f39 doesn't container go 1.22 so we can no longer build there. epel 9
has the same issue although it is likely that go will be updated there
at one point.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Many dependencies started using go 1.22 which means we have to follow in
order to update.
Disable the now depracted exportloopref linter as it was replaced by
copyloopvar as go fixed the loop copy problem in 1.22[1]
Another new chnage in go 1.22 is the for loop syntax over ints, the
intrange linter chacks for this but there a lot of loops that have to be
converted so I didn't do it here and disable th elinter for now, th eold
syntax is still fine.
[1] https://go.dev/blog/loopvar-preview
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
We want to update to golang 1.22 but f39 only ships 1.21. This means we
can no longer build on f39 once the update is done. Given we do not ship
podman 5.0 on f39 anyway this is not a problem.
So drop the f39 tasks but given we use them to test different db/storage
drivers we cannot just remove them. Instead we now test f40 twice with
the different db/storage drivers to ensure we keep full coverage.
It do however drop it from the container int task. I really do not see
the value there to test a different db backend as this doesn't depend on
anything in the container.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Since commit 26cd01ee51 we only test netavark and cni support was
dropped. Remove the leftover CI_DESIRED_NETWORK reference.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This check has a condition on the distro name to only run once, however
the prior fedora version doesn't have to exists necessarily as we might
have to drop support there due the outdated golang version.
The current fedora version should alway exists so this seems safer.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This changes the value emitted for HostConfig.Devices from 'null' to
'[]'. The 'null' value was preventing ansible's podman module from
working correctly on FreeBSD.
Signed-off-by: Doug Rabson <dfr@rabson.org>
This commit vendor pre-release version of `c/common:8483ef6022b4`.
It also adapts the code to the new `c/common/libimage` API, which
fixes an image listing race that was listing false warnings.
fixes: #23331
Signed-off-by: Jan Rodák <hony.com@seznam.cz>
- replace random_string with safename in container/network names
- add ci:parallel tags where possible.
- where not possible, add explanations
- fix a userns leak
Signed-off-by: Ed Santiago <santiago@redhat.com>
Since commit 55ad0d6e0e we do the conditions in the cirrus.yml directly
so there is no longer any need for this.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
The renovate config is used for the renovate bot, validating this in the
prior fedora prebuild setp is just confusing and hidden.
The problem is this image is very big so it is slow to download/extract.
To speed things up given it is only a single file we check the diff if
we even changed it.
Now one could argue this should be part of the validate Makefile target
but I given the size I do not want this run by default and I am not sure
if we should do the diff check in the Makefile.
Lastly remove -it, these is meant for interactive use and throws a
warning here because we have no actual tty attached.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This doesn't help us at all, first the list is outdated. AFAICT we no
longer connect to docker.io, registry.fedoraproject.org or
podman.cachix.org (seems to be a cache site for nix.dev?) anywhere in
our tests.
Second a simple port check is not helpful, in the most cases the
CDN's and or load balancer accept connections but return internal server
errors when the registy goes down.
This is very similar to commit 5b6de98ee8.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
They do not add much value to users, first of it compiles podman with
cgo disabled which means the included the podman binary is unusable
either way. The only goal of the build job is to ensure we can compile
on all arches, i.e. go build tags adn types work correctly. The upload
if these artifacts alone take over 90s so let's get rid of them to speed
up the total CI time.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
It is not used by anybody so we do not have to store these and can safe
some time by not having to generate it even if it is just ~500ms.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Two weeks ago we had a weird hang in the macos build job on the
persitent worker. The task just hang for an hour wasting time.
Most tests are fast so we do not need/want such high timeouts.
Therefore drop the default timeout to 20 minutes. The integration task
also should take under 20m so we can remove the longer timeout there as
well. Some system tests need a bit over 30m currently, timeout is set to
35m. For machine tests we use 30m on linux, 45m on windows and 35m on
macos to have some extra room there as machine tests have a much higher
run to run variance.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
It is flaking[1] from time to time on PRs and doesn't really add value
because noone is actually chaing the Containerfile. The task on main
that build a image based of main still exists as per Lokesh
podman-dekstop uses it for testing. In the near term we will get proper
builds in the podman-machine-os repo so this other workflow can
hopefully removed altogether.
[1] https://github.com/containers/podman/actions/runs/10618524888/job/29433963845?pr=23807
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
Podman machine list now supports a new option, --all-providers, which lists all machines from all providers.
Signed-off-by: Ashley Cui <acui@redhat.com>
Generated at build time from troubleshooting.md. Purpose is
to ship an actual man page to end users.
Much more complicated than initial guess, because there was
a bug in my Makefile man page filtering, the sed expression
that cleans up markdown that does not translate to roff.
All I've done here is reorder some of the expressions,
stripping off https links *before* we process
podman man page links.
Signed-off-by: Ed Santiago <santiago@redhat.com>
