00:04 <farkuhar> Guest323: SiFuh has something close to what you want, but using only primary partitions, not lvm. It should be easy enough to modify, though. https://raw.githubusercontent.com/SiFuh/Documentation/master/encrypted-CRUX-without-dracut.txt

00:07 <SiFuh> You can also skip the swap rc script if you don't have swap on a seperate partition

00:07 <Guest323> farkuhar: thanks for the text, it looks promising. about to give machine a reboot with lilo and see what happens. if it stays on same situation will look into the text for more details.

00:07 <Guest323> thank you!

00:09 <Guest323> it was fun building lilo with devicemapper support to support encryption because the lilo package coming with crux is built without DM, took a while to fix all deps and errors.

00:11 <SiFuh> Guest323: Also check out this one, with dracut. But as I mentioned before the swap setup might not be what you want. It is specifically how my system has the swap since i don't use hibernation

00:11 <SiFuh> https://raw.githubusercontent.com/SiFuh/Documentation/master/CRUX-3.6-Encrypted.txt

00:14 <Guest323> appreciate! saved both links. its a server so no need for hibernation :) will give it a read if the upcoming reboot fails

00:17 <SiFuh_> Cool, I have mine re-encrypt randomnly swap on boot.

00:18 <SiFuh_> I am also using serpent cypher not the aes in that text

00:24 <Guest323> ran a similar setup before moving to bsd. used serpent-twofish-aes + whirlpool and 2 boot partitions, first boot partition unencrypted that decrypts second boot partition with first password, makes checksums of all files separately and together (kernel,initramfs...) then another password to decrypt the main partition which holds the filesystem.

00:24 <Guest323> this way it protects the data before entering the main decryption pwd. ciphers were kinda slow though probably an overkill.

00:27 <Guest323> heh. "acpi: spcr: unexpected spcr access width. defaulting to byte size. the bios has corrupted hw-pmu resources (msr 38d is 330)"

03:10 <SiFuh_> Heh, adjust volume per program, press pause and the volume drops below the main volume when I continue, what an annoying feature in pulseaudio

08:54 <cruxbot> [opt.git/3.6]: [notify] jdk8-bin: update to 8.312.07.1

08:54 <cruxbot> [opt.git/3.6]: ruby-sqlite3: fix build with latest ruby

08:54 <cruxbot> [opt.git/3.6]: ruby-ncurses: fix build with latest ruby

08:54 <cruxbot> [opt.git/3.6]: ruby-mysql: fix build with latest ruby

08:54 <cruxbot> [opt.git/3.6]: gtk-theme-adwaita: dependency fix, new intltool

08:54 <cruxbot> [opt.git/3.6]: gtk-engines: dependency fix, new intltool

08:54 <cruxbot> [opt.git/3.6]: cups-filters: dependency fix, new xorg-font-dejavu-ttf

09:56 <cruxbot> [opt.git/3.6]: vte: dependency fix, new intltool

09:56 <cruxbot> [opt.git/3.6]: libwnck: dependency fix, new intltool

09:56 <cruxbot> [opt.git/3.6]: graphviz: update to 2.49.2

09:56 <cruxbot> [opt.git/3.6]: php-mbstring: new dependency libpcre2

09:56 <cruxbot> [opt.git/3.6]: quota: use bash for configure, new dependency rpcsvc-proto

09:56 <cruxbot> [opt.git/3.6]: lxappearance: dependency fix, new intltool

10:26 <Guest70> hello. im still trying to install crux+luks+lvm on dl360 g8. i've tried multiple bootloader configs with syslinux and lilo however im encountering the same issue all over again, it starts unpacking initramfs from what i see but fails with 'could not find /dev/sda2'. i find this weird cause same initramfs and same lilo config are being used over 5

10:26 <Guest70> other virtual machines. could this be something specific with bare metal, any ideas?

10:28 <Guest70> here's my lilo.conf http://ix.io/3Chv and my init in initramfs http://ix.io/3Chw

10:30 <Guest70> i would be happy to provide more information on request

10:35 <Guest70> ive tried earlier with nullspoons mkinitramfs on setup without lvm, also same issue. then i wrote custom initramfs and to have sda2 access used major 8 minor 2 block devices created with 'cp -a /dev/sda2 ./initramfs/tree/dev/'

10:36 <Guest70> however it is unable to find 'enc_root=/dev/sda2' from appended kernel parameter. sadly when it drops to initramfs shell it has no keyboard, trying to deal with this now

10:38 <Guest70> so the cryptsetup is unable to find /dev/sda2, maybe it needs some sleep before so the system can find the devices

10:39 <Guest70> ie, if the initramfs is too fast

10:59 <braewoods> Guest70: i would make it drop to a shell so you can probe /dev

11:00 <braewoods> chances are a failure to find boot devices is a timing issue (too soon), lack of kernel support, hardware issue, or configuration issue

11:05 <Guest70> braewoods: initramfs does drop to a shell with 'some_command | exec /bin/sh' however with no keyboard, so i rebooted back to iso to find the corresponding kernel module, it seems its 'hid-generic.ko' included it now. will be making few boot entries with lilo now and try all of them. will be giving it a shot in half an hour and report back. thank

11:05 <Guest70> you for helping out.

11:05 <Guest70> do you know why do these timing issues happen, i experienced these before, is it due to multi-threading?

11:07 <Guest70> thank you.

11:16 <braewoods> Guest70: it's to do with variable hardware detection or startup times.

11:16 <braewoods> to know if this is the issue, you'd need to see if they appear later on in recovery shell

11:17 <braewoods> if they do you got a timing issue where you are trying to use hardware before it is ready

11:17 <braewoods> nothing to do with threading in this case

11:18 <braewoods> the most often way to deal with this is introducing a sleep delay, perhaps in combination with code for testing for the existence of devices.

11:18 <braewoods> but a bulk sleep is far simpler

11:20 <Guest70> appreciate your explanation

12:53 <Guest70> a small update, tried going with dracut however it has its own issues with cryptsetup not asking decryption password. upon dropping to dracut shell and opening luks partition manually, more lvm issues arise. been playing around with my custom initramfs it seems for some reason 'hpsa' kmod for hp 420i raid controller is not loaded even though its

12:53 <Guest70> not built as module but built inside kernel. now im almost positive this is the issue - hpsa. will try building hpsa as module and including other module dependencies in initramfs and see where that goes.

12:54 <Guest70> the plan is to include hpsa.ko , scsi_mod.ko and scsi_transport_sas.ko to initramfs. last two are deps for hpsa.

12:55 <Guest70> all this takes time on a 0.4mbps connection :)

13:16 <SiFuh_> hpsa needs to be a module

13:17 <SiFuh_> Everything you need to access the disk to get to the kernel needs to be a module and packed into your initrd

13:22 <SiFuh_> But I don't have that issue since I use a modular kernel

13:22 <SiFuh_> https://raw.githubusercontent.com/SiFuh/Documentation/master/Linux-Config-5.4.98-modular

13:22 <SiFuh_> https://github.com/SiFuh/Documentation/blob/master/Linux-Config-5.11.1-modular

13:42 <Guest70> SiFuh_: that was it, had to build hpsa as module. now thinking about it, its logical, it needs to be module so it can prepare disks for decryption, since it wasnt it couldn't find /dev/sda2. now everything is clear. it sems i was too tired and not thinking straight. cryptsetup asks for luks password, however i forgot to include whirlpool (wp512.ko)

13:42 <Guest70> so it didn't go through as expect but that is it, should boot after this fix.

13:42 <Guest70> thank you for your help, all :)

13:45 <Guest70> this is my pre-final init http://ix.io/3Cj0 will post initramfs tree once machine boots :)

14:05 <Guest70> ls -R of initramfs tree http://ix.io/3Cjd

14:06 <Guest70> ls -laR http://ix.io/3Cje

14:06 <Guest70> hopefully somebody finds this helpful, thank you

14:13 <Guest-69> booted succesffuly, hurray

14:24 <SiFuh> :-D

14:25 <SiFuh> Guest-69: pretty compact and lite

14:34 <Guest-69> :))) thank you, i have saved your links and will a bit how you did the swap encryption since im interested in that. need to reboot back to bsd. using this windows only for java/ilo/console connections. brb :)

14:35 <Guest-69> s/a bit/a bit later see/

14:36 <cruxbot> [contrib.git/3.6]: hyperfine: initial commit, version 1.12.0

14:36 <cruxbot> [contrib.git/3.6]: lfs: 1.1.0 -> 1.2.0

14:36 <cruxbot> [contrib.git/3.6]: lua-language-server: 2.4.4 -> 2.4.5

14:36 <cruxbot> [contrib.git/3.6]: python3-tzlocal: 4.0 -> 4.0.1

14:36 <cruxbot> [contrib.git/3.6]: seatd: 0.6.2 -> 0.6.3

14:36 <cruxbot> [contrib.git/3.6]: wpebackend-fdo: 1.10.0 -> 1.12.0

14:36 <cruxbot> [contrib.git/3.6]: zellij: 0.18.1 -> 0.19.0

16:53 <SiFuh> bountyht: https://www.youtube.com/watch?v=7gSU_Xes3GQ check this out

17:16 <bountyht> No Youtube capabilities here.

17:17 <SiFuh_> Good news for you. It is a documentary about the Chinese surveiilance system

17:19 <SiFuh_> Surveillance*

17:25 <SiFuh_> jaeger: https://morefines.com/products/morefine-s500-mini-pc A little bigger than the NUC but holy wow. It runs nice

17:37 <Guest-69> its not only china surrveilance but everywhere surveillance. see, the same way i know you typed about surveillance but just entered the channel ;) joke aside. i haven't had yt capabilities either but used mplayer for music so thought why not yt-dlp since its python3 only dependant.

21:08 <SiFuh> Guest-69: Yeah, I use to live in Shanghai for a few years. Left just before the olympic games.