[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[bug#73654] [PATCH v4] mapped-devices: luks: Support passing --allow-dis
From: |
Maxim Cournoyer |
Subject: |
[bug#73654] [PATCH v4] mapped-devices: luks: Support passing --allow-discards during open |
Date: |
Thu, 20 Mar 2025 11:54:28 +0900 |
User-agent: |
Gnus/5.13 (Gnus v5.13) |
tag 73654 + moreinfo
quit
Hi!
soeren@soeren-tempel.net writes:
> From: Sören Tempel <soeren@soeren-tempel.net>
>
> * gnu/system/mapped-devices.scm (open-luks-device): Support opening
> LUKS devices with the --allow-discards option.
> * gnu/system/mapped-devices.scm (luks-device-mapping-with-options):
> Pass through the allow-discards? keyword argument.
> * doc/guix.texi (Mapped Devices): Update documentation for the
> luks-device-mapping-with-options procedure.
>
> Co-authored-by: Sisiutl <sisiutl@egregore.fun>
I was about to apply it with the following cosmetic changes (mostly to
meet the 80 max column width):
--8<---------------cut here---------------start------------->8---
> ---
> Changes since v3: Fix replacement of “Solid State Disks” with “solid
> state disks” in doc/guix.texi. That is, only perform this replacement
> locally on the added text and not the whole document.
>
> doc/guix.texi | 11 +++++++++-
> gnu/system/mapped-devices.scm | 39 +++++++++++++++++++++--------------
> 2 files changed, 33 insertions(+), 17 deletions(-)
>
> diff --git a/doc/guix.texi b/doc/guix.texi
> index b1b6d98e74..6eb9fcb8ee 100644
> --- a/doc/guix.texi
> +++ b/doc/guix.texi
> @@ -18402,7 +18402,7 @@ Mapped Devices
> @code{dm-crypt} Linux kernel module.
> @end defvar
>
> -@deffn {Procedure} luks-device-mapping-with-options [#:key-file]
> +@deffn {Procedure} luks-device-mapping-with-options [#:key-file
> #:allow-discards?]
> Return a @code{luks-device-mapping} object, which defines LUKS block
> device encryption using the @command{cryptsetup} command from the
> package with the same name. It relies on the @code{dm-crypt} Linux
> @@ -18424,6 +18424,15 @@ Mapped Devices
> (type (luks-device-mapping-with-options
> #:key-file "/crypto.key")))
> @end lisp
> +
> +
> +@code{allow-discards?} allows the use of discard (TRIM) requests for the
> +underlying device. This is useful for solid state drives. However,
> +this option can have a negative security impact because it can make
> +file system level operations visible on the physical device. For more
> +information, refer to the description of the @code{--allow-discards}
> +option in the @code{cryptsetup-open(8)} man page.
> +
> @end deffn
>
> @defvar raid-device-mapping
> diff --git a/gnu/system/mapped-devices.scm b/gnu/system/mapped-devices.scm
> index 931c371425..3a8f0d66fe 100644
> --- a/gnu/system/mapped-devices.scm
> +++ b/gnu/system/mapped-devices.scm
> @@ -194,9 +194,10 @@ (define (check-device-initrd-modules device
> linux-modules location)
> ;;; Common device mappings.
> ;;;
>
> -(define* (open-luks-device source targets #:key key-file)
> +(define* (open-luks-device source targets #:key key-file allow-discards?)
> "Return a gexp that maps SOURCE to TARGET as a LUKS device, using
> -'cryptsetup'."
> +'cryptsetup'. When ALLOW-DISCARDS? is true, the use of discard (TRIM)
> requests is
> +allowed for the underlying device."
> (with-imported-modules (source-module-closure
> '((gnu build file-systems)
> (guix build utils))) ;; For mkdir-p
> @@ -234,17 +235,21 @@ (define* (open-luks-device source targets #:key
> key-file)
> (loop (- tries-left 1))))))
> (error "LUKS partition not found" source))
> source)))
> - ;; We want to fallback to the password unlock if the keyfile
> fails.
> - (or (and keyfile
> - (zero? (system*/tty
> - #$(file-append cryptsetup-static
> "/sbin/cryptsetup")
> - "open" "--type" "luks"
> - "--key-file" keyfile
> - partition #$target)))
> - (zero? (system*/tty
> - #$(file-append cryptsetup-static "/sbin/cryptsetup")
> - "open" "--type" "luks"
> - partition #$target)))))))))
> + (let ((cryptsetup-flags (cons*
> + "open" "--type" "luks" partition
> #$target
> + (if allow-discards?
> + '("--allow-discards")
> + '()))))
> + ;; We want to fallback to the password unlock if the keyfile
> fails.
> + (or (and keyfile
> + (zero?
> + (apply system*/tty
> + #$(file-append cryptsetup-static
> "/sbin/cryptsetup")
> + "--key-file" keyfile
> + cryptsetup-flags)))
> + (zero? (apply system*/tty
> + #$(file-append cryptsetup-static
> "/sbin/cryptsetup")
> + cryptsetup-flags))))))))))
>
> (define (close-luks-device source targets)
> "Return a gexp that closes TARGET, a LUKS device."
> @@ -286,13 +291,15 @@ (define luks-device-mapping
> ((gnu build file-systems)
> #:select (find-partition-by-luks-uuid system*/tty))))))
>
> -(define* (luks-device-mapping-with-options #:key key-file)
> +(define* (luks-device-mapping-with-options #:key key-file allow-discards?)
> "Return a luks-device-mapping object with open modified to pass the
> arguments
> into the open-luks-device procedure."
> (mapped-device-kind
> (inherit luks-device-mapping)
> - (open (λ (source targets) (open-luks-device source targets
> - #:key-file key-file)))))
> + (open (λ (source targets)
> + (open-luks-device source targets
> + #:key-file key-file
> + #:allow-discards? allow-discards?)))))
>
> (define (open-raid-device sources targets)
> "Return a gexp that assembles SOURCES (a list of devices) to the RAID
> device
>
> base-commit: f2b3c36bee8c232b026a66de93db38e13fbd7076
--8<---------------cut here---------------end--------------->8---
But unfortunately it appears to hang at least the 'encrypted-root-os'
system test, which you can run like:
--8<---------------cut here---------------start------------->8---
$ make check-system TESTS=encrypted-root-os
[...]
cSeaBIOS (version 1.16.2/GNU Guix)
iPXE (https://ipxe.org) 00:03.0 CA00 PCI2.10 PnP PMM+0EFCB030+0EF0B030 CA00
Booting from Hard Disk...
GRUB loading..
Welcome to GRUB!
Enter passphrase for hd0,gpt2 (12345678-1234-1234-1234-123456789abc):
Attempting to decrypt master key...
lot 0 opened
C-c C-cmake: *** [Makefile:7562: check-system] Interrompre
--8<---------------cut here---------------end--------------->8---
Would you have an idea of why this happens and how we could avoid the
hang in the test?
Thanks,
--
Maxim