Native support was introduced with commit 13c6be96618c, QEMU 7.2.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Even libvirt itself will configure passt to write log, PID and socket
files to different locations depending on whether the domain is
started as root (/var/log/libvirt/...) or as a regular user
(/var/log/<PID>/libvirt/...), and user_tmp_t would only cover the
latter.
Create interfaces for log and PID files, so that callers can specify
different file contexts for those, and modify the interface for the
UNIX socket file to allow different paths as well.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Tested-by: Laine Stump <laine@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Laine reports that with a simple:
<portForward proto='tcp'>
<range start='2022' to='22'/>
</portForward>
in libvirt's domain XML, passt won't start as it fails to bind
arbitrary ports. That was actually the intention behind passt_port_t:
the user or system administrator should have explicitly configured
allowed ports on a given machine. But it's probably not realistic, so
just allow any port to be bound and forwarded.
Also fix up some missing operations on sockets.
Reported-by: Laine Stump <laine@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Tested-by: Laine Stump <laine@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Otherwise, it's unusable as stand-alone tool, or in foreground mode,
and it's also impossible to get output from --help or --version,
because for SELinux it's just a daemon.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Tested-by: Laine Stump <laine@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
One day, libvirt might actually support running passt to provide
guest connectivity. Should libvirtd (or virtqemud) start passt, it
will need to access socket and PID files in specific locations, and
passt needs to accept SIGTERM in case QEMU fails to start after passt
is already started.
To make this more convenient, split the current profile into two
abstractions, for passt and for pasta, so that external programmes
can include the bits they need (and especially not include the pasta
abstraction if they only need to start passt), plus whatever specific
adaptation is needed.
For stand-alone usage of passt and pasta, the 'passt' profile simply
includes both abstractions, plus rules to create and access PID and
capture files in default or reasonable ($HOME) locations.
Tested on Debian with libvirt 9.0.0 together with a local fix to start
passt as intended, namely libvirt commit c0efdbdb9f66 ("qemu_passt:
Avoid double daemonizing passt"). This is an example of how the
libvirtd profile (or virtqemud abstraction, or virtqemud profile) can
use this:
# support for passt network back-end
/usr/bin/passt Cx -> passt,
profile passt {
/usr/bin/passt r,
owner @{run}/user/[0-9]*/libvirt/qemu/run/passt/* rw,
signal (receive) set=("term") peer=/usr/sbin/libvirtd,
signal (receive) set=("term") peer=libvirtd,
include if exists <abstractions/passt>
}
translated:
- when executing /usr/bin/passt, switch to the subprofile "passt"
(not the "discrete", i.e. stand-alone profile), described below.
Scrub the environment (e.g. LD_PRELOAD is dropped)
- in the "passt" subprofile:
- allow reading the binary
- allow read and write access to PID and socket files
- make passt accept SIGTERM from /usr/sbin/libvirtd, and
libvirtd peer names
- include anything else that's needed by passt itself
Suggested-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Instead of restricting PID files to /var/run/passt.pid, which is a
single file and unlikely to be used, use the user_tmp_t type which
should cover any reasonable need.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Services running passt will commonly need to transition to its
domain, terminate it, connect and write to its socket.
The init_daemon_domain() macro now defines the default transition to
the passt_t domain, using the passt_exec_t type.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is an example interface, currently unused, so it went undetected:
m4 macros need a backtick at the beginning of a block instead of a
single quote.
Fixes: 1f4b7fa0d7 ("passt, pasta: Add examples of SELinux policy modules")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The development of the Debian package is now at:
https://salsa.debian.org/sbrivio/passt
Drop contrib/debian, it's finally obsolete.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
AppArmor resolves executable links before profile attachment rules
are evaluated, so, as long as pasta is installed as a link to passt,
there's no way to differentiate the two cases. Merge the two profiles
and leave a TODO note behind, explaining two possible ways forward.
Update the rules so that passt and pasta are actually usable, once
the profile is installed. Most required changes are related to
isolation and sandboxing features.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
fedora-review says:
Note: Directories without known owners:
/usr/share/selinux/packages/passt, /usr/share/doc/passt,
/usr/share/selinux, /usr/share/selinux/packages
and selinux-policy owns those two last ones.
While at it, split Requires: tags also for post and preun actions
onto different lines, for consistency.
Reported-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
fedora-review says:
Note: No known owner of /usr/share/selinux/packages/passt,
/usr/share/doc/passt
While at it, replace "passt" by "%{name}" in a few places for
consistency.
Reported-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...instead of PATH. This seems to be the only change needed in
existing pasta integrations after patch:
Use explicit --netns option rather than multiplexing with PID
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Fedora's parameters currently match the ones from the Makefile (which
is based on GNU recommendations), but that's not necessarily
guaranteed.
This should make the OpenSUSE Tumbleweed override for docdir
unnecessary: drop it.
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...as it's used twice. The short version, however, appears hardcoded
only once in the output, and it comes straight from the rpkg macro
building the version string -- leave that macro as it is.
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...which makes it fall under MIT licensing terms. Daniel reports that
it's very unusual for spec files to contain explicit licensing terms
and might cause minor inconveniences later on, on mass changes to
spec files.
I originally added licensing information using SPDX identifiers to
make the project fully compliant with the REUSE Specification 3.0
(https://reuse.software/spec/), but there are anyway a few more files
not including explicit licensing information. It might be worth to
fix that later on, in any case.
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
It turns out that, while on most distributions "docdir" would be
/usr/share/doc, it's /usr/share/doc/packages/ on OpenSUSE Tumbleweed.
Use an explicit docdir as shown in:
https://en.opensuse.org/openSUSE:Build_Service_cross_distribution_howto
and don't unnecessarily hardcode directory variables in the Makefile.
Otherwise, RPM builds for OpenSUSE will fail now that we have a README
there.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
If the man pages are not compressed, the current wildcards wouldn't
match them. Drop the trailing '.' from them.
Reported-by: Artur Frenszek-Iwicki <fedora@svgames.pl>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This will also set any distribution-specific LDFLAGS. It's not needed
anymore starting from Fedora 36, but the package might be built on
other versions and distributions too (including e.g. CentOS Stream 8).
Reported-by: Artur Frenszek-Iwicki <fedora@svgames.pl>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Otherwise, passt-selinux will be built separately for each supported
architecture.
Suggested-by: Artur Frenszek-Iwicki <fedora@svgames.pl>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is required as Fedora doesn't accept a temporary pointer to
a source URL.
Reported-by: Ralf Corsepius <rc040203@freenet.de>
Reported-by: Artur Frenszek-Iwicki <fedora@svgames.pl>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
It seems to be exposed by Koji (https://pagure.io/koji/issue/2541),
but it's not actually in use, so we have to drop that. The website
the URL tag points to reports all the needed information anyway.
Reported-by: Artur Frenszek-Iwicki <fedora@svgames.pl>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
git_dir_changelog is useful in theory, but it requires pairs of
annotated tags, which should be generated by rpkg itself to make any
sense, implying a relatively heavyweight interaction whenever I want
to push a new package version.
Also, the default content of the changelog entries include the full
list of changes, but the Fedora Packaging Guidelines specifically
mention that:
[t]hey must never simply contain an entire copy of the source
CHANGELOG entries.
We don't have a CHANGELOG file, but the full git history is
conceptually equivalent for this purpose, I guess.
Introduce our own passt_git_changelog() rpkg macro, building
changelog entries, using tags in the form DATE-SHA, where DATE
is an ISO 8601 date representation, and SHA is a short (7-digits)
form of the head commit at a given moment (git push).
These changelog entries mention, specifically, changes to the
packaging information itself (entries under contrib/fedora), and
simply report a link to cgit for the ranges between tags.
Reported-by: Benson Muite <benson_muite@emailplus.org>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Given that a three-way git merge was enough to cope with context
changes in man pages, it's probably a good idea to enable that for
'git am' in the demo too.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
ip(8)'s ability to take abbreviated arguments (e.g. "li sh" instead of
"link show") is very handy when using it interactively, but it doesn't make
for very readable scripts and examples when shown that way.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Rebase the patch for Podman on top of current upstream, and:
- add support for configuration of specific addresses for forwarded
ports
- by default, disable port forwarding, and reflect this in the man
page changes
- adjust processing to a new, incompatible format for port storage,
which I couldn't actually track down to a specific commit, but
that resulted in https://github.com/containers/podman/issues/13643
and commit eedaaf33cdbf ("fix slirp4netns port forwarding with
ranges")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...using dh_apparmor to ship and apply AppArmor profiles. Tried on
current Debian testing (Bookworm, 12).
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The patch introduces a "pasta" networking mode for rootless
container, similar to the existing slirp4netns mode. Notable
differences are described in the commit message.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
passt can be used to implement user-mode networking for the Kata
Containers runtime, so that networking setup doesn't need elevated
privileges or capabilities.
This commit adds the patch for Kata Containers runtime and agent
to support passt as networking model and endpoint, and some basic
documentation.
See contrib/kata-containers/README.md for more details and setup
steps.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
I'm about to add a new adaptation carrying out-of-tree patches
for a Kata Containers PoC -- move the existing out-of-tree patches
to their own directory to keep things easy to find in the main one.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
