| vagrantc | but then why are kernel upgrades "without effect" ? | 00:00 |
|---|---|---|
| josch | Because for them to take effect I have to mount a partition from the SD-Card to /boot | 00:00 |
| vagrantc | and if you do that *something* then puts some files in the wrong place? | 00:00 |
| josch | If I don't have a separate /boot partition on the SD-Card, then yes. | 00:01 |
| vagrantc | there seems to be something unconventional or maybe legacy in the configuration ... these sorts of workarounds aren't typically needed with most of the platforms i work with | 00:01 |
| vagrantc | unless you don't *want* the /boot to be from the microSD | 00:02 |
| josch | It has to be either on the SD-Card or on eMMC but I cannot boot directly from NVMe. | 00:02 |
| vagrantc | right | 00:02 |
| josch | So if my root filesystem is on another medium than the one used for boot, the separate /boot partition is needed. | 00:03 |
| vagrantc | the /boot/boot stuff might be from the extra flash-kernel options you pass | 00:03 |
| josch | No. | 00:04 |
| josch | Let me explain again step by step. In the booted system, I want to have /boot/vmlinuz, correct? | 00:04 |
| vagrantc | sure... or /boot/vmlinuz-ABI | 00:04 |
| josch | Yes, something like that. | 00:05 |
| josch | But if my booted system is not on the SD-Card or eMMC, then I have to populate /boot by mounting a partition from SD-Card or eMMC, correct? | 00:05 |
| vagrantc | sure | 00:05 |
| josch | Lets say we choose the SD-Card. | 00:05 |
| josch | If the SD-Card does *not* have a separate /boot partition then the partition on the SD-Card will have a subdirectory called /boot containing vmlinuz. | 00:06 |
| vagrantc | although, "populate" ... i would just normally have /boot mounted all the time | 00:06 |
| josch | So when I mount that partition onto /boot I will get /boot/boot/vmlinuz | 00:06 |
| josch | Yes, but that thing that is mounted must be a /boot partition with vmlinuz at its root | 00:07 |
| vagrantc | so ... when you mount that partition onto /boot is where this starts to go strange ... | 00:07 |
| josch | If there is no separate /boot partition then vmlinuz will be in the /boot subdirectory and that will put it into the wrong place when mounting it to /boot | 00:07 |
| vagrantc | where does the /boot/boot come from? | 00:07 |
| vagrantc | what will put it into the wrong place? | 00:08 |
| josch | I have no idea where you lost me. | 00:08 |
| vagrantc | < josch> So when I mount that partition onto /boot I will get /boot/boot/vmlinuz | 00:08 |
| vagrantc | i don't follow you here... | 00:08 |
| josch | Suppose my SD-Card does not have a separate /boot partition. | 00:09 |
| josch | Then vmlinuz will be in /boot/vmlinuz on that partition, correct? | 00:09 |
| vagrantc | sure | 00:09 |
| josch | If I mount that partition into /boot, then vmlinuz will be in /boot/boot/vmlinuz | 00:09 |
| vagrantc | ok, but, that sounds like a broken configuration ... ? | 00:09 |
| josch | Yes. | 00:10 |
| josch | That's why we use a separate /boot partition. | 00:10 |
| josch | Because on that partition, vmlinuz will be in the root of the partition. | 00:10 |
| vagrantc | that sounds normal. | 00:10 |
| josch | And that's why we want a separate /boot partition. | 00:10 |
| josch | If we wouldn't have it, we could still boot from NVMe but we couldn't upgrade the kernel. | 00:11 |
| vagrantc | but, shouldn't your NVMe install just mount a partition from micrSD or eMMC as /boot ? | 00:11 |
| josch | Yes, that's what it does. It mounts the separate /boot partition. | 00:11 |
| vagrantc | so what is placing the kernel at the wrong location? | 00:12 |
| josch | It would be placed in the wrong location if we would not have a separate /boot partition. | 00:12 |
| vagrantc | in that configuration ... e.g. rootfs on NVMe, /boot on micrSD | 00:12 |
| josch | Then everything works. | 00:12 |
| vagrantc | i guess i'm lost at how it gets placed in the "wrong" location at all, or what problem is trying to solve what issue :) | 00:13 |
| vagrantc | that just sounds like a normal configuration to me. | 00:13 |
| josch | I was trying to explain why we need a separate /boot partition. | 00:13 |
| josch | (see above) | 00:14 |
| josch | And why it will not work without one. | 00:14 |
| josch | i.e. the SD-Card having one single partition for / | 00:15 |
| vagrantc | do you mean you have the SD-card install have a split boot partition, even though it wouldn't strictly need it, to keep the configuration more similar for FDE or NVMe rootfs? | 00:16 |
| vagrantc | as, the SD-Card could work either way, split boot or no split boot ... | 00:16 |
| vagrantc | ACTION does have a vague memory of having to have /boot/boot -> . as a symlink in some cases to work around things | 00:18 |
| vagrantc | but those were very old configurations, haven't had to do that for ages | 00:18 |
| josch | Yes, one SD-Card image can boot from the SD-Card itself, from NVMe or from NVMe with FDE | 00:18 |
| - Christoph_ (QUIT: Remote host closed the connection) (~Christoph@p54bf6380.dip0.t-ipconnect.de) | 00:37 | |
| flowy | josch: in any case, i really appreciate the new configuration. thank you! | 01:29 |
| flowy | after booting it from sd i used the reform-setup-encrypted-nvme and reform-migrate scripts. i made some minor modifications: so that luks goes on a proper partition rather than the block device, and transfered /boot to a 1gb swissbit microsd | 01:34 |
| flowy | also liking the networking tweaks that were made. i was having dns struggles with the v2 image when using wireguard and that went away with v3 | 01:36 |
| flowy | also it's very nice to have all the kernel modules available in v3. tweaking the kernel config every time i found something else that i forgot... was getting tiresome | 01:41 |
| flowy | anyone got a bluetooth dongle recommendation for the reform | 01:44 |
| vagrantc | seems like i should get or build a v3 image... | 01:48 |
| vagrantc | oh, the reform-system-image stuff all assumes cross-compilation? | 01:49 |
| vagrantc | this is ringing bells | 01:49 |
| vagrantc | requires both multistrap and mmdebstrap? :) | 01:52 |
| - mtm (QUIT: Ping timeout: 246 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 02:04 | |
| * jfred -> jfred-server | 02:07 | |
| + jfred (uid534649@libera/sponsor/jfred) | 02:09 | |
| - oomono (QUIT: Quit: Connection closed for inactivity) (uid328183@id-328183.tinside.irccloud.com) | 02:11 | |
| - adjtm (QUIT: Remote host closed the connection) (~adjtm@2a0c:5a80:1c01:3a00:b499:a8d6:4d3e:78e8) | 02:13 | |
| + adjtm (~adjtm@2a0c:5a80:1c01:3a00:b499:a8d6:4d3e:78e8) | 02:13 | |
| + jomo8 (~jomo@user/jomo) | 02:14 | |
| + ndufresne3 (~ndufresne@gyros.collabora.co.uk) | 02:14 | |
| - tadeus_brick (QUIT: Quit: Stay Cheeki Breeki) (tadeustad@psifactor.pl) | 02:14 | |
| - jomo (QUIT: Read error: Connection reset by peer) (~jomo@user/jomo) | 02:14 | |
| * jomo8 -> jomo | 02:14 | |
| + TadeusTaD (tadeustad@psifactor.pl) | 02:14 | |
| - ndufresne (QUIT: Ping timeout: 252 seconds) (~ndufresne@gyros.collabora.co.uk) | 02:16 | |
| * ndufresne3 -> ndufresne | 02:16 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 03:22 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 04:10 | |
| - xktr (QUIT: Ping timeout: 252 seconds) (~xktr@37.120.147.186) | 07:04 | |
| josch | flowy: why would you put luks into a partition? | 07:10 |
| + xktr (~xktr@37.120.147.186) | 07:16 | |
| + MajorBiscuit (~MajorBisc@2a02:a461:129d:1:193d:75d8:745d:e91e) | 09:35 | |
| + Christoph_ (~Christoph@p4fe73dc6.dip0.t-ipconnect.de) | 11:08 | |
| + Major_Biscuit (~MajorBisc@c-001-024-026.client.tudelft.eduvpn.nl) | 11:26 | |
| - Major_Biscuit (QUIT: Client Quit) (~MajorBisc@c-001-024-026.client.tudelft.eduvpn.nl) | 11:27 | |
| + Major_Biscuit (~MajorBisc@c-001-024-026.client.tudelft.eduvpn.nl) | 11:27 | |
| - MajorBiscuit (QUIT: Ping timeout: 256 seconds) (~MajorBisc@2a02:a461:129d:1:193d:75d8:745d:e91e) | 11:29 | |
| - erle (QUIT: Ping timeout: 240 seconds) (~erle@ip5f5bf491.dynamic.kabel-deutschland.de) | 11:59 | |
| + erle (~erle@ip5f5bf491.dynamic.kabel-deutschland.de) | 12:12 | |
| - q66 (QUIT: Remote host closed the connection) (~q66@q66.moe) | 12:42 | |
| + q66 (~q66@q66.moe) | 13:31 | |
| + mjw (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 14:02 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 14:04 | |
| mjw | Your Crowd Supply order (117129) has shipped! | 15:24 |
| mjw | exciting | 15:24 |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 15:35 | |
| minute | mjw: nice | 15:35 |
| minute | mjw: which country are you in, and did you order wifi? | 15:36 |
| mjw | minute, I am in the Netherlands, but the order is going out to the US (Portland actually, so hopefully it will arrive in the next day or so). It is a gift of a fully assembled mnt reform, including wifi. | 15:46 |
| minute | i see, thanks | 15:52 |
| vkoskiv_ | mjw: When did you order? | 16:21 |
| mjw | vkoskiv_, Placed on Mar 23, 2021 | 16:23 |
| mjw | so almost precisely a year ago | 16:23 |
| vkoskiv_ | I ordered in september, the shipped orders are rapidly approaching it :D | 16:23 |
| mjw | Originally estimated to ship on Apr 28, 2022 | 16:24 |
| mjw | Actually shipped on Mar 28, 2022 | 16:24 |
| mjw | Although I believe they had an earlier ship estimate earlier. | 16:25 |
| vkoskiv_ | I had it on Dec 9th last year, but it went to Apr 28th and is still there. | 16:26 |
| mjw | too many earliers, sorry | 16:26 |
| vkoskiv_ | Feeling pretty good about April if this pace of shipments continues. | 16:26 |
| mjw | right, I believe the original estimate was also December for me, then it flipped to April, but it actually shipped in March. | 16:27 |
| flowy | josch: my nvme ssd is more than large enough (1TB) so i decided to leave some space at the beginning for any possible future migration stuff. but also, i did some reading and putting LUKS on a partition even if it is full-disk encryption seems to be the standard? i believe this is what the debian installer does. you always create a partition for LUKS. feels weird not having a partition table haha. but i'm | 16:49 |
| flowy | not sure if there are any concrete benefits. | 16:49 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 16:50 | |
| flowy | at the very least i can imagine future me incorrectly assuming there's no data on the disk if it was lying around for a year | 16:51 |
| flowy | i have a lot of disks | 16:51 |
| flowy | drives rather | 16:52 |
| flowy | much tb | 16:54 |
| + andrey_utkin (~andrey_ut@gentoo/developer/andrey-utkin) | 17:02 | |
| Boostisbetter | I got a tb in my Reform, and I don't seem myself EVER filling that thing up. Hahaha. | 17:07 |
| - andrey_utkin (QUIT: Quit: Gateway shutdown) (~andrey_ut@gentoo/developer/andrey-utkin) | 17:11 | |
| + andrey_utkin (~andrey_ut@gentoo/developer/andrey-utkin) | 17:11 | |
| Boostisbetter | minute: saw the tweet about the pocket reform display sample. How are you liking it? Does it seem like it will be a candidate? BTW, if you need someone to real-world test prototypes, I will happily pay for the spot. | 17:35 |
| Boostisbetter | josch: thanks for taking the time to post all the charts on the encryption performance overhead question. | 17:35 |
| flowy | yeah i think you don't really have to worry about overhead. it's gonna be pretty rare where that IO becomes your bottleneck on the reform | 17:37 |
| flowy | esp since dm-crypt utilises multiple cores, so it shouldn't steal much from single threaded stuff | 17:38 |
| - indefini[m] (QUIT: Quit: You have been kicked for being idle) (~indefinim@2001:470:69fc:105::1e2a) | 18:00 | |
| - Major_Biscuit (QUIT: Ping timeout: 260 seconds) (~MajorBisc@c-001-024-026.client.tudelft.eduvpn.nl) | 18:00 | |
| + Major_Biscuit (~MajorBisc@2a02:a461:129d:1:193d:75d8:745d:e91e) | 18:18 | |
| + MajorBiscuit (~MajorBisc@86-88-79-148.fixed.kpn.net) | 18:18 | |
| - MajorBiscuit (QUIT: Client Quit) (~MajorBisc@86-88-79-148.fixed.kpn.net) | 18:19 | |
| - Major_Biscuit (QUIT: Ping timeout: 250 seconds) (~MajorBisc@2a02:a461:129d:1:193d:75d8:745d:e91e) | 18:36 | |
| + alex4nder (~alexander@ip98-171-191-229.sb.sd.cox.net) | 18:43 | |
| alex4nder | hey | 18:43 |
| - mjw (QUIT: Quit: Leaving) (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 19:08 | |
| - alex4nder (QUIT: Ping timeout: 272 seconds) (~alexander@ip98-171-191-229.sb.sd.cox.net) | 19:11 | |
| + alex4nder (~alexander@wsip-184-186-225-187.sb.sd.cox.net) | 20:00 | |
| - xktr (QUIT: Ping timeout: 260 seconds) (~xktr@37.120.147.186) | 20:24 | |
| Boostisbetter | Howdy | 20:27 |
| alex4nder | hey | 20:27 |
| + xktr (~xktr@37.120.147.186) | 20:31 | |
| flowy | yo | 20:54 |
| - xktr (QUIT: Ping timeout: 256 seconds) (~xktr@37.120.147.186) | 20:59 | |
| + xktr (~xktr@37.120.147.186) | 21:01 | |
| * Guest7118 -> nsc | 21:25 | |
| - alex4nder (QUIT: Ping timeout: 256 seconds) (~alexander@wsip-184-186-225-187.sb.sd.cox.net) | 21:47 | |
| - mtm (QUIT: Ping timeout: 272 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 22:15 | |
| ex-parrot | hello | 23:26 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!