| 1 |
linux (6.18~rc4-1~exp1) experimental; urgency=medium |
| 2 |
|
| 3 |
The XFS V4 filesystem format is no longer supported. The V5 format, |
| 4 |
supported since Linux 3.15, should be used instead. You MUST convert |
| 5 |
any V4 filesystems before running this kernel version. |
| 6 |
|
| 7 |
To check which format your current XFS filesystems use, run this command |
| 8 |
as root: |
| 9 |
|
| 10 |
journalctl -k -b 0 --grep '^XFS .* Mounting V.' |
| 11 |
|
| 12 |
Unfortunately there is no support for in-place conversion of a V4 |
| 13 |
filesystem to V5. You will need to migrate the files onto newly created |
| 14 |
V5 filesystems. |
| 15 |
|
| 16 |
-- Ben Hutchings <[email protected]> Sun, 26 Apr 2026 22:47:35 +0200 |
| 17 |
|
| 18 |
linux (6.17.2-1~exp1) experimental; urgency=medium |
| 19 |
|
| 20 |
* The bcachefs filesystem has been disabled in Debian kernel packages. This |
| 21 |
filesystem is no longer being developed in the mainline Linux kernel, and |
| 22 |
the bcachefs-tools package is also not available in Debian. |
| 23 |
|
| 24 |
We recommend switching to a different filesystem if you are using bcachefs. |
| 25 |
|
| 26 |
Alternatively you can use the upstream provided bcachefs dkms packages |
| 27 |
available at https://apt.bcachefs.org/ . |
| 28 |
|
| 29 |
-- Salvatore Bonaccorso <[email protected]> Tue, 09 Sep 2025 20:41:27 +0200 |
| 30 |
|
| 31 |
linux (5.10.46-4) unstable; urgency=medium |
| 32 |
|
| 33 |
* From Linux 5.10.46-4, unprivileged calls to bpf() are disabled by |
| 34 |
default, mitigating several security issues. However, an admin can |
| 35 |
still change this setting later on, if needed, by writing 0 or 1 to |
| 36 |
the kernel.unprivileged_bpf_disabled sysctl. |
| 37 |
|
| 38 |
If you prefer to keep unprivileged calls to bpf() enabled, set the |
| 39 |
sysctl: |
| 40 |
|
| 41 |
kernel.unprivileged_bpf_disabled = 0 |
| 42 |
|
| 43 |
which is the upstream default. |
| 44 |
|
| 45 |
-- Salvatore Bonaccorso <[email protected]> Mon, 02 Aug 2021 22:59:24 +0200 |
| 46 |
|
| 47 |
linux (5.10~rc7-1~exp2) unstable; urgency=medium |
| 48 |
|
| 49 |
* From Linux 5.10, all users are allowed to create user namespaces by |
| 50 |
default. This will allow programs such as web browsers and container |
| 51 |
managers to create more restricted sandboxes for untrusted or |
| 52 |
less-trusted code, without the need to run as root or to use a |
| 53 |
setuid-root helper. |
| 54 |
|
| 55 |
The previous Debian default was to restrict this feature to processes |
| 56 |
running as root, because it exposed more security issues in the |
| 57 |
kernel. However, the security benefits of more widespread sandboxing |
| 58 |
probably now outweigh this risk. |
| 59 |
|
| 60 |
If you prefer to keep this feature restricted, set the sysctl: |
| 61 |
|
| 62 |
kernel.unprivileged_userns_clone = 0 |
| 63 |
|
| 64 |
-- Ben Hutchings <[email protected]> Sun, 13 Dec 2020 17:11:36 +0100 |
| 65 |
|
| 66 |
linux-latest (86) unstable; urgency=medium |
| 67 |
|
| 68 |
* From Linux 4.13.10-1, AppArmor is enabled by default. This allows |
| 69 |
defining a "profile" for each installed program that can mitigate |
| 70 |
security vulnerabilities in it. However, an incorrect profile might |
| 71 |
disable some functionality of the program. |
| 72 |
|
| 73 |
In case you suspect that an AppArmor profile is incorrect, see |
| 74 |
<https://lists.debian.org/debian-devel/2017/11/msg00178.html> and |
| 75 |
consider reporting a bug in the package providing the profile. The |
| 76 |
profile may be part of the program's package or apparmor-profiles. |
| 77 |
|
| 78 |
-- Ben Hutchings <[email protected]> Thu, 30 Nov 2017 20:08:25 +0000 |
| 79 |
|
| 80 |
linux-latest (81) unstable; urgency=medium |
| 81 |
|
| 82 |
* From Linux 4.10, the old 'virtual syscall' interface on 64-bit PCs |
| 83 |
(amd64) is disabled. This breaks chroot environments and containers |
| 84 |
that use (e)glibc 2.13 and earlier, including those based on Debian 7 |
| 85 |
or RHEL/CentOS 6. To re-enable it, set the kernel parameter: |
| 86 |
vsyscall=emulate |
| 87 |
|
| 88 |
-- Ben Hutchings <[email protected]> Fri, 30 Jun 2017 23:50:03 +0100 |
| 89 |
|
| 90 |
linux-latest (76) unstable; urgency=medium |
| 91 |
|
| 92 |
* From Linux 4.8, several changes have been made in the kernel |
| 93 |
configuration to 'harden' the system, i.e. to mitigate security bugs. |
| 94 |
Some changes may cause legitimate applications to fail, and can be |
| 95 |
reverted by run-time configuration: |
| 96 |
- On most architectures, the /dev/mem device can no longer be used to |
| 97 |
access devices that also have a kernel driver. This breaks dosemu |
| 98 |
and some old user-space graphics drivers. To allow this, set the |
| 99 |
kernel parameter: iomem=relaxed |
| 100 |
- The kernel log is no longer readable by unprivileged users. To |
| 101 |
allow this, set the sysctl: kernel.dmesg_restrict=0 |
| 102 |
|
| 103 |
-- Ben Hutchings <[email protected]> Sat, 29 Oct 2016 02:05:32 +0100 |
| 104 |
|
| 105 |
linux-latest (75) unstable; urgency=medium |
| 106 |
|
| 107 |
* From Linux 4.7, the iptables connection tracking system will no longer |
| 108 |
automatically load helper modules. If your firewall configuration |
| 109 |
depends on connection tracking helpers, you should explicitly load the |
| 110 |
required modules. For more information, see |
| 111 |
<https://home.regit.org/netfilter-en/secure-use-of-helpers/>. |
| 112 |
|
| 113 |
-- Ben Hutchings <[email protected]> Sat, 29 Oct 2016 01:53:18 +0100 |