Starting with this up-to-date release we can finally rely on
properly versioned falcosecurity-libs, support more platforms
and eventually add support for the eBPF probe, obviating the
need for the classic scap kernel module.
Note that this release has the "modern eBPF" probe disabled
until the necessary dependencies (llvm etc.) have been sorted
out and tested.
Closes: https://bugs.gentoo.org/934889
Closes: https://github.com/gentoo/gentoo/pull/37446
Signed-off-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: Yixun Lan <dlan@gentoo.org>
This adds the latest & greatest version of the scap kernel driver.
Since we're now finally able to use proper version numbers we no
longer need to rely on random git commits.
Unfortunately this means we have to travel back in version-time;
this should not pose a problem as sysdig-0.38.x is pinned to this
version and will automatically do the right thing.
Signed-off-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: Yixun Lan <dlan@gentoo.org>
- bring back the REQUIRED_USE for "wayland? ( dbus )". Again, getting
Firefox's dependencies installed without dbus (especially when +wayland is
specified) requires custom patching to GTK and its deps, and by that nature
is very unsupported already. I can't test on a dbusless system, but am
willing to include such a patch when provided (see bgo#860033).
Closes: https://bugs.gentoo.org/936583
Signed-off-by: Joonas Niilola <juippis@gentoo.org>
Attempted to patch it like usually do, but this time around some
methods were replaced by alternatives that I'm not sure sip even
supports and have no idea how to patch it properly (esp. since
hardly know sip and C++ as-is). Issues are in coreand USE=testlib,
the rest appears fine.
Not much can do other than wait for PyQt6-6.8.0 to release
(similarly to what pyside6 does) unless someone else can provide
a working patch. If there is a PyQt6 pre-release snapshot will
likely use it once available, but these haven't been happening
*before* Qt releases lately.
Technically no need for a revbump thanks to the binding operator,
but, if we do not do it, portage backtracks a lot more and gives
confusing messages when the whole Qt stack is involved. So want
to propagate the upper bound on user's systems now *long* before
the bump (may be wasted if we get a snapshot in time, but that
seems unlikely). Will spare stable users given we can just stabilize
6.7.1-r1 to cause a rebuild and odds are we will not stable 6.8.0
either way (may wait for 6.8.1 + PyQt6-6.8.x).
Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
This new eclass includes various improvements over the existing
readme.gentoo-r1.eclass.
First, the content of README.gentoo will only be shown on new
installations or if it has changed between updates.
Second, it allows the composition of readme via bash's heredoc.
And finally, the eclass exports phase functions, which helps to reduce
the boilerplate code in many cases.
Signed-off-by: Florian Schmaus <flow@gentoo.org>
- mask USE http3, quic globally
- unmask on arches with support
We will switch to a per-arch mask down the line as keyword requests
are fulfilled.
Signed-off-by: Matt Jolly <kangie@gentoo.org>
Michał Górny
Holger Hoffstätte
Sam James
Alfred Wingate
Volkmar W. Pogatzki
Jakov Smolić
Joonas Niilola
Ionen Wolkens
Florian Schmaus
Ulrich Müller
Andrew Ammerlaan
Petr Vaněk
Agostino Sarubbo
Matt Jolly
Matoro Mahri
Eray Aslan