We want to modify /etc/passwd to add an entry for the user in
question, but at the same time we don't want to require the
container provide a /etc/passwd (a container with a single,
statically linked binary and nothing else is perfectly fine and
should be allowed, for example). We could create the passwd file
if it does not exist, but if the container doesn't provide one,
it's probably better not to make one at all. Gate changes to
/etc/passwd behind a stat() of the file in the container
returning cleanly.
Fixes#7515
Signed-off-by: Matthew Heon <mheon@redhat.com>
With previous versions of Podman (like v1.9.2) it was always possible to
specify the log level in any case, for example `INFO`. This behavior has
silently changed, where the `--log-level` flag only accepts lower case
levels. This commit re-enables the old behavior and adds an e2e test for
it.
Signed-off-by: Sascha Grunert <sgrunert@suse.com>
The test that does 'adduser' in a keep-id container had a
really dumb bug: if the user running the test has UID 1000,
then podman itself (via keep-id) will add the "1000" passwd
entry, and the in-container "adduser" will allocate 1001,
making our test fail. This triggered in f31/f32 podman gating
tests, but (?!?) never in rawhide gating tests.
Solution: explicitly feed a UID to adduser. Make sure that
it's not the same as the UID of the current user.
Also (unrelated): fix a ridiculous "run mkdir || die". At
the time I wrote that I probably had no idea how BATS works.
Signed-off-by: Ed Santiago <santiago@redhat.com>
because a pod's network information is dictated by the infra container at creation, a container cannot be created with network attributes. this has been difficult for users to understand. we now return an error when a container is being created inside a pod and passes any of the following attributes:
* static IP (v4 and v6)
* static mac
* ports -p (i.e. -p 8080:80)
* exposed ports (i.e. 222-225)
* publish ports from image -P
Signed-off-by: Brent Baude <bbaude@redhat.com>
Fixes: 4c75fe3f70 ("fix pod creation with "new:" syntax")
Commit 4c75fe3f70 passes all net options to the pod but forgot
to unset the options for the container creation. This leads to
erros when using flags like `--ip` since we tried setting
the ip on the pod and container which obviously fails.
I didn't notice the bug because we don't throw an error when
specifing port bindings on a container which joins the pods
network namespace. (#7373)
Also allow the use of `--hostname` and pass that option to the
pod and unset it for the container. The container has to use
the pods hostname anyway. This would error otherwise.
Added tests to prevent regression.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
info, images, run, networking tests: remove some skip_if_remote()s
that were added in the varlink days. All of these tests now seem
to work with APIv2.
help test: check that first output line from 'podman --help'
is the program description (regression check for #7273).
load test: clean up stray images, rewrite test to make it conform
to existing convention. In the process, discover and file #7337
exec test (and networking): file #7360, and add FIXME comment
to skip()s suggesting evaluating those tests once that is fixed.
pod test: now that #6328 is fixed, use 'podman pod inspect --format'
instead of relying on jq
Various other tests: add an explanation of why test is disabled
so we can more easily distinguish "this will never be meaningful
under remote" vs "hey, doesn't work for now, but maybe someday".
Signed-off-by: Ed Santiago <santiago@redhat.com>
The output of 'podman image trust' is in random order; but
its e2e test was assuming a specific one. This caused flakes.
Fixes: #6764
Signed-off-by: Ed Santiago <santiago@redhat.com>
One of the --iidfile tests was flaking:
Error: failed to write image ID to file "/tmp/dir/idFile": open /tmp/dir/idFile: no such file or directory
Root cause: test was actually not mkdir'ing /tmp/dir. Test was
mostly passing because _other_ tests in the suite were mkdir'ing
it, but once in a while this test ran before the others.
Solution: fixed this test to use CreateTempDirInTempDir(). And,
since hardcoded tempdirs are bad practice, grepped for '"dir"'
and fixed all other instances too.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Some CI tests are hanging, timing out in 60 or 120 minutes.
I wonder if it's #7316, the bug where all podman commands
hang forever if NOTIFY_SOCKET is set?
Signed-off-by: Ed Santiago <santiago@redhat.com>
It's not possible to run any of the scripts on distributions which do
have `bash` not in `/bin`. This is being fixed by using `/usr/bin/env
bash` instead.
Signed-off-by: Sascha Grunert <sgrunert@suse.com>
podman containers using IPv6 were missing the default route, breaking
deployments trying to use them.
The problem is that the default route was hardcoded to IPv4, this
takes into consideration the podman subnet IP family to generate
the corresponding default route.
Signed-off-by: Antonio Ojea <aojea@redhat.com>
Oops. PR #6693 (sdnotify) added tests, but they were disabled
due to broken crun on f31. I tried for three weeks to get a
magic CI:IMG PR to update crun on the CI VMs ... but in that
time I forgot to actually enable those new tests.
This PR removes a 'skip', replacing it with a check that systemd
is running plus one more to make sure our runtime is crun. It
looks like sdnotify just doesn't work on Ubuntu (it hangs), and
my guess is that it's a crun/runc issue.
I also changed the test image from fedora:latest to :31, because,
sigh, fedora:latest removed the systemd-notify tool.
WARNING WARNING WARNING: the symptom of a missing systemd-notify
is that podman will hang forever, not even stopped by the timeout
command in podman_run! (Filed: #7316). This means that if the
sdnotify-in-container test ever fails, the symptom will be that
Cirrus itself will time out (2 hours?). This is horrible. I
don't know what to do about it other than push for a fix for 7316.
Signed-off-by: Ed Santiago <santiago@redhat.com>
it allows to customize the options passed down to the OCI runtime for
setting up the /proc mount.
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
This adds support for the --cidr parameter that is supported
by slirp4netns since v0.3.0. This allows the user to change
the ip range that is used for the network inside the container.
Signed-off-by: Adis Hamzić <adis@hamzadis.com>
the deepcopy in the remote history code path was throwing an uncaught error on a type mismatch. we now manually do the conversion and fix the type mismatch on the fly.
Fixes: #7122
Signed-off-by: Brent Baude <bbaude@redhat.com>
To sync the behavior between AppArmor and seccomp it is now possible to
also specify seccomp profiles for privileged containers.
Signed-off-by: Sascha Grunert <sgrunert@suse.com>
Buildah and podman build can create images without a working dir.
FROM fedora
WORKDIR /test
If you build this image with caching twice, the second time the image
will not have a working dir.
Similarly if you execute
podman run --workdir /foobar fedora
It blows up since the workingdir is not created automatically.
Finally there was duplicated code for getting the workingdir
out of an image, that this PR removes.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Refactor the processing of Repository and Tag fields to default to <none>
when printing via --format flag. Previously, the default format would
print <none> but --format {{.Tag}} would not in some cases.
Fixes#7123
Signed-off-by: Jhon Honce <jhonce@redhat.com>
Fix a bug in the error handling which returned nil instead of an error
and ultimately lead to nil dereferences in the client. To prevent
future regressions, add a test and check for the error message.
Fixes: #7271
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
This matches Docker behavior, and seems to make sense - the CMD
may have been specific to the original entrypoint and probably
does not make sense if it was changed.
While we're in here, greatly simplify the logic for populating
the SpecGen's Command. We create the full command when making the
OCI spec, so the client should not be doing any more than setting
it to the Command the user passed in, and completely ignoring
ENTRYPOINT.
Fixes#7115
Signed-off-by: Matthew Heon <mheon@redhat.com>
- new sanity checks for podman-remote:
- first, confirm that when PODMAN is "-remote",
we actually talk to a server (validated by
presence of "Server:" string in "podman version").
- second, add test for #7212, in which we run
"podman --remote" (podman with --remote flag,
not podman-remote command) and make sure --remote
is allowed both as the first option and also
with other flag options preceding.
- new test for "podman image tree" (piggybacking on
top of a "podman build" test, because that gives
us lots of layers).
- skip "podman exec - basic test" when remote. It is consistently
causing CI failures, breaking all of CI, due to #7241.
Signed-off-by: Ed Santiago <santiago@redhat.com>
