take to arrive? | 13:29 |
| vkoskiv | dustfing`: Curious, when did you order? From CS? | 13:29 |
| vkoskiv | Mine was estimated on apr. 28th for many months but I haven't gotten a shipping notification yet | 13:29 |
| vkoskiv | Hah, reading comprehension. You ordered in august :D | 13:30 |
| minute | dustfing`: huh. can you forward me a screenshot of your order/history to lukas@mntre.com? i would like to look into that | 13:34 |
| dustfing` | minute: Sure. Thank you! | 13:38 |
| dustfing` | vkoskiv: I ordered around the end of Aug 2021. | 13:38 |
| * dustfing` -> dustfinger | 13:51 | |
| dustfinger | minute: I sent the email with my order summary attached. | 13:52 |
| dustfinger | I am considering installing the guix distro on my reform. I am normally a gentoo user, but guix has caught my attention. Has anyone tried that distro? | 13:53 |
| dustfinger | a.f.k for a bit. going to make coffee and breakfast. | 13:55 |
| - sundog (QUIT: Quit: The Lounge - https://thelounge.chat) (~sundog@reclaim.technology) | 14:05 | |
| + sundog2 (~sundog@reclaim.technology) | 14:05 | |
| + joeyh (joeyh@kitenet.net) | 14:17 | |
| - Kooda (QUIT: Quit: WeeChat 3.3) (~kooda@natsu.upyum.com) | 14:24 | |
| + Kooda (~kooda@natsu.upyum.com) | 14:25 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@66.115.189.236) | 15:12 | |
| + bkeys (~Thunderbi@66.115.189.236) | 15:21 | |
| - bkeys (QUIT: Quit: With every step we take, danger will follow closely) (~Thunderbi@66.115.189.236) | 15:34 | |
| + bkeys (~Thunderbi@66.115.189.236) | 15:38 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@66.115.189.236) | 15:40 | |
| + bkeys (~Thunderbi@66.115.189.236) | 15:40 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 16:18 | |
| jfred | dustfinger: I'm also interested in running Guix System on my Reform, but I don't think anyone has done so yet successfully. There was a thread on help-guix on it last year: https://mail.gnu.org/archive/html/help-guix/2021-09/msg00027.html | 17:04 |
| jfred | I've been planning to install NixOS on mine in the interim, though it appears I need a bigger SD card because I ran out of space trying to build the NixOS image haha | 17:06 |
| josch | Getting any userspace GNU OS to boot should not be too hard. The tricky part (in my case Debian) was to properly set it up so that upgrades of the kernel work and don't make the system unbootable. To get it fully working a few tweaks are needed but most of them can be found in the reform-tools package: https://source.mnt.re/reform/reform-tools | 17:13 |
| jfred | Guix System is like NixOS in that the bootloader has the previous system configs selectable, so even if a kernel upgrade made the system unbootable you can always pick the last one to boot from. It seems like the bootloader can’t/doesn’t output to the display on the Reform though, I was only able to interact with it over the serial console | 17:25 |
| jfred | I’m still fairly new to ARM Linux machines so I’m not sure if display output/KB input for the bootloader (u-boot?) is possible | 17:26 |
| josch | jfred: possible yes, but nobody did the work yet. While the internal display doesn't work, hdmi works though. | 17:27 |
| jfred | Ahh, interesting | 17:27 |
| josch | So you are back at: a wrong kernel update makes the system unbootable. :) | 17:28 |
| sknebel | or you need to type blind :D | 17:28 |
| josch | yes | 17:28 |
| josch | (assuming you know what you should see) | 17:28 |
| josch | Next question is, if the keyboard works in u-boot. | 17:29 |
| sknebel | (hm, would it be easier to make the bootloader talk to the keyboard display?) | 17:29 |
| jfred | What work needs to be done for that to work? Drivers for the display in u-boot? | 17:30 |
| jfred | Also the keyboard display is *very* small for interacting with the bootloader haha | 17:48 |
| sknebel | would be more a "press X now to boot old kernel" symbol, yes :D | 17:50 |
| vagrantc | wow... that would be interesting :) | 18:04 |
| jfred | Huh yeah that would be nice, a simplified “oh no I broke it” button. Wonder if you could also get u-boot on the OLED display to have a scrolling menu of boot menu items | 18:10 |
| jfred | If you really limited the line length… you can definitely fit a datestamp on each line at least, and for Guix maybe the (possibly truncated) generation number | 18:12 |
| vagrantc | been toying with the idea of testing out guix on the mnt/reform ... had it running on top of an earlier debian install on the mnt/reform | 18:20 |
| vagrantc | it's a little underpowered, although i was running off of microsd ... might perform better on NVMe | 18:20 |
| jfred | I do know from previous experience running Guix on my old ThinkPad that it can be fairly slow on slower hardware… Nix seems to be snappier, I wonder why | 18:25 |
| - MajorBiscuit (QUIT: Quit: WeeChat 3.4) (~MajorBisc@145.94.235.5) | 18:31 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 18:41 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149-210-16-210.mobile.nym.cosmote.net) | 20:38 | |
| + Ar|stote|is (~linx@149-210-16-210.mobile.nym.cosmote.net) | 20:38 | |
| dustfinger | jfred: I will let you know if I get it working. | 20:43 |
| jfred | :) | 20:43 |
| dustfinger | jfred: Thanks for the tip about using the serial console to interact with the boot loader. That is good to know. | 20:45 |
| jfred | Yeah that and a lot of other useful info is in the Reform handbook https://mntre.com/reform2/handbook/advanced.html#serial-console | 20:46 |
| jfred | Definitely very helpful for troubleshooting boot issues | 20:46 |
| dustfinger | cool, thanks! | 20:47 |
| josch | vagrantc: I enabled distro_bootcmd but am now getting "ERROR : memory not allocated" which seems to be some failed malloc -- how would I go about finding out when and how that happens? | 20:56 |
| - piroko (QUIT: Ping timeout: 240 seconds) (~piroko@104.225.216.16) | 21:04 | |
| - tarxvf (QUIT: Ping timeout: 246 seconds) (~tarxvf@mail.tarxvf.tech) | 21:04 | |
| - qbit (QUIT: Ping timeout: 276 seconds) (~qbit@h.suah.dev) | 21:06 | |
| + tarxvf (~tarxvf@mail.tarxvf.tech) | 21:11 | |
| + qbit (~qbit@h.suah.dev) | 21:12 | |
| - qbit (QUIT: Client Quit) (~qbit@h.suah.dev) | 21:14 | |
| - tarxvf (QUIT: Ping timeout: 276 seconds) (~tarxvf@mail.tarxvf.tech) | 21:16 | |
| + tarxvf (~tarxvf@mail.tarxvf.tech) | 21:22 | |
| vagrantc | josch: whe booting or building? ... do you have the kernel_addr_r, ramdisk_addr_r and fdt_addr_r environment variables? | 21:35 |
| josch | vagrantc: booting and yes: https://paste.debian.net/1239727/ | 21:37 |
| vagrantc | josch: what does printenv bootcmd show you? | 21:39 |
| josch | vagrantc: there is no way for me to get into a u-boot interactive session because the S1 on my reform is broken -- I'm setting "bootcmd" to try out new stuff, flash to the sd-card and then carry the reform over to where I can plug it into hdmi to see something... | 21:41 |
| vagrantc | josch: i'd put the fdt_addr_r lower than the kernel_addr_r just in case it grows | 21:41 |
| vagrantc | it'll happily overwrite previously loaded bits | 21:42 |
| josch | is there a limit to the amount of memory u-boot has access to? | 21:42 |
| vagrantc | sometimes ... but ... eesh | 21:43 |
| josch | is there something i can set to get the equivalent of 'set -x'? | 21:44 |
| vagrantc | don't see anything obviously wrong in your patch otherwise | 21:44 |
| josch | that's already good to know :) | 21:45 |
| vagrantc | you can echo ... but if you don't have access to it, where will you see it? | 21:45 |
| vagrantc | josch: you could also change which serial port u-boot uses ... you had access to another if i recall? | 21:45 |
| josch | right now i can only see the output of u-boot commands on hdmi but I cannot input anything -- the usb keyboard for example doesn't work even after setting CONFIG_USB_KEYBOARD, CONFIG_SYS_USB_EVENT_POLL_VIA_CONTROL_EP and running "usb start" | 21:46 |
| vagrantc | ah | 21:46 |
| vagrantc | guess the alternate serial port still applies, then ? :) | 21:46 |
| + qbit (~qbit@h.suah.dev) | 21:47 | |
| josch | vagrantc: yes, we tried to use my S2 (which is working) with patches suggested by minute but it didn't work in the end so I gave up that route | 21:47 |
| vagrantc | ah, you could do something like ... | 21:47 |
| josch | so, hardware wise, my S2 is working but while linux can use it, I didn't get u-boot to use it | 21:47 |
| vagrantc | CONFIG_BOOTCMD="printenv ; run bootcmd" (or distro_bootcmd ...) | 21:48 |
| josch | I had replaced "run distro_bootcmd" by "run bootcmd" because it said "command not found" for the former | 21:49 |
| josch | one sec... | 21:49 |
| josch | (from doc/README.distro I understood that it sets $bootcmd and not $distro_bootcmd) | 21:50 |
| vagrantc | fair | 21:50 |
| vagrantc | more important, was the printenv ... to get some output on the screen | 21:50 |
| josch | yes, i'm on it :) | 21:50 |
| vagrantc | and then you can follow the commands ... | 21:51 |
| vagrantc | although, it will probably scroll off the screen | 21:51 |
| josch | because $bootcmd will become super long, right? | 21:51 |
| josch | i just recorded a movie with my camera when this happened in the past and then went frame-by-frame to watch it in slow motion... | 21:52 |
| josch | (suck not to have serial) | 21:52 |
| vagrantc | also, try disabling some of the boot devices and only use the ones you're using for now | 21:52 |
| vagrantc | it might initialize some buggy bus or something that's triggering the issue | 21:52 |
| vagrantc | josch: you're courageous and/or foolish to try without serial, but thanks! :) | 21:53 |
| josch | vagrantc: nobody else with a working serial is doing it, so... ;) | 21:54 |
| vagrantc | i should fire mine up while the sunshine is kicking and see if i can't build my own working u-boot | 21:54 |
| josch | okay, so the printenv might've revealed a problem -- it just doesn't stop printing (maybe an infinite loop?) I'll try to reduce the number of boot devices and try again | 21:54 |
| vagrantc | ACTION is in the midst of a barrage of somewhat tedious debian QA/reproducible builds uploads | 21:55 |
| vagrantc | gotta make the graphs more green! | 21:55 |
| josch | nice! :) | 21:56 |
| josch | vagrantc: okay, so the infinite loop problem is, that CONFIG_BOOTCOMMAND sets bootcmd and setting that to "printenv ; run bootcmd" obviously leads to infinite recursion. The remaining environment is this https://mister-muffin.de/p/o94c.jpg (the lower part happened to be in focus). But with bootcmd set to this it can obviously not work, right? I wonder where I should see the effect of | 22:05 |
| josch | config_distro_bootcmd.h | 22:05 |
| josch | oh and I'm now reading that config_distro_bootcmd.h is already included by include/configs/nitrogen8m_som.h | 22:12 |
| + mjw (~mark@gnu.wildebeest.org) | 22:17 | |
| - mtm (QUIT: Ping timeout: 240 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 22:27 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 22:28 | |
| vagrantc | josch: oh, that was fixed in newer versions | 22:31 |
| vagrantc | i think you can override it by setting BOOTCMD in the board specific file | 22:31 |
| vagrantc | ACTION forgets how these five year old antiques worked | 22:32 |
| vagrantc | josch: probably after include config_distro_boot.h ... something like define #BOOTCMD | 22:36 |
| josch | vagrantc: are there some popular good defaults for memory addresses for u-boot on devices with several gigs of memory and booting big distros like Debian with initramfs? Or is everybody just choosing different arbitrary values? | 22:42 |
| josch | I think the reason that I don't have distro_bootcmd in my env is that config_distro_bootcmd.h populates BOOTENV and that is then appended to CONFIG_EXTRA_ENV_SETTINGS but got commented out in https://source.mnt.re/reform/reform-boundary-uboot/-/commit/b3a8287b13832cd25b2f8568cafdece62043a3db | 22:44 |
| josch | vagrantc: it boots!! :D | 22:50 |
| vagrantc | josch: it's slightly platform dependent | 22:50 |
| vagrantc | i prefer to put the fdt_addr_r the smallest, as it's the least likely to grow, followed by the kernel with a hefty buffer for the initrd | 22:51 |
| vagrantc | e.g. basically in order of the size i expect things to need to grow | 22:54 |
| vagrantc | the patches submitted upstream were > +"fdt_addr_r=0x43000000\0" \ | 22:55 |
| vagrantc | > +"kernel_addr_r=" __stringify(CONFIG_SYS_LOAD_ADDR) "\0" \ | 22:55 |
| vagrantc | > +"ramdisk_addr_r=0x44000000\0" \ | 22:55 |
| vagrantc | +CONFIG_SYS_LOAD_ADDR=0x40480000 | 22:56 |
| vagrantc | huh, guess that put the kernel below the initrd | 22:56 |
| vagrantc | those are things that can change as long as you don't hard-code your environment or scripts | 22:57 |
| vagrantc | josch: there's also #u-boot, but they'll look at your askance for using an ancient u-boot... | 22:59 |
| josch | vagrantc: no need, I got it working now with distro_bootcmd: https://paste.debian.net/1239735/ | 23:04 |
| vagrantc | josch: just getting the right addressses? | 23:05 |
| josch | vagrantc: no, partly reverting b3a8287b13832cd25b2f8568cafdece62043a3db | 23:06 |
| josch | that commit resulted in BOOTENV not being appended to CONFIG_EXTRA_ENV_SETTINGS and thus including config_distro_bootcmd.h had no effect | 23:06 |
| vagrantc | ah! | 23:07 |
| vagrantc | well, you're now well on your way... | 23:07 |
| josch | vagrantc: thanks for all your help!! :D | 23:07 |
| vagrantc | josch: helping you is helping me too :) | 23:09 |
| josch | I'm happy you see it that way. XD | 23:10 |
| vagrantc | josch: your sysimage-v3 has a number of missing pieces i was struggling with to build an mmdebstrap based live image generator... | 23:22 |
| josch | Nice to hear that it works for you now! I was essentially just dogfooding. ;) | 23:23 |
| bluerise | Oh what a re y ou doing? | 23:42 |
| + littlebobeep (~alMalsamo@gateway/tor-sasl/almalsamo) | 23:49 | |
| dustfinger | I heard a rumor that there is a module being worked on to upgrade the reform to 16GB. Is that so? I think it was a third partyi project. | 23:51 |
| littlebobeep | Hi all I just discovered this project, it seems very interesting! Still researching, but I am slightly confused by serpent's post here: https://community.mnt.re/t/why-hdmi-out-instead-of-display-port-out/581/5 | 23:51 |
| littlebobeep | Does this mean there is limited HDMI output that is kinda broken with NO proprietary blob? | 23:51 |
| littlebobeep | also 16GB RAM is quite interesting, but does this mean a different SOC? Who even makes SODIMMs like these besides Boundary Devices in California? Are there any Asian suppliers? | 23:53 |
| littlebobeep | Also I was discouraged by "leather" sleeves but on a different web page it says vegan? | 23:54 |
| littlebobeep | Also does the WLE200NX have a Bluetooth chip on it? I only see mention of AR9280 chip for WLAN... | 23:55 |
| littlebobeep | Also are the trackpad + keyboard firmware upgradeable? If so how do I write to them and what chips store the code? | 23:56 |
| littlebobeep | (sorry for bombing with all these questions....) | 23:56 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
