Category Archives: Linux

[$] Hash table memory usage and a BPF interpreter bug

Posted on by

Anton Protopopov led a short discussion at the 2025 Linux Storage, Filesystem,
Memory-Management, and BPF Summit about amount of memory used
by hash tables in BPF programs. He thinks that the current memory layout is
inefficient, and wants to split the structure that holds table entries into two
variants for different kinds of maps. When that proposal proved
uncontroversial, he also took the chance to talk about a bug in BPF’s call
instruction.

[$] Debian’s AWKward essential set

Posted on by

The Debian project has the concept of essential
packages, which provide the bare minimum functionality considered
absolutely necessary (or “essential”) for a system to
function. Packages tagged as essential, and the packages that are
required by the set of essential packages, are always installed as
part of a Debian system. However, Debian’s packaging rules do not
require developers to explicitly declare dependencies on that set of
packages (the essential set) but they can simply rely on the fact that those
will always be present. That means that changing the essential set, as
the project may wish to do occasionally, is more complicated than it
should be. This came to light recently when a Debian developer asked
what might be required to remove mawk to slim down
the project’s container images.

Deepin Desktop removed from openSUSE

Posted on by

The SUSE Security Team has announced the removal of the Deepin
Desktop from openSUSE due to violations of the project’s packaging
policy.

The discovery of the bypass of the security whitelistings via the
deepin-feature-enable package marks a turning point in our assessment
of Deepin. We don’t believe that the openSUSE Deepin packager acted
with bad intent when he implemented the “license agreement” dialog to
bypass our whitelisting restrictions. The dialog itself makes the
security concerns we have transparent, so this does not happen in a
sneaky way, at least not towards users. It was not discussed with us,
however, and it violates openSUSE packaging policies. Beyond the
security aspect, this also affects general packaging quality
assurance: the D-Bus configuration files and Polkit policies installed
by the deepin-feature-enable package are unknown to the package
manager and won’t be cleaned up upon package removal, for
example. Such bypasses are not deemed acceptable by us.

The combination of these factors led us to the decision to remove
the Deepin desktop completely from openSUSE Tumbleweed and from the
future Leap 16.0 release. In openSUSE Leap 15.6 we will remove the
offending deepin-feature-enable package only. It is a difficult
decision given that the Deepin desktop has a considerable number of
users. We firmly believe the Deepin packaging and security assessment
in openSUSE needs a reboot, however, ideally involving new people that
can help get the Deepin packages into shape, establish a relationship
with Deepin upstream and keep an eye on bugfixes, thus avoiding
fruitless follow-up reviews that just waste our time. In such a new
setup we would be willing to have a look at all the sensitive Deepin
components again one by one.

The announcement goes into detail about the bypass of
openSUSE packaging policy and the history of security reviews of
Deepin components. It also offers guidance on continuing
to use Deepin Desktop on openSUSE.

AMD Strix Point & Intel Lunar Lake: Windows 11 vs. Ubuntu 25.04 Linux Performance

Posted on by

Last month was a fresh look at the Intel Lunar Lake graphics performance between Windows and Linux while this article is Microsoft Windows 11 Pro vs. Ubuntu 25.04 again but looking at the CPU performance between these competing operating systems. For additional reference, some of the recently completed AMD Strix Point Windows vs. Linux benchmarks were also included for additional insight.

Intel P-State Energy Aware Scheduling May Be Ready For Lunar Lake

Posted on by

Going back to last year there have been patches for adapting the Intel P-State Linux driver with support for Energy Aware Scheduling (EAS) as what began as an Arm big.LITTLE feature but since being explored for use by the Intel P-State driver for hybrid CPUs without SMT which for the moment means Lunar Lake SoCs. A new version of the patches were posted on Tuesday and likely to be the last refinement to this patch series…