util-linux.git
15 months agolscpu: optimize query virt pci device
Guixin Liu [Wed, 21 Aug 2024 08:49:54 +0000 (16:49 +0800)]
lscpu: optimize query virt pci device

The lscpu command needs to traverse the /proc/bus/pci/devices file
three times to check for any PCI devices related to virtualization.
If there are many PCI devices on the machine, this can lead to
increased execution time for lscpu. It would be beneficial to
consolidate these queries into a single check to optimize the
execution time of lscpu.

Use time command to test on a machine with 13000 PCI devices:
before:
real  0m13.506s
user  0m0.028s
sys   0m13.409s
after:
real  0m5.834s
user  0m0.018s
sys   0m5.771s

Signed-off-by: Guixin Liu <kanie@linux.alibaba.com>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name lscpu-optimize-query-virt-pci-device.patch

15 months agomore: make sure we have data on stderr
Karel Zak [Thu, 22 Aug 2024 06:56:52 +0000 (08:56 +0200)]
more: make sure we have data on stderr

more(1) uses more_poll() to monitor data on stdin, stderr, and
signals. It is used before read_command(), but this function only
reads from stderr. Therefore, if any other non-stderr event occurs,
this function will wait on read(). In this case, more(1) will not
react to signals anymore. We need to ensure that more(1) only waits in
more_poll().

Try

 for x in {1..1000}; do echo "line $x"; done | more

to reproduce.

Reported-by: Radka Skvarilova <rskvaril@redhat.com>
Signed-off-by: Karel Zak <kzak@redhat.com>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name more-make-sure-we-have-data-on-stderr.patch

15 months agoTreat out of memory as error
Tobias Stoeckmann [Mon, 12 Aug 2024 15:37:03 +0000 (17:37 +0200)]
Treat out of memory as error

Otherwise it might happen that a configured file location is not taken
into account. A NULL value would lead to default file location.

Gbp-Pq: Topic upstream-master
Gbp-Pq: Name Treat-out-of-memory-as-error.patch

15 months agoagetty: add "systemd" to --version output
Karel Zak [Thu, 11 Jul 2024 11:26:30 +0000 (13:26 +0200)]
agetty: add "systemd" to --version output

Signed-off-by: Karel Zak <kzak@redhat.com>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name agetty-add-systemd-to-version-output.patch

15 months agounshare: load binfmt_misc interpreter
Laurent Vivier [Wed, 19 Jun 2024 08:19:51 +0000 (10:19 +0200)]
unshare: load binfmt_misc interpreter

add -l, --load-interp <file> to load a binfmt_misc interpreter at startup.

The interpreter is loaded from the initial fileystem if the 'F' flags is
provided, otherwise from inside the new namespace
This makes possible to start a chroot of another architecture without
being root.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name unshare-load-binfmt_misc-interpreter.patch

15 months agounshare: mount binfmt_misc
Laurent Vivier [Wed, 19 Jun 2024 08:19:50 +0000 (10:19 +0200)]
unshare: mount binfmt_misc

add --mount-binfmt[=<dir>] to mount binfmt_misc filesystem,
this results in clearing inherited interpreters from the previous namespace

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name unshare-mount-binfmt_misc.patch

15 months agoswapoff: avoid being killed by OOM
Karel Zak [Thu, 20 Jun 2024 10:07:12 +0000 (12:07 +0200)]
swapoff: avoid being killed by OOM

Based on patch from dparalen <vetrisko@gmail.com>.

Fixes: https://github.com/util-linux/util-linux/issues/3095
Signed-off-by: Karel Zak <kzak@redhat.com>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name swapoff-avoid-being-killed-by-OOM.patch

15 months agolibmount: Fix atime remount for new API
Karel Zak [Tue, 14 May 2024 09:58:20 +0000 (11:58 +0200)]
libmount: Fix atime remount for new API

All atime settings are mutually exclusive, and the attr_set mask for
the mount_setattr() syscall cannot contain multiple MOUNT_ATTR_ atime
related options.

Unfortunately, during a remount, the list of options is composed of
both old and new options. In this case, libmount sets more atime
options to the mask. The correct behavior is to use the last atime
related option from the list.

Fixes: https://github.com/util-linux/util-linux/issues/3032
Signed-off-by: Karel Zak <kzak@redhat.com>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name libmount-Fix-atime-remount-for-new-API.patch

15 months agofsck: warn if fsck.<type> not found and device is specified
Karel Zak [Thu, 8 Feb 2024 11:42:51 +0000 (12:42 +0100)]
fsck: warn if fsck.<type> not found and device is specified

The fsck(8) command prints an error for certain 'required'
filesystems, such as extN, if the corresponding fsck.<type> helper is
not found. However, for other filesystems, it silently ignores the
device. While this behavior is acceptable for the -A option, if a user
explicitly specifies the device on the command line, it is probably a
good idea to be more verbose and issue a warning that the device will
be ignored. The warning is enabled only for interactive mode (only one
device specified or -s).

Note that some filesystems do not have an fsck utility, so caution is
needed when issuing such warnings to prevent the generation of a large
number of unwanted entries in system logs, etc."

Reported-by: наб <nabijaczleweli@nabijaczleweli.xyz>
Signed-off-by: Karel Zak <kzak@redhat.com>
Closes: #1062208
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name fsck-warn-if-fsck.-type-not-found-and-device-is-specified.patch

15 months agoaudit-arch.h: add defines for m68k, sh
Chris Hofstaedtler [Tue, 9 Apr 2024 08:30:32 +0000 (10:30 +0200)]
audit-arch.h: add defines for m68k, sh

John Paul Adrian Glaubitz points out that libseccomp needs to release 2.6.0,
too.

Upstream commit 5ebf0edb0a0531cba801c7791543c610725bf944.

Signed-off-by: Chris Hofstaedtler <zeha@debian.org>
Gbp-Pq: Topic upstream-master
Gbp-Pq: Name audit-arch.h-add-defines-for-m68k-sh.patch

15 months agologin: turn off btmp, utmp, lastlog writing
Chris Hofstaedtler [Mon, 5 Aug 2024 14:03:11 +0000 (16:03 +0200)]
login: turn off btmp, utmp, lastlog writing

Should be done by PAM instead. login from src:shadow had turned it off, especially with --enable-logind.

Gbp-Pq: Topic debian
Gbp-Pq: Name login-turn-off-btmp-utmp-lastlog-writing.patch

15 months agologin: do not print motd, done by PAM
Chris Hofstaedtler [Mon, 5 Aug 2024 13:54:17 +0000 (15:54 +0200)]
login: do not print motd, done by PAM

Gbp-Pq: Topic debian
Gbp-Pq: Name login-do-not-print-motd-done-by-PAM.patch

15 months agologin: support LOG_OK_LOGINS
Chris Hofstaedtler [Mon, 5 Aug 2024 13:53:44 +0000 (15:53 +0200)]
login: support LOG_OK_LOGINS

Gbp-Pq: Topic debian
Gbp-Pq: Name login-support-LOG_OK_LOGINS.patch

15 months agotests: mark fadvise/drop as known failing on buildds
Chris Hofstaedtler [Tue, 2 Apr 2024 19:52:47 +0000 (21:52 +0200)]
tests: mark fadvise/drop as known failing on buildds

buildds build on a tmpfs, but this is not visible inside the build environment.

Gbp-Pq: Topic debian
Gbp-Pq: Name tests-mark-fadvise-drop-as-known-failing-on-buildds.patch

15 months agotests: mark lsfd mkfds-socketpair as failing in sbuild
Chris Hofstaedtler [Thu, 29 Feb 2024 22:22:41 +0000 (23:22 +0100)]
tests: mark lsfd mkfds-socketpair as failing in sbuild

Gbp-Pq: Topic debian
Gbp-Pq: Name tests-mark-lsfd-tests-failing-in-sbuild.patch

15 months agoUse sensible-pager
Chris Hofstaedtler [Thu, 14 Jul 2022 09:54:01 +0000 (09:54 +0000)]
Use sensible-pager

Closes: #1014368
Gbp-Pq: Topic debian
Gbp-Pq: Name sensible-pager.patch

15 months agoInstall lsfd into /usr/bin
Chris Hofstaedtler [Thu, 14 Apr 2022 11:47:20 +0000 (11:47 +0000)]
Install lsfd into /usr/bin

Gbp-Pq: Topic debian
Gbp-Pq: Name lsfd-usrbin.patch

15 months agohardlink tests: set known failed
Chris Hofstaedtler [Tue, 12 Apr 2022 20:35:57 +0000 (20:35 +0000)]
hardlink tests: set known failed

sha256 cannot be initialized on some archs / buildds.

Gbp-Pq: Topic debian
Gbp-Pq: Name hardlink-tests-known-failed.patch

15 months agoverbose-tests
Andreas Henriksson [Wed, 30 Jul 2014 12:28:17 +0000 (14:28 +0200)]
verbose-tests

Enable verbose tests and print content of failing tests output file.

Gbp-Pq: Topic debian
Gbp-Pq: Name verbose-tests.patch

15 months agoutil-linux: getopt(1) should point to where examples are on a Debian system
Matthew Vernon [Tue, 6 Nov 2018 11:49:21 +0000 (11:49 +0000)]
util-linux: getopt(1) should point to where examples are on a Debian system

Debian BTS #913049

Gbp-Pq: Topic debian
Gbp-Pq: Name man-getopt-examples.patch

15 months agoutil-linux (2.40.2-8) unstable; urgency=medium
Chris Hofstaedtler [Mon, 9 Sep 2024 21:20:25 +0000 (23:20 +0200)]
util-linux (2.40.2-8) unstable; urgency=medium

  * Mark nologin, newgrp binaries and man pages linux-any
  * Install manpages using dh_install
  * Import patches from upstream master
    * newgrp: Optionally execute a program after group change
    * Treat out of memory as error
    * agetty: add "systemd" to --version output
    * lscpu: optimize query virt pci device
    * more: make sure we have data on stderr
    Thanks to Gábor Németh <homar@riseup.net> (Closes: #1078122)
  * Revert "Remove sg symlink to newgrp"
  * Install sg symlink only on Linux
  * Install write(1) manpage
  * salsa-ci: add nodoc profile

[dgit import unpatched util-linux 2.40.2-8]

15 months agoImport util-linux_2.40.2-8.debian.tar.xz
Chris Hofstaedtler [Mon, 9 Sep 2024 21:20:25 +0000 (23:20 +0200)]
Import util-linux_2.40.2-8.debian.tar.xz

[dgit import tarball util-linux 2.40.2-8 util-linux_2.40.2-8.debian.tar.xz]

17 months agoImport util-linux_2.40.2.orig.tar.xz
Chris Hofstaedtler [Sat, 6 Jul 2024 13:57:30 +0000 (15:57 +0200)]
Import util-linux_2.40.2.orig.tar.xz

[dgit import orig util-linux_2.40.2.orig.tar.xz]