Conversation
|
@GrahamcOfBorg build cryptsetup |
|
Success on x86_64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
|
No attempt on x86_64-darwin (full log) The following builds were skipped because they don't evaluate on x86_64-darwin: cryptsetup Partial log (click to expand)
|
|
Success on aarch64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
GrahamcOfBorg
added
10.rebuild-darwin: 0
|
No attempt on x86_64-darwin (full log) The following builds were skipped because they don't evaluate on x86_64-darwin: cryptsetup Partial log (click to expand)
|
|
Success on aarch64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
|
Success on x86_64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
|
I recall there were some patches needed for compatibility but not sure what the package was. |
|
Oh, right that was https://pagure.io/volume_key/c/ecef526a51c5a276681472fd6df239570c9ce518?branch=master and I am already applying it in #35551 |
|
Hi! I tried to use this patch to boot a nixos installed in a a LUKS2 encrypted device, but couldn't make it work. Appearently when running cryptsetup 2 from the initrd image to mount LUKS2 devices doesn't work. I'm not sure if this should be discused here, so ping me if you want more information about the bug. |
|
I only use this for LUKS encrypted offline storage on some thumb drives, and this has worked for me for the past 2 weeks perfectly. |
supersedes NixOS#35551 closes NixOS#34999 /cc NixOS#38382
|
@sh4r3m4n Please tell more about the bug. |
|
@yegortimoshenko: I had to modify the cryptsetup derivation in order to boot from a LUKS2 partition. I had to make two important changes to the pull request patch
After this two fixes I was able to boot with LVM inside a LUKS 2 device Here is my custom version of the cryptsetup derivation: ({ stdenv, fetchurl, devicemapper, json_c, openssl, libuuid, pkgconfig, popt
, enablePython ? false, python2 ? null, ...
}:
assert enablePython -> python2 != null;
stdenv.mkDerivation rec {
name = "cryptsetup-2.0.2";
NIX_LDFLAGS = "-lgcc_s";
src = fetchurl {
url = "mirror://kernel/linux/utils/cryptsetup/v2.0/${name}.tar.xz";
sha256 = "15wyjfgcqjf0wy5gxnmjj8aah33csv5v6n1hv9c8sxdzygbhb0ag";
};
configureFlags = [ "--enable-cryptsetup-reencrypt" "--with-crypto_backend=openssl" "--disable-kernel_crypto"]
++ stdenv.lib.optional enablePython "--enable-python";
nativeBuildInputs = [ pkgconfig ];
buildInputs = [ devicemapper json_c openssl libuuid popt ]
++ stdenv.lib.optional enablePython python2;
meta = {
homepage = https://gitlab.com/cryptsetup/cryptsetup/;
description = "LUKS for dm-crypt";
license = stdenv.lib.licenses.gpl2;
maintainers = with stdenv.lib.maintainers; [ viric chaoflow ];
platforms = with stdenv.lib.platforms; linux;
};
}) (import <nixpkgs> {}) |
|
@sh4r3m4n Thank you a lot! Hopefully will be able to commit updated cryptsetup to the tree this week. |
|
No attempt on x86_64-darwin (full log) The following builds were skipped because they don't evaluate on x86_64-darwin: cryptsetup Partial log (click to expand)
|
|
Success on x86_64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
|
Success on aarch64-linux (full log) Attempted: cryptsetup Partial log (click to expand)
|
| NIX_LDFLAGS = "-lgcc_s"; | ||
|
|
||
| configureFlags = [ | ||
| "--disable-kernel_crypto" |
There was a problem hiding this comment.
Why disable kernel crypto? I tried and cryptsetup builds fine without this flag.
There was a problem hiding this comment.
Because it doesn't seem to be available in initrd image, see #38382 (comment).
There was a problem hiding this comment.
I see. Is this setting likely to hurt performance of the encrypted disks? If yes, it might be better to add the missing module(s) to initrd at some point.
There was a problem hiding this comment.
It likely will hurt performance. I agree that adding missing modules to initrd would be preferable.
There was a problem hiding this comment.
I would suggest to merge this now anyway, observe the effects and optimize later if needed. What do you think?
There was a problem hiding this comment.
This is long overdue, so yes. I've opened #42163 to track this issue.
|
This change broke |
7 participants
build-use-sandboxinnix.confon non-NixOS)nix-shell -p nox --run "nox-review wip"./result/bin/)