| minute | eibachd: awesome | 00:00 |
|---|---|---|
| - robin (QUIT: Quit: Leaving) (~robin@user/terpri) | 00:05 | |
| + robin (~robin@user/terpri) | 00:52 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 01:02 | |
| - eibachd (QUIT: Ping timeout: 260 seconds) (~eibachd@p200300dcf73d8b00537e4b194d65f380.dip0.t-ipconnect.de) | 01:23 | |
| + eibachd (~eibachd@p200300dcf73d8b008af5fa8d029d5a67.dip0.t-ipconnect.de) | 01:23 | |
| - mjw (QUIT: Remote host closed the connection) (~mjw@gnu.wildebeest.org) | 01:36 | |
| + mjw (~mjw@gnu.wildebeest.org) | 01:36 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:09 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:08 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-140-33.tukw.qwest.net) | 03:33 | |
| + colinsane (~colinunin@97-113-140-33.tukw.qwest.net) | 03:35 | |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@21-98-142-46.pool.kielnet.net) | 03:58 | |
| + nsc (~nicolas@46.142.49.137) | 04:00 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 04:02 | |
| josch | eibachd: i really honestly appreciate that you wrote your article in a way that makes even people like me understand what is going on. Really impressive! Thank you for the work and the write-up! :D | 06:13 |
| josch | eibachd: your copy of u-boot did have sata_boot available but it did not automatically run that with the default distro_bootcmd, right? | 06:15 |
| josch | i think it would not hurt to modify the u-boot config so that it would automatically try running sata_boot as part of the default boot sequence | 06:15 |
| - vkoskiv (QUIT: Ping timeout: 268 seconds) (~vkoskiv@87-95-190-139.bb.dnainternet.fi) | 07:31 | |
| eibachd | josch: yes, I had to modify the bootcmd. Including that in the standard config would be cool. I did not find where the root entry for the kernel command line would come from in that case. | 08:40 |
| eibachd | josch: really glad you liked the article. | 08:41 |
| josch | :) | 08:59 |
| josch | # CONFIG_SATA_BOOT is not set | 09:01 |
| josch | changing this in the u-boot config might already be enouguh | 09:01 |
| + vkoskiv (~vkoskiv@87-100-163-117.bb.dnainternet.fi) | 09:03 | |
| josch | eibachd: i created https://source.mnt.re/reform/reform-ls1028a-uboot/-/merge_requests/3 and will test it as soon as I have my second mainboard that i can install my ls1028a to | 09:07 |
| josch | because testing this should yield the same results with u-boot on sd-card (which is of course the case for me) | 09:07 |
| f_ | eibachd: Nice article :) | 09:07 |
| eibachd | josch: Thanks, I will give it a test on my system. | 09:08 |
| eibachd | f_: thanks, glad you liked it | 09:09 |
| f_ | Wish I could write good blog posts | 09:10 |
| f_ | I'm terrible at blogs. | 09:10 |
| eibachd | I have set up an environment, where I enjoy writing. That helps a lot. | 09:11 |
| f_ | an ssg you mean? | 09:14 |
| f_ | https://blog.vitali64.duckdna.org/ :P | 09:15 |
| f_ | * https://vitali64.duckdns.org/ | 09:15 |
| f_ | * https://blog.vitali64.duckdns.org/ | 09:15 |
| eibachd | Yes, there is also some CI involved ;) https://zeromips.org/posts/2022-03-04-website/ | 09:22 |
| f_ | I also use an ssg for my blog | 09:24 |
| eibachd | So it comes down to text editor and beverage of choice and that's it. | 09:27 |
| f_ | For me too, the problem I have is...no inspiration :) | 09:29 |
| eibachd | For me it is about writing down my own adventures, so I will remember them later. | 09:32 |
| f_ | Yeah that, but I mean, I have no inspiration on how to write it all. | 09:48 |
| f_ | Anyway yes, that's a good blog post ;) | 09:48 |
| - Boostisbetter (QUIT: Ping timeout: 256 seconds) (4a410829d7@irc.cheogram.com) | 09:51 | |
| eibachd | f_: The link to the sourcecode in your ssg post seems to be broken | 10:20 |
| f_ | Yeah, it's actually https://git.vitali64.duckdns.org/utils/seen.git branch: flask | 10:20 |
| hramrach | there isn't really 'how to write it' - you write something, and it's written | 10:32 |
| hramrach | btw duckdns using google analytics, eww | 10:33 |
| f_ | duckdns.org? | 10:41 |
| f_ | my website doesn't use google analytics. | 10:41 |
| + reform20442 (~mauro@2a02:a03f:ea78:f501:cedc:c4c3:9993:354f) | 10:48 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 11:12 | |
| - XYZ (QUIT: Ping timeout: 268 seconds) (~XYZ@37-48-32-137.nat.epc.tmcz.cz) | 11:30 | |
| - reform20442 (QUIT: Quit: Leaving) (~mauro@2a02:a03f:ea78:f501:cedc:c4c3:9993:354f) | 11:36 | |
| + XYZ (~XYZ@37-48-40-37.nat.epc.tmcz.cz) | 11:46 | |
| - XYZ (QUIT: Read error: Connection reset by peer) (~XYZ@37-48-40-37.nat.epc.tmcz.cz) | 11:46 | |
| - eibachd (QUIT: Ping timeout: 260 seconds) (~eibachd@p200300dcf73d8b008af5fa8d029d5a67.dip0.t-ipconnect.de) | 12:10 | |
| + eibachd (~eibachd@p200300dcf73d8b007e9b42d38e52e335.dip0.t-ipconnect.de) | 12:10 | |
| josch | eibachd: i didn't test it yet but looked at the raw flash.bin and it shouldn't work yet because we also have to adjust boot_targets... let me figure out how... | 12:12 |
| + mjw (~mjw@gnu.wildebeest.org) | 12:14 | |
| josch | minute: in include/configs/ls1028a_common.h, is the "setenv xtrace yes" still necessary? | 12:21 |
| eibachd | We also will have to modify build.sh in reform-ls1028a-uboot, so birth images for SD and eMMC get built. | 12:22 |
| eibachd | *both* | 12:22 |
| josch | eibachd: what's the difference for u-boot that is put on emmc? | 12:24 |
| eibachd | The ATF build must be called with a different BOOT_MODE parameter, emmc instead of sd. This results on a different PBI code. Remember the block copy in my | 12:28 |
| eibachd | Write-up? It has a src parameter 9 for eMMC, 8 for SD. | 12:28 |
| josch | ah BOOT_MODE=sd | 12:29 |
| josch | right | 12:29 |
| josch | is setting BOOT_MODE=emmc all that needs to be done? | 12:30 |
| eibachd | Exactly | 12:30 |
| josch | okay, i'll write that up after lunch | 12:31 |
| - lexik (QUIT: Quit: Bella ciao.) (~lexik@93.185.97.209) | 12:31 | |
| josch | i pushed a new commit to that MR which now sets | 12:31 |
| josch | boot_targets=mmc0 mmc1 usb0 sata0 | 12:31 |
| josch | (not tested yet of course) | 12:31 |
| eibachd | I don't know if there is more changes resulting from this parameter. As I wrote, I cannot access SD anymore. I will examine this later. | 12:31 |
| hramrach | yes. I am looking at the dynamic dns people use because mine was too annoying to keep going | 12:42 |
| hramrach | also it turns out my ISP no longer provides public IP addresses, it was swallowed by another one, and apparently the IP space is no longer sufficient, and they cannot do v6 because they suck | 12:43 |
| - ggoes (QUIT: Quit: ZNC - https://znc.in) (~gregf@fsf/staff/ggoes) | 12:46 | |
| + ggoes (~gregf@fsf/staff/ggoes) | 12:48 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 13:04 | |
| - Boostisbetter (QUIT: Ping timeout: 255 seconds) (4a410829d7@irc.cheogram.com) | 13:13 | |
| + lexik (~lexik@93.185.97.209) | 13:36 | |
| + BoostisBetter (4a410829d7@irc.cheogram.com) | 13:46 | |
| erle | they cannot do v6 because they suck, in $CURRENT_YEAR … oof | 13:47 |
| erle | eibachd i like how detailed your blog posts are. and all images are on-point, feels kinda like a rarity nowadaysj | 13:49 |
| hramrach | not quite detailed enough if you wanted to replicate it | 14:16 |
| eibachd | erle: thanks for the feedback, that is really encouraging | 14:35 |
| eibachd | hramrach: right, I am not doing step by step tutorials. My goal is to write texts that are actually fun reading. | 14:37 |
| erle | hramrach that is something i would probably notice at some point if i tried to reproduce it, but i do not. i must say detailed texts that are not *quite* detailed enough are frustrating (most commonly scientific papers) | 14:37 |
| minute | btw not sure if i mentioned it, but the rest of the 50x LS1028A modules batch is on the way to us | 14:49 |
| minute | should arrive in the coming week if there are no customs issues | 14:49 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 15:10 | |
| erle | eibachd as (maybe?) part of your audience i would appreciate it even more detailed actually, but take care: only a fraction of users will try to replicate it at all, i mean i probably won't. | 15:12 |
| erle | eibachd maybe you can just link to a repo or tarball containing your full instructions at the end? | 15:12 |
| erle | eibachd assuming you have actually written down everything ^_^ | 15:13 |
| minute | i think some of it is already in an MR to official repos | 15:20 |
| - eibachd (QUIT: Ping timeout: 246 seconds) (~eibachd@p200300dcf73d8b007e9b42d38e52e335.dip0.t-ipconnect.de) | 15:21 | |
| + eibachd (~eibachd@2a01:599:207:e431:cd2e:4731:8d7a:8ccb) | 15:25 | |
| - eibachd (QUIT: Read error: Connection reset by peer) (~eibachd@2a01:599:207:e431:cd2e:4731:8d7a:8ccb) | 15:45 | |
| + eibachd (~eibachd@p200300dcf73d8b009df7d895d963fc5b.dip0.t-ipconnect.de) | 15:46 | |
| josch | okay, the u-boot changes from above MR indeed seem to trigger the correct changes in the u-boot environment: https://paste.debian.net/hidden/aae08273/ | 15:53 |
| josch | now i'll let the pipeline create two artifacts: one for sd and one for emmc | 15:54 |
| josch | i'll use the chance for a bit more clean-up of build.sh | 15:54 |
| eibachd | erle: I like the idea of adding additional documents for download. There are no full written down instructions, but I could add some notes and drawings. | 15:55 |
| eibachd | As the process of replicating will void your warranty, involves delicate soldering and might result in a non-working system I really don't want to push anybody into this. | 15:55 |
| josch | minute: do you think we can move ls1028a u-boot from v2023.10-rc4 to v2023.10 | 15:57 |
| eibachd | josch: great, thanks. Have you found where the "root" parameter for the kernel commandline is coming from? | 15:57 |
| eibachd | Because 'run sata_boot' won't set it. | 15:58 |
| josch | eibachd: it should work without root= as the initramfs should contain all the information about where the rootfs is and what needs to be mounted where | 15:58 |
| josch | my /proc/cmdline certainly does not contain it | 15:58 |
| josch | so if your /etc/fstab is correct, re-generating your initramfs should just do the right thing | 15:59 |
| minute | josch: sure @ move | 15:59 |
| josch | nice -- i'll leave that MR as draft until everything has been tested | 16:00 |
| eibachd | Okay, that sounds good. How would I properly regenerate the initramfs? | 16:00 |
| josch | eibachd: sudo mkinitramfs -u | 16:01 |
| josch | errr sorry | 16:01 |
| josch | sudo update-initramfs -u | 16:01 |
| eibachd | Great, will try later. | 16:02 |
| hramrach | well, there is the thing ho you reference the root filesystem | 16:15 |
| hramrach | if by device name like mmcblk0p1 then you are screwed when you move the filesystem to another device | 16:16 |
| hramrach | if by uuid or label it should just work | 16:16 |
| erle | eibachd you can be assured that voiding the warranty is a thing with MNT customers ;) | 16:17 |
| erle | eibachd i mean i would definitely *know* i am doing that if i was doing it | 16:17 |
| erle | of course my reform is pristine hahaha | 16:18 |
| erle | :P | 16:18 |
| erle | eibachd in any case, i like your blog posts. please continue! | 16:18 |
| BoostisBetter | eibachd: can I get a link to your blog? | 16:25 |
| eibachd | zeromips.org | 16:26 |
| - sundog20629 (QUIT: Quit: The Lounge - https://thelounge.chat) (~sundog@164.subnet-151-210-103.ellijay.com) | 16:26 | |
| josch | eibachd: so the u-boot binaries for sd and emmc differ a lot but the important change is probably this one at the very beginning: https://paste.debian.net/1299284/ | 16:28 |
| josch | where it says 09 for emmc and 08 for sd | 16:28 |
| eibachd | Yes, that is the PBI code. | 16:29 |
| josch | i wonder if it's possible to just patch a u-boot.bin for sd like that to create a working u-boot.bin for emmc? | 16:30 |
| josch | then i could avoid building atf twice and just patch the correct location | 16:31 |
| eibachd | As I said,I am not sure what else gets changed. | 16:33 |
| eibachd | I will investigate further, as my SD card is not working at the moment. | 16:35 |
| josch | i'll leave it as a double-build for now but verify at the end that this is the only diff between the two binaries | 16:35 |
| eibachd | Makes sense | 16:42 |
| josch | okay, sd and emmc ls1028a u-boot binaries now here: https://source.mnt.re/reform/reform-ls1028a-uboot/-/jobs/2621/artifacts/browse | 17:06 |
| josch | the list of changes in the MR became a bit bigger than intended :) https://source.mnt.re/reform/reform-ls1028a-uboot/-/merge_requests/3 | 17:06 |
| josch | will test once i'm back on my ls1028a soon! | 17:07 |
| - eibachd (QUIT: Ping timeout: 256 seconds) (~eibachd@p200300dcf73d8b009df7d895d963fc5b.dip0.t-ipconnect.de) | 17:33 | |
| + eibachd (~eibachd@p200300dcf73d8b0165c2c554b91925ff.dip0.t-ipconnect.de) | 17:39 | |
| - eibachd (QUIT: Ping timeout: 268 seconds) (~eibachd@p200300dcf73d8b0165c2c554b91925ff.dip0.t-ipconnect.de) | 18:22 | |
| + eibachd (~eibachd@p200300dcf73d8b00e0eb9b18b91c2691.dip0.t-ipconnect.de) | 18:23 | |
| - ggoes (QUIT: Quit: ZNC - https://znc.in) (~gregf@fsf/staff/ggoes) | 18:26 | |
| + ggoes (~gregf@fsf/staff/ggoes) | 18:27 | |
| - eibachd (QUIT: Ping timeout: 240 seconds) (~eibachd@p200300dcf73d8b00e0eb9b18b91c2691.dip0.t-ipconnect.de) | 18:37 | |
| + eibachd (~eibachd@p200300dcf73d8b0158b3e67d26667452.dip0.t-ipconnect.de) | 18:38 | |
| - eibachd (QUIT: Ping timeout: 276 seconds) (~eibachd@p200300dcf73d8b0158b3e67d26667452.dip0.t-ipconnect.de) | 19:16 | |
| + eibachd (~eibachd@p200300dcf73d8b007cce257666741f27.dip0.t-ipconnect.de) | 19:16 | |
| + eibachd1 (~eibachd@p200300dcf73d8b0055a01349ef6828fc.dip0.t-ipconnect.de) | 20:54 | |
| + XYZ (~XYZ@37-48-40-37.nat.epc.tmcz.cz) | 21:02 | |
| eibachd1 | josch: Damn. That did not go so well: ERROR: EMMC driver not present. Check your BUILD | 21:11 |
| eibachd1 | Returned to my manually built flash.bin for eMMC and everything works. Phew. And you were right, after update-initramfs I do not need the root option in the kernel commandline anymore. | 21:25 |
| eibachd1 | I guess I will leave it there cause my australian exchange student will need it this week for some more EtherCAT work. | 21:28 |
| - eibachd1 (QUIT: Quit: Client closed) (~eibachd@p200300dcf73d8b0055a01349ef6828fc.dip0.t-ipconnect.de) | 21:52 | |
| hramrach | josch: looks like you are missing some EOL backslashes in that boot script https://source.mnt.re/reform/reform-ls1028a-uboot/-/merge_requests/3/diffs?commit_id=0bc4da6c6a2832755279be720bf760382b5fb0a0 | 22:08 |
| josch | hramrach: reading the diff of a diff is hard. Here is the actual diff after getting patched: https://source.mnt.re/reform/reform-ls1028a-uboot/-/blob/0bc4da6c6a2832755279be720bf760382b5fb0a0/patches/0001-tweak-boot-commands.patch | 22:11 |
| josch | hramrach: do you still see a missing backslash somewhere? | 22:11 |
| josch | if it's still there then i need help because i don't see it :D | 22:12 |
| hramrach | when applied it looks fine | 22:16 |
| josch | the error message "EMMC driver not present." seems to come from ATF not u-boot | 22:18 |
| hramrach | I guess the coloring gitlab uses does not help | 22:18 |
| josch | funnily in the code just below it says /* Should never reach here */ | 22:18 |
| josch | now i wonder how i built my stuff differently from how Dirk built it | 22:18 |
| hramrach | did you build the ATF? | 22:19 |
| josch | atf is built from https://source.mnt.re/reform/reform-ls1028a-atf.git | 22:20 |
| eibachd | josch: I did a full rebuild of the ATF. Maybe some dependency in the makefile is wrong. | 22:21 |
| josch | yes, that's a possiblity -- maybe i just should do a "git clean -fdx" instead | 22:21 |
| eibachd | That should do it. | 22:22 |
| josch | we'll get back to this topic once you and/or I can test this again :) | 22:22 |
| eibachd | Yep | 22:23 |
| josch | eibachd: do you have an account on source.mnt.re? Otherwise i can also leave that error message you got in that MR so that we don't forget the current status by then. | 22:25 |
| eibachd | josch: I have | 22:27 |
| hramrach | the message is gnerated when #if defined(EMMC_BOOT) && defined(NXP_ESDHC2_ADDR) is not true | 22:27 |
| BoostisBetter | anyone know if it is possible getting protonmail bridge running on the Reform? | 22:35 |
| f_ | BoostisBetter: no idea, but I've been using hydroxide for proton | 22:37 |
| eibachd | josch: done | 22:37 |
| f_ | perhaps try that | 22:37 |
| minute | BoostisBetter: i've taken a quick look out of curiosity, it's a amd64 binary | 22:38 |
| minute | box64 crashes > 18176|0x46ad69: Unimplemented Opcode (00 01 75 10) 0F 01 F9 48 C1 E2 20 48 01 D0 48 89 44 24 08 | 22:39 |
| minute | so, no dice. better use open source solutions i guess | 22:39 |
| minute | BoostisBetter: f_'s recommendation looks good (https://github.com/emersion/hydroxide) | 22:40 |
| - bkeys (QUIT: Ping timeout: 255 seconds) (~Thunderbi@45.134.140.153) | 23:13 | |
| + bkeys (~Thunderbi@45.134.140.153) | 23:20 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!