| 4 Aug 2022 |
 @elvishjerricco:matrix.org | Winter (she/her): The problem is if those FSes can't be unmounted, such as with / | 03:08:11 |
| @elvishjerricco:matrix.org | So you have to transition into another ramfs to unmount oldroot, so that you can close the mapper devices | 03:08:30 |
| @elvishjerricco:matrix.org | It's nowhere near as complicated as initrd though. It's literally just the systemd-shutdown binary. So it just loops over killing live processes, unmounting file systems, and closing devices, until a loop has nothing to do. Then it runs shutdown hooks | 03:09:34 |
| 8 Aug 2022 |
| @elvishjerricco:matrix.org | man systemd-network-generator
Ah the many features we just get for free out of the box by using systemd stuff in initrd
| 03:21:37 |
 Winter (she/her) | (this is good, and not sarcasm, right?) | 03:28:27 |
| @elvishjerricco:matrix.org | Winter (she/her): Yea, I'm being serious. I noticed there's a whole chunk of code for handling cmdline stuff for networking in initrd, and then I noticed systemd just does that already | 03:30:43 |
| @elvishjerricco:matrix.org | And that's happened several times. Like IIRC we were surprised to find that hibernation resume just worked when none of us had actually put any effort into supporting it | 03:33:38 |
| Winter (she/her) | thank you lennart & co | 03:47:15 |
| Winter (she/her) | is it okay if I PR a test for a feature that was merged a few days ago (bind mounts in fstab), or is too small of a feature to test? (maybe too small for its own file, but maybe we should make a systemd-initrd-misc.nix file for those?) | 03:55:52 |
| @elvishjerricco:matrix.org | It's pretty rare for a nixos test to be a bad idea :P | 03:56:45 |
| Winter (she/her) | fair, but organization wise, thoughts? are there any more misc tests we could group together with it eventually? | 04:00:41 |
| @elvishjerricco:matrix.org | I wouldn't worry too much about it. IIRC there are plenty of very small tests in nixpkgs/nixos/tests | 04:01:25 |
| Winter (she/her) | TBH we should probably do a wider organization effort for those... | 04:04:22 |
| Winter (she/her) | Like why didn't we just copy the module structure? | 04:04:30 |
| Winter (she/her) | Who's idea was it to have one flat tree lol | 04:04:38 |
| @elvishjerricco:matrix.org | Lol I'm not too worried about it. There's probably some test for file systems in initrd that you could throw this one into | 04:05:30 |
| Winter (she/her) | Doesn't look like we have one for fs, just simple + sw raid | 04:08:15 |
| Winter (she/her) | * Doesn't look like we have one for fs, just sw raid | 04:08:35 |
| Winter (she/her) | (I'm looking at the systemd initrd specific ones) | 04:08:50 |
| @elvishjerricco:matrix.org | I cannot for the life of me figure out why this nixos test is failing... It's the inird-network-ssh test and I can't get the client to connect to the server. It gets permission denied (public key | 04:20:47 |
| @elvishjerricco:matrix.org | * I cannot for the life of me figure out why this nixos test is failing... It's the inird-network-ssh test and I can't get the client to connect to the server. It gets permission denied (public key) | 04:20:49 |
| @elvishjerricco:matrix.org | * I cannot for the life of me figure out why this nixos test is failing... It's the inird-network-ssh test and I can't get the client to connect to the server. It gets permission denied (publickey) | 04:20:56 |
| Winter (she/her) | Is it failing on master? | 04:24:08 |
| Winter (she/her) | (as in, could i repro it right now?) | 04:24:15 |
| @elvishjerricco:matrix.org | nah it's my worktree. Trying to adapt the test for the systemd based initrd networking | 04:24:41 |
| Winter (she/her) | ah | 04:24:47 |
| Winter (she/her) | make sure to tell us when you find the one character typo causing it 😆 | 04:25:14 |
| @elvishjerricco:matrix.org | is there a convenient command to double check that a given public key goes with a particular private key? | 04:25:48 |
| @elvishjerricco:matrix.org | because at this point that's all I can think of... | 04:25:59 |
| @elvishjerricco:matrix.org | well they definitely match | 04:29:16 |
