linux (5.10.46-4) unstable; urgency=medium

  * From Linux 5.10.46-4, unprivileged calls to bpf() are disabled by
    default, mitigating several security issues. However, an admin can
    still change this setting later on, if needed, by writing 0 or 1 to
    the kernel.unprivileged_bpf_disabled sysctl.

    If you prefer to keep unprivileged calls to bpf() enabled, set the
    sysctl:

    kernel.unprivileged_bpf_disabled = 0

    which is the upstream default.

 -- Salvatore Bonaccorso <carnil@debian.org>  Mon, 02 Aug 2021 22:59:24 +0200

linux (5.10~rc7-1~exp2) unstable; urgency=medium

  * From Linux 5.10, all users are allowed to create user namespaces by
    default.  This will allow programs such as web browsers and container
    managers to create more restricted sandboxes for untrusted or
    less-trusted code, without the need to run as root or to use a
    setuid-root helper.

    The previous Debian default was to restrict this feature to processes
    running as root, because it exposed more security issues in the
    kernel.  However, the security benefits of more widespread sandboxing
    probably now outweigh this risk.

    If you prefer to keep this feature restricted, set the sysctl:

        kernel.unprivileged_userns_clone = 0

 -- Ben Hutchings <benh@debian.org>  Sun, 13 Dec 2020 17:11:36 +0100

linux-latest (86) unstable; urgency=medium

  * From Linux 4.13.10-1, AppArmor is enabled by default.  This allows
    defining a "profile" for each installed program that can mitigate
    security vulnerabilities in it.  However, an incorrect profile might
    disable some functionality of the program.

    In case you suspect that an AppArmor profile is incorrect, see
    <https://lists.debian.org/debian-devel/2017/11/msg00178.html> and
    consider reporting a bug in the package providing the profile.  The
    profile may be part of the program's package or apparmor-profiles.

 -- Ben Hutchings <ben@decadent.org.uk>  Thu, 30 Nov 2017 20:08:25 +0000

linux-latest (81) unstable; urgency=medium

  * From Linux 4.10, the old 'virtual syscall' interface on 64-bit PCs
    (amd64) is disabled.  This breaks chroot environments and containers
    that use (e)glibc 2.13 and earlier, including those based on Debian 7
    or RHEL/CentOS 6.  To re-enable it, set the kernel parameter:
    vsyscall=emulate

 -- Ben Hutchings <ben@decadent.org.uk>  Fri, 30 Jun 2017 23:50:03 +0100

linux-latest (76) unstable; urgency=medium

  * From Linux 4.8, several changes have been made in the kernel
    configuration to 'harden' the system, i.e. to mitigate security bugs.
    Some changes may cause legitimate applications to fail, and can be
    reverted by run-time configuration:
    - On most architectures, the /dev/mem device can no longer be used to
      access devices that also have a kernel driver.  This breaks dosemu
      and some old user-space graphics drivers.  To allow this, set the
      kernel parameter: iomem=relaxed
    - The kernel log is no longer readable by unprivileged users.  To
      allow this, set the sysctl: kernel.dmesg_restrict=0

 -- Ben Hutchings <ben@decadent.org.uk>  Sat, 29 Oct 2016 02:05:32 +0100

linux-latest (75) unstable; urgency=medium

  * From Linux 4.7, the iptables connection tracking system will no longer
    automatically load helper modules.  If your firewall configuration
    depends on connection tracking helpers, you should explicitly load the
    required modules.  For more information, see
    <https://home.regit.org/netfilter-en/secure-use-of-helpers/>.

 -- Ben Hutchings <ben@decadent.org.uk>  Sat, 29 Oct 2016 01:53:18 +0100