The project was sunset;
> After more than a decade of serving the ecosystem, the time
> has come to sunset Go Report Card. Following the loss of our
> primary infrastructure sponsor, maintaining the web app is
> no longer sustainable.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Given that the majority of reviewers favor the 'nr_inodes=2' retry logic,
we propose reverting #5353 and implementing the new approach.
1. Revert "libct: add a fallback for nr_inodes=2"
This reverts commit 79ac57770f.
2. Revert "libct: Enforce nr_inodes=2 to fix Focal mount errors"
This reverts commit feea25820e.
3. The new approach:
On most kernels `nr_inodes=1` works fine. However, Ubuntu 20.04 (Focal) with
the official 5.4 kernel carries a private patch in "mm/shmem.c" that rejects
`nr_inodes<2`, so retry with `nr_inodes=2` here.
For reference, search for "case Opt_nr_inodes" in:
https://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux/+git/focal/plain/mm/shmem.c?h=Ubuntu-5.4.0-216.236
Signed-off-by: lifubang <lifubang@acmcoder.com>
We don't know whether some kernels will fail with "nr_inodes=2",
so let's fall back to mount a tmpfs without "nr_inodes".
Signed-off-by: lifubang <lifubang@acmcoder.com>
On most kernels `nr_inodes=1` works fine. However, Ubuntu 20.04 (Focal) with
the official 5.4 kernel carries a private patch in mm/shmem.c that rejects
"nr_inodes<2", so let's keep `nr_inodes=2` here!
Signed-off-by: lifubang <lifubang@acmcoder.com>
We try to delete the interface, but it lot of tests it won't be there
unless we failed to move it to the container. Let's just clarify that in
a comment and redirect the error output to /dev/null, as it seems an
error otherwise while it is completely normal.
Signed-off-by: Rodrigo Campos <rodrigo@amutable.com>
The cleaning is condition on this variable being set. So let's unset it
after we clean the resources.
Signed-off-by: Rodrigo Campos <rodrigo@amutable.com>
We are creating the interface for every test, but there is only one
using it. Let's just call the function to create the netdev on the test
that uses it.
I guess that was the reason we had the "ip link del ..." in teardown.
Because in a lot of tests we were just creating and deleting the
interface on the host.
While we are there, as suggested by lifubang, let's make the "ip link
add" line specify the mtu and mac addr. This way, the interface is not
created without that info and we race with host daemons (like udev) that
_might_ want to change it.
Signed-off-by: Rodrigo Campos <rodrigo@amutable.com>
Once we add a new network device, systemd-udev may execute some rules.
In particular, we see that on Fedora it sets the MAC address (presumably
based on the host name and device name). This setting races with ours
'ip link set address', as a result, "checkpoint and restore with netdevice"
test sometimes fails telling the MAC address is not as expected.
In the future there may be some other udev rules etc., so the overall
solution is to wait until systemd-udev is finished applying the rules,
thus eliminating the race.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of using a standalone modernize (and a separate job), add
modernize and go vet (which should be same-as/similar-to go fix) linters
to golangci-lint config.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This update includes a few breaking API changes that I needed to get in
before an actual runc release depends on it, so that we don't need to
deal with compatibility shims for them (or bumping the SOVERSION).
From a Go API perspective, there were no major changes -- though this
bump did also require a bump to github.com/cyphar/filepath-securejoin
because one of the wrapped APIs changed from int to uint64 as a flag
argument type. Again, better to get this done before we really depend on
this in a public way.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
go-criu v8.3.0 switches to protobuf-go-lite, which helps to remove
google.golang.org/protobuf dependency from here, reducing the runc
binary size from ~16M to ~14M.
The only missing piece is proto.String, proto.Bool, proto.Int32 etc.
helpers that return a pointer to a given variable. Those are replaced
by a generic mkPtr, which in turn is to be replaced by the new builtin
once Go < 1.26 is no longer supported.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As with /dev symlinks, this was missed in commit d40b3439a9 ("rootfs:
switch to fd-based handling of mountpoint targets"). It's not really
clear to what extent this was exploitable (/sys/fs/cgroup is a tmpfs we
create) but it's better to just fix this anyway.
Fixes: d40b3439a9 ("rootfs: switch to fd-based handling of mountpoint targets")
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
These codepaths are very old and operate on pure paths but before
pivot_root(2), meaning that a bad image with a malicious /dev symlink
could cause us to operate on host paths instead.
In practice this means that we could be tricked into removing a file
called "ptmx" (note that /dev/pts/ptmx and /dev/ptmx are both immune for
different reasons) or creating a very restricted set of symlinks (with
fixed targets and names). The scope of these bugs is thus quite limited,
but we definitely need to harden against it.
These codepaths were unfortunately missed during the fd-based rework in
commit d40b3439a9 ("rootfs: switch to fd-based handling of mountpoint
targets") -- I must've assumed they were called after pivot_root(2)...
Fixes: GHSA-xjvp-4fhw-gc47
Fixes: CVE-2026-41579
Fixes: d40b3439a9 ("rootfs: switch to fd-based handling of mountpoint targets")
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
This argument order matches most other helpers we have and will also
match the changes we are about to make to setupPtmx and
setupDevSymlinks.
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
Replace the goroutine + channel + 100ms time.After + blocking open
in handleFifo with a poll(2) loop on a non-blocking open. Use
pidfd_open(2) where available to wait for init exit without timeout,
falling back to /proc state checks with 100ms timeout on older
kernels.
Fixes#5251
Signed-off-by: Mohammed Aminu Futa <mohammedfuta2000@gmail.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
The hardcoded architecture list was little-endian only, causing
seccomp_arch_add() to fail with -EDOM on s390x.
Drop it. It's optional and libseccomp automatically adds the native
architecture when the filter is created.
Fixes: https://github.com/opencontainers/runc/issues/4835
Signed-off-by: Ricardo Branco <rbranco@suse.de>
The test checked for the exact BusyBox ash diagnostic "sh: can't fork".
With BusyBox 1.38, ash reports the failure as:
/bin/sh: line 0: can't fork: Resource temporarily unavailable
Match the stable "can't fork" part of the error message instead.
Signed-off-by: Ricardo Branco <rbranco@suse.de>
TestPids used long hand-written /bin/true pipelines for the 4-, 32- and
64-command cases. This made the test easy to typo and hard to review, as
seen by the earlier "bin/true" entries.
Build the shell pipelines instead, preserving the existing test coverage
while making the command counts explicit.
Signed-off-by: Ricardo Branco <rbranco@suse.de>
The idea of commit d1fca8e was right (report errors for non-existent
root, unless using the default root dir) but the logic was inverted.
Fix the logic.
Test case for default root requires non-existent /root/runc, which is
not always possible.
Reported-by: RedMakeUp <girafeeblue@gmail.com>
Co-authored-by: RedMakeUp <girafeeblue@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Migrate from urfave/cli v1 (maintenance mode) to v3 to benefit from
active development, improved features, and long-term support.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Close the root file descriptor immediately after use in maskPaths to
reduce the window during which an attacker could potentially exploit
an open fd to access or manipulate the root filesystem. This follows
the principle of least privilege and mitigates risks in compromised
or malicious container scenarios.
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
This is a follow-up to #5275. That change reused a single tmpfs mount
to mask multiple directories, which is efficient when masking more than
one path. However, it introduced unnecessary overhead when only one
directory is masked. This commit restores the original behavior for the
single-path case while preserving shared tmpfs logic for multiple paths.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Switch from sync.Once to sync.OnceFunc and sync.OnceValues.
Keep Root a function (rather than a variable) because godoc
renders function doc better than a variable doc.
Switch to using internal function root internally.
Modify tests accordingly (and simplify NewIntelRdtTestUtil to
fakeRoot).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The whole struct intelRdtStatus with its methods and a sync.Once is not
needed, since intelrtd.Is*Enabled methods are already run-once (or use
run-once and a simple comparison).
Yet it is still needed for the test to fake values returned by *Enabled.
Simplify to use func pointers which a test case overwrites.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It is not doing anything, and tests can just instantiate the &Manager{}.
Suggested-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Use sync.OnceValue.
2. Fix the len(buf) check -- we only need 1 byte. Real kernel output
is "Y\n" so practically this change is a no-op.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As runc binary grows in size over time (new features, more
dependencies) some tests start to flake because of low memory limits.
One such test is "runc run (cgroup v2 resources.unified override)";
it obviously fails because of 1M memory limit:
> runc run failed: unable to start container process: container init was OOM-killed (memory limit too low?)
Increase the limits 4x. Do the same for the "unified only" test.
Fixes issue 5264.
Reported-by: Kevin Berry <kpberry11@gmail.com>
Reported-by: Ricardo Branco <rbranco@suse.de>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kubernetes may add one sysfs thermal_throttle entry per CPU to
maskedPaths. On large Intel systems this can produce many directory
masks for a single container. runc currently handles each directory
mask with a separate read-only tmpfs mount, and therefore a separate
tmpfs superblock.
On Linux 4.18/RHEL 8 kernels, creating and tearing down many tmpfs
superblocks can contend on the global shrinker_rwsem when containers
start or stop concurrently.
Use one read-only tmpfs for directory masks and bind-mount it over the
remaining directory targets. The first non-procfs-fd directory mount is
reopened through the container root fd before it is reused. File masks
still bind /dev/null, and procfs fd targets keep the existing
one-tmpfs-per-target behaviour because they are fd aliases rather than
stable rootfs paths.
If the bind-mount of the shared source fails (e.g. due to kernel
restrictions), fall back to individual tmpfs mounts for all remaining
directories. Tmpfs mounts use nr_blocks=1,nr_inodes=1 to minimise
kernel resource usage.
The bind mounts do not create additional tmpfs superblocks. They also
retain the read-only mount flag inherited from the source vfsmount, so
the masking semantics remain unchanged.
xref: kubernetes/kubernetes#138512
xref: kubernetes/kubernetes#138388
xref: kubernetes/kubernetes#131018
Co-authored-by: Davanum Srinivas <davanum@gmail.com>
Refactored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
Previously, masked directories (e.g., /proc/acpi, /proc/scsi) were
mounted as read-only tmpfs without explicit size or inode limits.
Although these mounts are meant to be empty and unwritable, the lack
of resource constraints means that—should an attacker bypass the
read-only protection (e.g., via container escape, mount namespace
manipulation, or a kernel vulnerability)—the tmpfs could consume up
to 50% of system memory by default (the kernel's default tmpfs limit).
To mitigate this risk in high-density container environments and
adhere to the principle of least privilege, we now explicitly set:
- nr_blocks=1 (sufficient for at most one block size)
- nr_inodes=1 (sufficient for at most one inode)
Ref: https://man7.org/linux/man-pages/man5/tmpfs.5.html
These limits ensure that even if compromised, kernel memory usage
remains strictly bounded and negligible.
This change aligns with best practices used by other container
runtimes and strengthens defense-in-depth for sensitive masked paths.
Co-authored-by: Davanum Srinivas <davanum@gmail.com>
Refactored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
Apparently, lima's experimental/fedora-rawhide image does not include
which rpm, and we don't really want to bother installing it.
Replace "which" with "command -v". Looks like this was the only place;
we already use "command -v" everywhere else.
This should fix lima (experimental/fedora-rawhide) CI.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kir Kolyshkin (3):
runc exec -p: fix adding HOME to nil env
tests/int/env.bats: add test for runc exec -p
tests/int: amend runc exec --env test
LGTMs: rata cyphar
Before commit 7dc24868, when process.env was nil, prepareEnv
returned a flag telling HOME is not set, and it was added.
Commit 7dc24868 moved the functionality of adding HOME into
prepareEnv but did not properly handle nil case. As a result,
runc exec -p with process.json having no env set resulted in
an exec with no HOME set.
Fix this, and add unit and integration tests.
Fixes: 7dc24868 ("libct: switch to numeric UID/GID/groups")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
All existing tests check runc run, and there is no single runc exec
environment test except for one in exec.bats.
Add it (no new issues found).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This tests checks that "runc exec --env VAR=VAR ..." actually appends
VAR=VAL to the exec's environment.
Add additional checks that:
- process.env from config.json is also inherited;
- HOME is set.
Those checks do not reveal any new issues.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
runc's build scripts and CI setup used a mix of wget (4 sites) and
curl (8 sites) for HTTPS fetches. Standardise the three script sites on
curl, which is already required by the majority of call sites.
* script/build-libpathrs.sh and script/build-seccomp.sh: replace
wget "<url>"{,.asc} with curl -fsSLO "<url>"{,.asc}. Bash brace
expansion still yields two separate downloads (the tarball and its
.asc signature).
* script/setup_host.sh: replace wget with curl in the Fedora RPM
install list.
Leaving the single wget call in Dockerfile untouched per review
feedback -- the trixie base image already ships wget, and switching
would add an extra apt-get update && install step before the existing
single-pass package install.
Refs #5240
Signed-off-by: Ali <alliasgher123@gmail.com>
Since commit 3cdda46 the poststart hooks runs after the container
process start, and so they race.
Move the poststart hook check to a separate step after the container
process has exited.
Fixes 5245.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test is racy for a long time now. All the logs I could find in CI
seem to be dangling symlinks, like the test shows "23 -> ". This means
the fd was closed before we did the call to readlink().
Let's try to disable the GC. This should get rid of the "fds are getting
closed before we read them" part.
Updates: #4297
Signed-off-by: Rodrigo Campos <rodrigo@amutable.com>
By default, readlink is silent about any errors. Make it verbose so we
can better interpret any test failures.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When running a process inside a container, make sure its stderr is not
nil (except for some trivial cases like cat). Modify waitProcess to show
failed command's stderr, if possible.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since Wait returns an ExitError if process' exit status is not 0,
checking process status is redundant and this code is never reached.
Remove it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
add bat integration test for rootfs propagation test, expect to
see the mount propagation is slave, the test will create a isolate mntns
to run the test as the test will mutate the rootfs propagation
Signed-off-by: sean <xujihui1985@gmail.com>
When rootfsPropagation is set to rslave, prepareRoot() was forcing the
rootfs parent mount to MS_PRIVATE before bind-mounting and pivoting into
the rootfs. That breaks the slave relationship needed for HostToContainer
propagation, so later unmount/remount events on host mountpoints under
the rootfs are not reflected inside the running container.
Fix this by keeping the rootfs parent mount as MS_SLAVE for slave-like
rootfs propagation settings, while leaving the final root propagation
remount in place.
Signed-off-by: sean <xujihui1985@gmail.com>
The RUNC_ALLOW_UNSAFE_TESTS variable set in the Cirrus CI env block
does not reach the integration tests because they are executed via
"ssh -tt localhost make ...", which starts a new login shell that
does not inherit the caller's environment. As a result, unsafe tests
are always skipped in Cirrus CI even though the intent is to run them.
Fix this by exporting the variable in /root/.bashrc (same way we
already handle PATH), so the ssh session picks it up.
See #5212 (comment).
Fixes: 9932ad19 ("tests/int: introduce the concept of unsafe tests")
Signed-off-by: RedMakeUp <girafeeblue@gmail.com>
Those tests were added by commit 8d180e96 ("Add support for Linux
Network Devices"), apparently by copy-pasting the test cases which
call simple_cr (all four of them).
While different simple_cr tests make sense as they cover different
code paths in runc and/or check for various regression, the same
variations with netdevice do not make sense, as having a net device
is orthogonal to e.g. bind mount, --debug, or cgroupns.
Remove those.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When RUNC_USE_SYSTEMD is set, tests/rootless.sh is using
ssh -tt rootless@localhost
to run tests as rootless user. In this case, local environment is not
passed to the user's ssh session (unless explicitly specified), and so
the tests do not get ROOTLESS_FEATURES.
As a result, idmap-related tests are skipped when running as rootless
using systemd cgroup driver:
integration test (systemd driver)
...
[02] run rootless tests ... (idmap)
...
ok 286 runc run detached ({u,g}id != 0) # skip test requires rootless_idmap
...
Fix this by creating a list of environment variables needed by the
tests, and adding those to ssh command line (in case of ssh) or
exporting (in case of sudo) so both cases work similarly.
Also, modify disable_idmap to unset variables set in enable_idmap so
they are not exported at all if idmap is not in features.
Fixes: bf15cc99 ("cgroup v2: support rootless systemd")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These helpers all make more sense as a self-contained package and moving
them has the added benefit of removing an unneeded libpathrs dependency
(from libcontainer/utils's import of pathrs-lite) from several test
binaries.
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
This allows users to automaticaly migrate to the new location
using `go fix`. It has some limitations, but can help smoothen
the transition; for example, taking this file;
```
package main
import (
"github.com/opencontainers/runc/libcontainer/devices"
)
func main() {
_, _ = devices.DeviceFromPath("a", "b")
_, _ = devices.HostDevices()
_, _ = devices.GetDevices("a")
}
```
Running `go fix -mod=readonly ./...` will migrate the code;
```
package main
import (
devices0 "github.com/moby/sys/devices"
)
func main() {
_, _ = devices0.DeviceFromPath("a", "b")
_, _ = devices0.HostDevices()
_, _ = devices0.GetDevices("a")
}
```
updates b345c78dca
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Some of runc integration tests may do something that I would not like
when running those on my development laptop. Examples include
- changing the root mount propagation [1];
- replacing /root/runc [2];
- changing the file in /etc (see checkpoint.bats).
Yet it is totally fine to do all that in a throwaway CI environment,
or inside a Docker container.
Introduce a mechanism to skip specific "unsafe" tests unless an
environment variable, RUNC_ALLOW_UNSAFE_TESTS, is set. Use it
from a specific checkpoint/restore test which modifies
/etc/criu/default.conf.
[1]: https://github.com/opencontainers/runc/pull/5200
[2]: https://github.com/opencontainers/runc/pull/5207
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The runtime-spec [1] currently says:
> 6. Runtime's start command is invoked with the unique identifier of
> the container.
> 7. The startContainer hooks MUST be invoked by the runtime. If any
> startContainer hook fails, the runtime MUST generate an error, stop
> the container, and continue the lifecycle at step 12.
> 8. The runtime MUST run the user-specified program, as specified by
> process.
> 9. The poststart hooks MUST be invoked by the runtime. If any
> poststart hook fails, the runtime MUST generate an error, stop the
> container, and continue the lifecycle at step 12.
> ...
> 11. Runtime's delete command is invoked with the unique identifier of
> the container.
> 12. The container MUST be destroyed by undoing the steps performed
> during create phase (step 2).
> 13. The poststop hooks MUST be invoked by the runtime. If any poststop
> hook fails, the runtime MUST log a warning, but the remaining hooks
> and lifecycle continue as if the hook had succeeded.
Currently, we do 9 before 8 (heck, even before 6), which is clearly
against the spec and results in issues like the one described in [2].
Let's move running poststart hook to after the user-specified process
has started.
NOTE this patch only fixes the order and does not implement removing
the container when the poststart hook failed (as this part of the spec
is controversial -- destroy et al and should probably be, and currently
are, part of "runc delete").
[1]: https://github.com/opencontainers/runtime-spec/blob/main/runtime.md#lifecycle
[2]: https://github.com/opencontainers/runc/issues/5182
Reported-by: ningmingxiao <ning.mingxiao@zte.com.cn>
Reported-by: Erik Sjölund <erik.sjolund@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Rename c.signal to c.signalInit, and add c.signal which is a lock-less
form of c.Signal.
To be used by the next patch.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The libcontainer/devices package has been moved to moby/sys/devices, so
we can just point users to that and keep some compatibility shims around
until runc 1.6. We don't use it at all so there are no other changes
needed.
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
When signal installation was moved to a goroutine for performance,
containers that exited quickly could complete before SIGCHLD was
registered, causing runc to hang waiting for the signal.
This fix ensures SIGCHLD is registered immediately in the main thread
before other signals are handled in the goroutine, maintaining performance
while guaranteeing no missed SIGCHLD notifications for fast-exiting
containers.
Reported-by: Ayato Tokubi <atokubi@redhat.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
For detached container, we don't need to setup signal notifier, because
there is no customer to consume the signals in `forward()`.
Signed-off-by: lifubang <lifubang@acmcoder.com>
In fact, notifySocket has no relationship to signalHandler, we
can move it out of signalHandler to make the code more clear.
Signed-off-by: lifubang <lifubang@acmcoder.com>
We should use a named return value of error, or else we can't
catch all errors when calling defer function, for example we
used a block scope var name `err` for `setupPidfdSocket`.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Since switching to Go 1.25 in go.mod, the "detect fd leaks" test fails
like this:
> not ok 57 runc create[detect fd leak as comprehensively as possible]
> # (in test file tests/integration/create.bats, line 76)
> # `[ "$violation_found" -eq 0 ]' failed
> ...
> # Violation: FD 9 -> '/system.slice/runc-test_busybox.scope/cpu.cfs_quota_us'
> # Violation: FD 10 -> '/system.slice/runc-test_busybox.scope/cpu.cfs_period_us'
> ...
This happens because Go 1.25 adds a feature to dynamically set GOMAXPROC
based on current CPU quota values. This feature can be disabled by setting
GODEBUG=containermaxprocs=0,updatemaxprocs=0
but it is harmless to keep it (except for the above test failure).
Add an exception to the test case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This uses preopened rootfs in Chdir and pivotRoot.
While at it, add O_PATH when opening oldroot in pivotRoot.
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: lfbzhm <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A lot of filesystem-related stuff happens inside the container root
directory, and we have used its name before. It makes sense to pre-open
it and use a *os.File handle instead.
Function names in internal/pathrs are kept as is for simplicity (and it
is an internal package), but they now accept root as *os.File.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Indeed, it does not make sense to prepend c.root once we started using
MkdirAllInRoot in commit 63c29081.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Li Fubang (3):
test: check mount source fds are cleaned up with idmapped mounts
libct: close mount source fd as soon as possible
libct: add a nil check for mountError
LGTMs: kolyshkin rata cyphar
word boundary anchor \> is present only in GNU awk implementation. This
will not work if we are building libpathrs inside a container like
debian/ubuntu which uses mawk. Therefore switch to a POSIX compatible
regex that can be used with all implementations of awk to find the
system libc path
Signed-off-by: Akhil Mohan <akhilerm@gmail.com>
Signed-off-by: Aleksa Sarai <aleksa@amutable.com>
This commit factors out setupAndMountToRootfs without changing any
logic. Use "Hide whitespace changes" during review to focus on the
actual changes.
The refactor ensures the mount source file descriptor is closed via
defer in each loop iteration, reducing the total number of open FDs
in runc. This helps avoid hitting the file descriptor limit under
high concurrency or when handling many mounts.
Signed-off-by: lifubang <lifubang@acmcoder.com>
A bit of history. EXTRA_BUILDTAGS was introduced in commit dac417174,
as a quick way to add some extra Go build tags to the runc build.
Later, commit 767bc008 changed Makefile to not get EXTRA_TAGS from the
shell environment, as the name is quite generic and some unrelated
environment variable with that name can affect runc build. While such
change does make sense, it makes it more complicated to pass build tags
in CI and otherwise (see e.g. commit 0e1fe368a).
Moreover, runc build uses some Go build tags by default (via Makefile),
and while it is easy to add more build tags (via EXTRA_BUILDTAGS), in
order to remove some existing tags one has to redefine BUILDTAGS from
scratch, which is not very convenient (again, see commit 0e1fe368a which
gets the current value of BUILDTAGS from the Makefile in order to remove
a single tag).
To handle all of the above, let's do this:
- implement RUNC_BUILDTAGS, fixing the issue of not-so-unique name;
- allow to get RUNC_BUILDTAGS from shell environment;
- implement a feature to remove a build tag from default set by
prefixing it with "-" (as in RUNC_BUILDTAGS="-seccomp");
- document all this in README;
- make CI use the new feature;
- keep EXTRA_BUILDTAGS for backward compatibility, add a make warning
and a TODO to remove it for runc 1.6.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This adds support for WaitKillableRecv seccomp flag
(also known as SCMP_FLTATR_CTL_WAITKILL in libseccomp and
as SECCOMP_FILTER_FLAG_WAIT_KILLABLE_RECV in the kernel).
This requires:
- libseccomp >= 2.6.0
- libseccomp-golang >= 0.11.0
- linux kernel >= 5.19
Note that this flag does not make sense without NEW_LISTENER, and
the kernel returns EINVAL when SECCOMP_FILTER_FLAG_WAIT_KILLABLE_RECV
is set but SECCOMP_FILTER_FLAG_NEW_LISTENER is not set.
For runc this means that .linux.seccomp.listenerPath should also be set,
and some of the seccomp rules should have SCMP_ACT_NOTIFY action. This
is why the flag is tested separately in seccomp-notify.bats.
At the moment the only adequate CI environment for this functionality is
Fedora 43. On all other platforms (including CentOS 10 and Ubuntu 24.04)
it is skipped similar to this:
> ok 251 runc run [seccomp] (SECCOMP_FILTER_FLAG_WAIT_KILLABLE_RECV) # skip requires libseccomp >= 2.6.0 and API level >= 7 (current version: 2.5.6, API level: 6)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Found out that these changes were backported to release-1.4 (PR 5040)
and made its way into runc v1.4.0, but were missing from its CHANGELOG.
Move the item to v1.4.0 changelog, and add a backport PR reference.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Aleksa Sarai (4):
VERSION: back to development
VERSION: release v1.5.0-rc.1
CHANGELOG: forward-port v1.4.1 entry
CHANGELOG: add original PR references to backports
LGTMs: rata lifubang
Some of the patches in the "unreleased" section were backported and so
they can be moved to the v1.4.1 section to some unneeded avoid
duplication.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
When going through the changelog for v1.5.0-rc.1, these PRs were merged
but not referenced by their original PR number in the changelog, making
it harder to figure out which patches since v1.4.0-rc.1 are in
v1.5.0-rc.1 or v1.4.0.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Commit 192e3d416f ("ci: add conmon tests run") was merged without
rebasing on top of commit e2c989b7e1 ("build: enable libpathrs by
default"), causing build failures when it was merged.
The solution is to just use the same install script as the rest of CI
from commit 7322b05f41 ("ci: build and install libpathrs").
Fixes: 192e3d416f ("ci: add conmon tests run")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Aleksa Sarai (11):
gha: test both with and without libpathrs
build: enable libpathrs by default
ci: build and install libpathrs
build: enable builds with libpathrs
deps: update to cyphar.com/go-pathrs@v0.2.4
README: document libpathrs build tag
script: seccomp.sh -> build-seccomp.sh
build: rename /opt/libseccomp cdylib directory
build: treat armhf as ARMv7
dockerfile: switch to Debian 13
integration: output debug information in fd leak test
LGTMs: rata kolyshkin AkihiroSuda
This entry was added by commit 653161f6d8 ("docs/spec-conformance.md:
update for spec v1.3.0") but the spec was updated before the v1.3.0
release to remove this requirement for poststart hooks in order to match
runc's current behaviour.
Fixes: 653161f6d8 ("docs/spec-conformance.md: update for spec v1.3.0")
Ref: opencontainers/runtime-spec#1262
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We do plan to make libpathrs required in the future, but in the meantime
we should test both with and without libpathrs in our CI to catch
regressions for users that will not use libpathrs initially.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
libpathrs has better hardening against certain attacks (most notably on
older kernels) so we should use it by default. This opens the door to
us using cyphar.com/go-pathrs in the future in order to remove some of
our internal/pathrs wrappers (that reimplement bits of libpathrs).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
pathrs-lite supports transparently switching to libpathrs.so as the
backend with the "libpathrs" build tag. In order to make this work
properly with our CI and release build scripts, we we need to have a
similar setup as with we do with libseccomp.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Ever since v0.6.0 of github.com/cyphar/filepath-securejoin, pathrs-lite
has been able to transparently switch to using libpathrs as the backend
for safe path resolution (at compile-time, using a build tag). Note that
because build-tags apply globally, this allows for us to easily opt
pure-Go dependencies into all using libpathrs as well for our binaries.
In a future patch this will likely be enabled by default, but document
that this is an option for downstreams that want to opt-in to using
libpathrs.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The intention of commit 531e29e192 ("script/lib.sh: set GOARM=5 for
armel, GOARM=6 for armhf") was to properly support older ARM platforms
with our release builds.
However, we have never been able to support ARMv6 for our builds because
we use the Debian compiler to build the libseccomp we statically compile
into our binaries and (as per the now-deleted comment itself) Debian
treats armhf as being ARMv7 so the final binaries we produced were
always only ever compatible with ARMv7+.
This was a bit of an oddity before but when building libpathrs for
releases we will need to use Rust which makes the target more explicit
(and while it does support armhf, we are using the Debian-packaged Rust
cross-compiler and thus are in the same dilemma with what Debian
considers "armhf" to be).
All-in-all, it's better to just bite the bullet and just follow Debian
here properly.
Fixes: 531e29e192 ("script/lib.sh: set GOARM=5 for armel, GOARM=6 for armhf")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Debian 13 (trixie) was released a few months ago and it's probably
prudent to just upgrade. This is also necessary to get access to riscv64
repositories when we build libpathrs for inclusion in our runc binaries.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
When deprecating Relabel field, its json attributes were mistakenly
removed, so now it is:
- saved to JSON under "Relabel" (rather than "relabel");
- won't be ignored if empty.
Let's fix it before it's too late.
Fixes: 8b2b5e94 ("libct: remove relabeling dead code")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It makes more sense to save whether we should cleanup the directory
after it gets created (to avoid error cases deleting a different
directory) as well as tying this check to the existing os.ErrExist
check rather than doing an extra stat(2).
Fixes: e2baa3ad10 ("Intel RDT: update according to spec changes.")
Suggested-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This fixes random failures to start a container in conmon integration
tests (see issue 5151).
I guess we need to find another way to fix issue 4645.
This reverts commit 1b39997e73.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no way to set Mount.Relabel field via OCI spec (config.json),
and so the relabeling code is never used.
My guess it's a leftover from times when runc used to be part of Docker.
Remove it, and mark Relabel field as deprecated.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is one proposed clarification to the OCI spec: the subdirectory
needs to be deleted. Runc already does that, but the clarification adds
for directory removal only if the directory was created by us.
Signed-off-by: Ismo Puustinen <ismo.puustinen@intel.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
unix.CPUSet is limited to 1024 CPUs. Calling
unix.SchedSetaffinity(pid, cpuset) removes all CPUs starting from 1024
from allowed CPUs of pid, even if cpuset is all ones. As a
consequence, when runc tries to reset CPU affinity to "allow all" by
default, it prevents all containers from CPUs 1024 onwards.
This change uses a huge CPU mask to play safe and get all possible
CPUs enabled with a single sched_setaffinity call.
Fixes: #5023
Signed-off-by: Antti Kervinen <antti.kervinen@intel.com>
Some maintainers appear to have removed their PGP keys, which causes
"gpg --import" during "make validate-keyring" to fail. The solution is
to switch to a non-fatal warning if no keys were imported.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This was a really ugly hack to try to reduce the impact of our original
set of CVE-2019-5736 mitigations, but unfortunately had too many caveats
to its use to ever be really useful. In addition, it was completely
obsoleted by the migration to using an detached overlayfs mount in
commit 515f09f7b1 ("dmz: use overlayfs to write-protect /proc/self/exe
if possible").
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These were all marked deprecated in commit a75076b4a4 ("Switch to
opencontainers/cgroups") when we switched maintenance of our cgroup code
to opencontainers/cgroups.
Users have had ample time to switch to opencontainers/cgroups
themselves, so we can finally remove this.
Note that the whole libcontainer/devices package will be moved to
moby/sys in the near future, so this whole package will be marked
deprecated soon.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These were inadvertently added to our exported APIs by commit
eeda7bdf80cca ("Add memory policy support"). We couldn't remove them
from runc 1.4.x, but we deprecated them in commit 3741f9186d
("libct/configs: mark MPOL_* constants as deprecated") and marked them
for removal in runc 1.5. Users should never have used these in the first
place.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This was deprecated in commit e6a4870e4ac40 ("libct: better errors for
hooks"), and users have had ample time to migrate to Hooks.Run since.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These were all marked deprecated in commit a75076b4a4 ("Switch to
opencontainers/cgroups") when we switched maintenance of our cgroup code
to opencontainers/cgroups.
Users have had ample time to switch to opencontainers/cgroups
themselves, so we can finally remove this.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
There is a chance of panic here -- eliminate it.
Add a test case (which panics before the fix).
Reported-by: Luke Hinds <luke@stacklok.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Example code in README is outdated (especially since cgroups is moved to
a separate repository) and lacks proper import statements. And, since it
is not code, it is hard to keep it up to date.
Let's move it out to the example_test.go file and refer to it. Note we
still don't run it, but it will be compiled and linted in CI.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Fix *some* of the prealloc linter warnings. While it does not make sense
to address all warnings (or add prealloc to the list of linters we run
in CI), some do make sense.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
SCMP_ACT_KILL terminates the process with a fatal signal, which may
produce a core dump depending on the host configuration.
While this is harmless on ephemeral CI instances, it can leave unwanted
core files on developer or customer systems. It also interferes with
test environments that detect unexpected core dumps.
Signed-off-by: Ricardo Branco <rbranco@suse.de>
When container-selinux 4:2.246.0-1.el10 is installed, it produces the
following %post script warnings:
> ...
> Running scriptlet: container-selinux-4:2.246.0-1.el10.noarch 26/37
> Installing : container-selinux-4:2.246.0-1.el10.noarch 26/37
> Running scriptlet: container-selinux-4:2.246.0-1.el10.noarch 26/37
> libsemanage.semanage_pipe_data: Child process /usr/libexec/selinux/hll/pp failed with code: 255. (No data available).
> libsemanage.semanage_compile_module: container: libsepol.policydb_read: policydb module version 24 does not match my version range 4-23.
> libsemanage.semanage_compile_module: container: libsepol.sepol_module_package_read: invalid module in module package (at section 0).
> libsemanage.semanage_compile_module: container: libsepol.sepol_ppfile_to_module_package: Failed to read policy package.
> libsemanage.semanage_direct_commit: Failed to compile hll files into cil files. (No data available).
> semodule: Failed!
> ...
For some reason, dnf install still succeeds, but when the selinux tests
fail with:
> chcon: failed to change context of '/tmp/bats-run-3MMyYP/runc.szTqBc/bundle/runc' to ‘system_u:object_r:container_runtime_exec_t:s0’: Invalid argument
All this is fixed once policycoreutils is added to the list of RPMS so
it is updated (from 3.9-3.el10 to 3.10-1.el10) during the same
transaction.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 67f6c37b ("ci/gha: switch to ubuntu 24.04") switched most GHA CI
to Ubuntu 24.04 except for one job. It says:
> Leave ubuntu-22.04 for ci/cross-i386 (issue with systemctl restart hang
> after apt install). This can be addressed separately later.
Assuming the issue it already fixed (updated systemd or something),
let's finalize the 24.04 switch.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Previously, when prepareCgroupFD would not open container's cgroup
(as configured in config.json and saved to state.json), it returned
a fatal error, as we presumed a container can't exist without its own
cgroup.
Apparently, it can. In a case when container is configured without
cgroupns (i.e. it uses hosts cgroups), and /sys/fs/cgroup is mounted
read-write, a rootful container's init can move itself to an entirely
different cgroup (even a new one that it just created), and then the
original container cgroup is removed by the kernel (or systemd?) as
it has no processes left. By the way, from the systemd point of view
the container is gone. And yet it is still there, and users want
runc exec to work!
And it worked, thanks to the "let's try container init's cgroup"
fallback as added by commit c91fe9aeba ("cgroup2: exec: join the
cgroup of the init process on EBUSY"). The fallback was added for
the entirely different reason, but it happened to work in this very
case, too.
This behavior was broken with the introduction of CLONE_INTO_CGROUP
support.
While it is debatable whether this is a valid scenario when a container
moves itself into a different cgroup, this very setup is used by e.g.
buildkitd running in a privileged kubernetes container (see issue 5089).
To restore the way things are expected to work, add the same "try
container init's cgroup" fallback into prepareCgroupFD.
While at it, simplify the code flow.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Refactor addIntoCgroupV2 in an attempt to simplify it.
2. Fix the bug of not trying the init cgroup fallback if
rootlessCgroup is set. This is a bug because rootlessCgroup
tells to ignore cgroup join errors, not to never try the fallback.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Separate initProcessCgroupPath code out of addIntoCgroupV2.
To be used by the next patch.
While at it, describe the new scenario in which the container's
configured cgroup might not be available.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When parsing mount options into recAttrSet and recAttrClr,
the code sets attr_clr to individual atime flags (e.g.
MOUNT_ATTR_NOATIME or MOUNT_ATTR_STRICTATIME) when clearing
atime attributes. However, this violates the kernel's
requirement documented in mount_setattr(2)[1]:
> Note that, since the access-time values are an enumeration
> rather than bit values, a caller wanting to transition to a
> different access-time setting cannot simply specify the
> access-time setting in attr_set, but must also include
> MOUNT_ATTR__ATIME in the attr_clr field. The kernel will
> verify that MOUNT_ATTR__ATIME isn't partially set in
> attr_clr (i.e., either all bits in the MOUNT_ATTR__ATIME
> bit field are either set or clear), and that attr_set
> doesn't have any access-time bits set if MOUNT_ATTR__ATIME
> isn't set in attr_clr.
Passing only a single atime flag (e.g. MOUNT_ATTR_RELATIME) in
attr_clr causes mount_setattr() to fail with EINVAL.
This change ensures that whenever an atime mode is updated,
attr_clr includes MOUNT_ATTR__ATIME to properly reset the
entire access-time attribute field before applying the new mode.
[1] https://man7.org/linux/man-pages/man2/mount_setattr.2.html
Signed-off-by: lifubang <lifubang@acmcoder.com>
Since [PR 4812], runc exec tries to use clone3 syscall with
CLONE_INTO_CGROUP, falling back to the old method if it is not
supported.
One issue with that approach is, a
> Cmd cannot be reused after calling its [Cmd.Start], [Cmd.Run],
> [Cmd.Output], or [Cmd.CombinedOutput] methods.
(from https://pkg.go.dev/os/exec#Cmd).
This is enforced since Go 1.26, see [CL 728642], and so runc exec
actually fails in specific scenarios (go1.26 and no CLONE_INTO_CGROUP
support).
The easiest workaround is to pre-copy the p.cmd structure (copy = *cmd).
From the [CL 734200] it looks like it is an acceptable way, but it might
break in the future as it also copies the private fields, so let's do a
proper field-by-field copy. If the upstream will add cmd.Clone method,
we will switch to it.
Also, we can probably be fine with a post-copy (once the first Start has
failed), but let's be conservative here and do a pre-copy.
[PR 4812]: https://github.com/opencontainers/runc/pull/4812
[CL 728642]: https://go.dev/cl/728642
[CL 734200]: https://go.dev/cl/734200
Reported-by: Efim Verzakov <efimverzakov@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Theoretically, exec.Command can set cmd.Err.
Practically, this should never happen (Linux, Go <= 1.26, exePath is
absolute), but in the unlikely case it does, let's fail early.
This is related to the cloneCmd (to be introduced by the following
commit) which chooses to not copy the Err field. Theoretically,
exec.Command can set Err and so the first call to cmd.Start will fail
(since Err != nil), and the second call to cmd.Start may succeed because
Err == nil. Yet, this scenario is highly unlikely, but better be safe
than sorry.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The Process type is quite big (currently 368 bytes on a 64 bit Linux)
and using non-pointer receivers in its methods results in copying which
is totally unnecessary.
Change the methods to use pointer receivers.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The Config type is quite big (currently 554 bytes on a 64 bit Linux)
and using non-pointer receivers in its methods results in copying which
is totally unnecessary.
Change the methods to use pointer receivers.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Rename a function parameter (containerId -> containerID) to avoid a
linter warning:
> var-naming: method parameter containerId should be containerID (revive)
In many other places, including config.json (.linux.uidMappings and
.gidMappings) it is already called containerID, so let's rename.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently Write (and WriteString) must return an error (apparently
io.ErrShortWrite) on short writes (see [1], [2]), so no explicit check
for a short write is needed.
While at it, use (*os.File).WriteString directly rather than
io.WriteString.
[1]: https://pkg.go.dev/os#File.Write
[2]: https://pkg.go.dev/io#Writer
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The modernize documentation used to suggest -test flag but it's not
needed as it is enabled by default. Drop it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since we use modernize@latest, it may require latest Go as well (and now it does),
so use "go-version: stable" explicitly (which resolves to latest Go).
This fixes the issue with CI:
> go: golang.org/x/tools/gopls/internal/analysis/modernize/cmd/modernize@latest: golang.org/x/tools/gopls@v0.21.0 requires go >= 1.25 (running go 1.24.11; GOTOOLCHAIN=local)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Alas, these new constants are already in v1.4.0 release so we can't
remove those right away, but we can mark them as deprecated now
and target removal for v1.5.0.
So,
- mark them as deprecated;
- redefine via unix.MPOL_* counterparts;
- fix the validator code to use unix.MPOL_* directly.
This amends commit a0e809a8.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It appears that when we import github.com/coreos/go-systemd/activation,
it brings in the whole crypto/tls package (which is not used by runc
directly or indirectly), making the runc binary size larger and
potentially creating issues with FIPS compliance.
Let's copy the code of function we use from go-systemd/activation
to avoid that.
The space savings are:
$ size runc.before runc.after
text data bss dec hex filename
7101084 5049593 271560 12422237 bd8c5d runc.before
6508796 4623281 229128 11361205 ad5bb5 runc.after
Reported-by: Dimitri John Ledkov <dimitri.ledkov@surgut.co.uk>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In view of recent criu-dev failure, let's not fail the
required "all-done" job when criu-dev tests fail.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This function calls Init what normally never returns, so the defer only
works if there is an error and we can safely use it to close those fds
we opened. This was done for most but not all fds.
Reported in issue 5008.
Reported-by: Arina Cherednik <arinacherednik034@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Because we add the runc-$version/ prefix to the archive we generate,
including the version in the name makes it easier for some tools to
operate on as it matches most other projects (for openSUSE we rename the
archive file to this format in order for the automated RPM scripts to
work properly).
Also, when doing several releases at the same time, being able to
double-check that the correct artefact versions were uploaded for each
release can be quite handy.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Now that runc 1.4.0 has been released, there is no need to single out
1.1.x and earlier as no longer being supported, as latest-2 is now 1.2.x
and thus 1.1.x would no longer be supported even with the new support
model.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We have used this release policy for a year and it seems to work well
for everyone and we haven't received much feedback, so it seems
reasonable to say that we are committed to this policy now.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We intentionally broke this in commit d40b3439a9 ("rootfs: switch to
fd-based handling of mountpoint targets") under the assumption that most
users do not need this feature. Sadly it turns out they do, and so
commit 3f925525b4 ("rootfs: re-allow dangling symlinks in mount
targets") added a hotfix to re-add this functionality.
This patch adds some much-needed tests for this behaviour, since it
seems we are going to need to keep this for compatibility reasons (at
least until runc v2...).
Co-developed-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While CreateInRoot supports hallucinating the target path, we do not use
it directly when constructing device inode targets because we need to
have different handling for mknod and bind-mounts.
The solution is to simply have a more generic MkdirAllParentInRoot
helper that MkdirAll's the parent directory of the target path and then
allows the caller to create the trailing component however they like.
(This can be used by CreateInRoot internally as well!)
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In order to maintain compatibility with previous releases of runc (which
permitted dangling symlinks as path components by permitting
non-existent path components to be treated like real directories) we
have to first do SecureJoin to construct a target path that is
compatible with the old behaviour but has all dangling symlinks (or
other invalid paths like ".." components after non-existent directories)
removed.
This is effectively a more generic verison of commit 3f925525b4
("rootfs: re-allow dangling symlinks in mount targets") and will let us
remove the need for open-coding SecureJoin workarounds.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Now that MkdirAllInRoot has been removed, we can make MkdirAllInRootOpen
less wordy by renaming it to MkdirAllInRoot. This is a non-functional
change.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This probably should've been done as part of commit d40b3439a9
("rootfs: switch to fd-based handling of mountpoint targets") but it
seems I missed them when doing the rest of the conversions.
This also lets us remove utils.WithProcfd entirely, as well as
pathrs.MkdirAllInRoot. Unfortunately, WithProcfd was exposed in the
externally-importable "libcontainer/utils" package and so we need to
have a deprecation notice to remove it in runc 1.5.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These helpers will be needed for the compatibility code added in future
patches in this series, but because "internal/pathrs" is imported by
"libcontainer/utils" we need to move them so that we can avoid circular
dependencies.
Because the old functions were in a non-internal package it is possible
some downstreams use them, so add some wrappers but mark them as
deprecated.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In case early stage of runc init (nsenter) fails for some reason, it
logs error(s) with FATAL log level, via bail().
The runc init log is read by a parent (runc create/run/exec) and is
logged via normal logrus mechanism, which is all fine and dandy, except
when `runc init` fails, we return the error from the parent (which is
usually not too helpful, for example):
runc run failed: unable to start container process: can't get final child's PID from pipe: EOF
Now, the actual underlying error is from runc init and it was logged
earlier; here's how full runc output looks like:
FATA[0000] nsexec-1[3247792]: failed to unshare remaining namespaces: No space left on device
FATA[0000] nsexec-0[3247790]: failed to sync with stage-1: next state
ERRO[0000] runc run failed: unable to start container process: can't get final child's PID from pipe: EOF
The problem is, upper level runtimes tend to ignore everything except
the last line from runc, and thus error reported by e.g. docker is not
very helpful.
This patch tries to improve the situation by collecting FATAL errors
from runc init and appending those to the error returned (instead of
logging). With it, the above error will look like this:
ERRO[0000] runc run failed: unable to start container process: can't get final child's PID from pipe: EOF; runc init error(s): nsexec-1[141549]: failed to unshare remaining namespaces: No space left on device; nsexec-0[141547]: failed to sync with stage-1: next state
Yes, it is long and ugly, but at least the upper level runtime will
report it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since the recent CVE fixes, TestFdLeaksSystemd sometimes fails:
=== RUN TestFdLeaksSystemd
exec_test.go:1750: extra fd 9 -> /12224/task/13831/fd
exec_test.go:1753: found 1 extra fds after container.Run
--- FAIL: TestFdLeaksSystemd (0.10s)
It might have been caused by the change to the test code in commit
ff6fe13 ("utils: use safe procfs for /proc/self/fd loop code") -- we are
now opening a file descriptor during the logic to get a list of file
descriptors. If the file descriptor happens to be allocated to a
different number, you'll get an error.
Let's try to filter out the fd used to read a directory.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
On some systems (e.g., AlmaLinux 8), systemd automatically removes cgroup paths
when they become empty (i.e., contain no processes). To prevent this, we spawn
a dummy process to pin the cgroup in place.
Fix: https://github.com/opencontainers/runc/issues/5003
Signed-off-by: lifubang <lifubang@acmcoder.com>
This is to ensure that Go version in Dockerfile (which is used to build
release binaries) is:
- currently supported;
- used in CI tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For some reason, some jobs in .github/workflows/validate.yml
have "fetch-depth: 0" argument to actions/checkout, meaning
"all history for all branches and tags". Obviously this is
not needed here.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The main update is actually in github.com/opencontainers/cgroups, but we
need to also update runtime-spec to a newer pre-release version to get
the updates from there as well.
In short, the behaviour change is now that "0" is treated as a valid
value to set in "pids.max", "-1" means "max" and unset/nil means "do
nothing". As described in the opencontainers/cgroups PR, this change is
actually backwards compatible because our internal state.json stores
PidsLimit, and that entry is marked as "omitempty". So, an old runc
would omit PidsLimit=0 in state.json, and this will be parsed by a new
runc as being "nil" -- and both would treat this case as "do not set
anything".
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Commit b2f8a74d "clothed" the naked return as inflicted by gofumpt
v0.9.0. Since gofumpt v0.9.2 this rule was moved to "extra" category,
not enabled by default. The only other "extra" rule is to group adjacent
parameters with the same type, which also makes sense.
Enable gofumpt "extra" rules, and reformat the code accordingly.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is mostly a mechanical change, but we also need to change some
types to match the "mode int" argument that golang.org/x/sys/unix
decided to use.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This was always the intended behaviour but commit 72fbb34f50 ("rootfs:
switch to fd-based handling of mountpoint targets") regressed it when
adding a mechanism to create a file handle to the target if it didn't
already exist (causing the later stat to always succeed).
A lot of people depend on this functionality, so add some tests to make
sure we don't break it in the future.
Fixes: 72fbb34f50 ("rootfs: switch to fd-based handling of mountpoint targets")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This new version includes the fixes for CVE-2025-52881, so we can remove
the internal/third_party copy of the library we added in commit
ed6b1693b8 ("selinux: use safe procfs API for labels") as well as the
"replace" directive in go.mod (which is problematic for "go get"
installs).
Fixes: ed6b1693b8 ("selinux: use safe procfs API for labels")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This will result in slower runs but we are having issues with
golangci-lint (false positives) that are most probably related
to caching.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
On MIPS arches, Rdev is uint32 so we have to convert it.
Fixes issue 4962.
Fixes: 8476df83 ("libct: add/use isDevNull, verifyDevNull")
Fixes: de87203e ("console: verify /dev/pts/ptmx before use")
Fixes: 398955bc ("console: add fallback for pre-TIOCGPTPEER kernels")
Reported-by: Tianon Gravi <admwiggin@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Aleksa Sarai (21):
rootfs: re-allow dangling symlinks in mount targets
openat2: improve resilience on busy systems
selinux: use safe procfs API for labels
rootfs: switch to fd-based handling of mountpoint targets
libct/system: use securejoin for /proc/$pid/stat
init: use securejoin for /proc/self/setgroups
init: write sysctls using safe procfs API
utils: remove unneeded EnsureProcHandle
utils: use safe procfs for /proc/self/fd loop code
apparmor: use safe procfs API for labels
ci: add lint to forbid the usage of os.Create
rootfs: avoid using os.Create for new device inodes
internal: add wrappers for securejoin.Proc*
go.mod: update to github.com/cyphar/filepath-securejoin@v0.5.0
console: verify /dev/pts/ptmx before use
console: avoid trivial symlink attacks for /dev/console
console: add fallback for pre-TIOCGPTPEER kernels
console: use TIOCGPTPEER when allocating peer PTY
*: switch to safer securejoin.Reopen
internal: move utils.MkdirAllInRoot to internal/pathrs
internal/sys: add VerifyInode helper
Li Fubang (1):
libct: align param type for mountCgroupV1/V2 functions
Kir Kolyshkin (3):
libct: maskPaths: don't rely on ENOTDIR for mount
libct: maskPaths: only ignore ENOENT on mount dest
libct: add/use isDevNull, verifyDevNull
Fixes: CVE-2025-31133 GHSA-9493-h29p-rfm2
Fixes: CVE-2025-52565 GHSA-qw9x-cqr3-wc7r
Fixes: CVE-2025-52881 GHSA-cgrx-mc8f-2prm
Reported-by: Lei Wang <ssst0n3@gmail.com>
Reported-by: Li Fubang <lifubang@acmcoder.com>
Reported-by: Tõnis Tiigi <tonistiigi@gmail.com>
Reported-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It seems there are a fair few images where dangling symlinks are used as
path components for mount targets, which pathrs-lite does not support
(and it would be difficult to fully support this in a race-free way).
This was actually meant to be blocked by commit 63c2908164 ("rootfs:
try to scope MkdirAll to stay inside the rootfs"), followed by commit
dd827f7b71 ("utils: switch to securejoin.MkdirAllHandle"). However, we
still used SecureJoin to construct mountpoint targets, which means that
dangling symlinks were "resolved" before reaching pathrs-lite.
This patch basically re-adds this hack in order to reduce the breakages
we've seen so far.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Previously, we would see a ~3% failure rate when starting containers
with mounts that contain ".." (which can trigger -EAGAIN). To counteract
this, filepath-securejoin v0.5.1 includes a bump of the internal retry
limit from 32 to 128, which lowers the failure rate to 0.12%.
However, there is still a risk of spurious failure on regular systems.
In order to try to provide more resilience (while avoiding DoS attacks),
this patch also includes an additional retry loop that terminates based
on a deadline rather than retry count. The deadline is 2ms, as my
testing found that ~800us for a single pathrs operation was the longest
latency due to -EAGAIN retries, and that was an outlier compared to the
more common ~400us latencies -- so 2ms should be more than enough for
any real system.
The failure rates above were based on more 50k runs of runc with an
attack script (from libpathrs) running a rename attack on all cores of a
16-core system, which is arguably a worst-case but heavily utilised
servers could likely approach similar results.
Tested-by: Phil Estes <estesp@gmail.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Due to the sensitive nature of these fixes, it was not possible to
submit these upstream and vendor the upstream library. Instead, this
patch uses a fork of github.com/opencontainers/selinux, branched at
commit opencontainers/selinux@879a755db5.
In order to permit downstreams to build with this patched version, a
snapshot of the forked version has been included in
internal/third_party/selinux. Note that since we use "go mod vendor",
the patched code is usable even without being "go get"-able. Once the
embargo for this issue is lifted we can submit the patches upstream and
switch back to a proper upstream go.mod entry.
Also, this requires us to temporarily disable the CI job we have that
disallows "replace" directives.
Fixes: GHSA-cgrx-mc8f-2prm CVE-2025-52881
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
An attacker could race with us during mount configuration in order to
trick us into mounting over an unexpected path. This would bypass
checkProcMount() and would allow for security profiles to be left
unapplied by mounting over /proc/self/attr/... (or even more serious
outcomes such as killing the entire system by tricking runc into writing
strings to /proc/sysrq-trigger).
This is a larger issue with our current mount infrastructure, and the
ideal solution would be to rewrite it all to be fd-based (which would
also allow us to support the "new" mount API, which also avoids a bunch
of other issues with mount(8)). However, such a rewrite is not really
workable as a security fix, so this patch is a bit of a compromise
approach to fix the issue while also moving us a bit towards that
eventual end-goal.
The core issue in CVE-2025-52881 is that we currently use the (insecure)
SecureJoin to re-resolve mountpoint target paths multiple times during
mounting. Rather than generating a string from createMountpoint(), we
instead open an *os.File handle to the target mountpoint directly and
then operate on that handle. This will make it easier to remove
utils.WithProcfd() and rework mountViaFds() in the future.
The only real issue we need to work around is that we need to re-open
the mount target after doing the mount in order to get a handle to the
mountpoint -- pathrs.Reopen() doesn't work in this case (it just
re-opens the inode under the mountpoint) so we need to do a naive
re-open using the full path. Note that if we used move_mount(2) this
wouldn't be a problem because we would have a handle to the mountpoint
itself.
Note that this is still somewhat of a temporary solution -- ideally
mountViaFds would use *os.File directly to let us avoid some other
issues with using bare /proc/... paths, as well as also letting us more
easily use the new mount API on modern kernels.
Fixes: GHSA-cgrx-mc8f-2prm CVE-2025-52881
Co-developed-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
sysctls could in principle also be used as a write gadget for arbitrary
procfs files. As this requires getting a non-subset=pid /proc handle we
amortise this by only allocating a single procfs handle for all sysctl
writes.
Fixes: GHSA-cgrx-mc8f-2prm CVE-2025-52881
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
All of the callers of EnsureProcHandle now use filepath-securejoin's
ProcThreadSelf to get a file handle, which has much stricter
verification to avoid procfs attacks than EnsureProcHandle's very
simplistic filesystem type check.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
From a safety perspective this might not be strictly required, but it
paves the way for us to remove utils.ProcThreadSelf.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
EnsureProcHandle only protects us against a tmpfs mount, but the risk of
a procfs path being used (such as /proc/self/sched) has been known for a
while. Now that filepath-securejoin has a reasonably safe procfs API,
switch to it.
Fixes: GHSA-cgrx-mc8f-2prm CVE-2025-52881
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
os.Create is shorthand for open(O_CREAT|O_TRUNC) *without* O_EXCL, which
is incredibly unsafe for us to do when interacting with a container
rootfs (especially before pivot_root) as an attacker could swap the
target path with a symlink that points to the host filesystem, causing
us to delete the contents of or create host files.
We did have a similar bug in CVE-2024-45310, but in that case we
(luckily) didn't have O_TRUNC set which avoided the worst possible case.
However, os.Create does set O_TRUNC and we were using it in scenarios
that may have been exploitable.
Because of how easy it us for us to accidentally introduce this kind of
bug, we should simply not allow the usage of os.Create in our entire
codebase.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If an attacker were to make the target of a device inode creation be a
symlink to some host path, os.Create would happily truncate the target
which could lead to all sorts of issues. This exploit is probably not as
exploitable because device inodes are usually only bind-mounted for
rootless containers, which cannot overwrite important host files (though
user files would still be up for grabs).
The regular inode creation logic could also theoretically be tricked
into changing the access mode and ownership of host files if the
newly-created device inode was swapped with a symlink to a host path.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In order to avoid lint errors due to the deprecation of the top-level
securejoin methods ported from libpathrs, we need to adjust
internal/pathrs to use the new pathrs-lite subpackage instead.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This is primarily done out of an abudance of caution against runc exec
being attacked by a container where /dev/pts/ptmx has been replaced with
some other bad inode (a disconnected NFS handle, a symlink that goes
through a leaked runc file descriptor to reference a host ptmx, etc).
Unfortunately, we cannot trivially verify that /dev/pts/ptmx is actually
the /dev/pts from the container without storing stuff like the fsid in
the runc state.json, which is probably not worth the extra effort. This
should at least avoid the most concerning cases.
Reported-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
An attacker could make /dev/console a symlink. This presents two
possible issues:
1. os.Create will happily truncate targets, which could have resulted
in a worse version of CVE-2024-4531. Luckily, this all happens after
pivot_root(2) so the scope of that particular attack is fairly
limited (you are unlikely to be able to easily access host rootfs
files -- though it might be possible to take advantage of leaks such
as in CVE-2024-21626). However, O_CREAT|O_NOFOLLOW is what we should
be doing for all file creations.
2. Because we passed /dev/console as the only mount path (as opposed to
using a /proc/self/fd/$n path), an attacker could swap the symlink
to point to any other path and thus cause us to mount over some
other path. This is not as big of a problem because all the mounts
are in the container namespace after pivot_root(2), and users
usually can create arbitrary mount targets inside the container.
These issues don't seem particularly exploitable, but they deserve to be
hardened regardless.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The pty driver has very consistent allocation rules for the major:minor
numbers of /dev/pts/$n inodes, so it is possible to somewhat safely open
/dev/pts/* paths if we validate that the inode is the one we expect.
It is possible for an attacker to have over-mounted a pts peer from a
different devpts instance, but to fix this would require more tracking
of devpts instances than runc currently can do.
This means runc should continue to work on very old kernels.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
When opening the peer end of a pty, the old kernel API required us to
open /dev/pts/$num inside the container (at least since we fixed console
handling many years ago in commit 244c9fc426 ("*: console rewrite")).
The problem is that in a hostile container it is possible for
/dev/pts/$num to be an attacker-controlled symlink that runc can be
tricked into resolving when doing bind-mounts. This allows the attacker
to (among other things) persist /proc/... entries that are later masked
by runc, allowing an attacker to escape through the kernel.core_pattern
sysctl (/proc/sys/kernel/core_pattern). This is the original issue
reported by Lei Wang and Li Fu Bang in CVE-2025-52565.
However, it should be noted that this is not entirely a newly-discovered
problem. Way back in Linux 4.13 (2017), I added the TIOCGPTPEER ioctl,
which allows us to get a pty peer without touching the /dev/pts inside
the container. The original threat model was around an attacker
replacing /dev/pts/$n or /dev/pts/ptmx with some malicious inode (a DoS
inode, or possibly a PTY they wanted a confused deputy to operate on).
Unfortunately, there was no practical way for runc to cache a safe
O_PATH handle to /dev/pts/ptmx (unlike other runtimes like LXC, which
switched to TIOCGPTPEER way back in 2017). Since it wasn't clear how we
could protect against the main attack TIOCGPTPEER was meant to protect
against, we never switched to it (even though I implemented it
specifically to harden container runtimes).
Unfortunately, It turns out that mount *sources* are a threat we didn't
fully consider. Since TIOCGPTPEER already solves this problem entirely
for us in a race free way, we should just use that. In a later patch, we
will add some hardening for /dev/pts/$num opening to maintain support
for very old kernels (Linux 4.13 is very old at this point, but RHEL 7
is still kicking and is stuck on Linux 3.10).
Fixes: GHSA-qw9x-cqr3-wc7r CVE-2025-52565
Reported-by: Lei Wang <ssst0n3@gmail.com> (CVE-2025-52565)
Reported-by: lfbzhm <lifubang@acmcoder.com> (CVE-2025-52565)
Reported-by: Aleksa Sarai <cyphar@cyphar.com> (TIOCGPTPEER)
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Currently, we rely on mount returning ENOTDIR when the destination is a
directory (and so mount tells us that the source is not), and fall back
to read-only tmpfs bind mount for such cases.
Theoretically, ENOTDIR can also be returned in some other cases,
resulting in the wrong type of mount being used.
Let's be more straightforward here -- call fstat on destination file
descriptor, and use the proper mount depending on whether it is a
directory.
Reported-by: Rodrigo Campos <rodrigoca@microsoft.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When mounting a path being masked, the /dev/null might disappear from
under us, and mount (even on an opened /dev/null file descriptor) will
return ENOENT, which we deliberately ignore, as there's no need to mask
non-existent paths.
Let's open the destination path and ignore ENOENT during open, then
mount via the destination file descriptor, not ignoring ENOENT.
Reported-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The /dev/null in a container should not be trusted, because when /dev
is a bind mount, /dev/null is not created by runc itself.
1. Add isDevNull which checks the fd minor/major and device type,
and verifyDevNull which does the stat and the check.
2. Rewrite maskPath to open and check /dev/null, and use its fd to
perform mounts. Move the loop over the MaskPaths into the function,
and rename it to maskPaths.
3. reOpenDevNull: use verifyDevNull and isDevNull.
4. fixStdioPermissions: use isDevNull instead of stat.
Fixes: GHSA-9493-h29p-rfm2 CVE-2025-31133
Co-authored-by: Rodrigo Campos <rodrigoca@microsoft.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
filepath-securejoin v0.3 gave us a much safer re-open primitive, we
should use it to avoid any theoretical attacks. Rather than using it
direcly, add a small pathrs wrapper to make libpathrs migrations in the
future easier...
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We will have more wrappers around filepath-securejoin, and so move them
to their own specific package so that we can eventually use libpathrs
fairly cleanly (by swapping out the implementation).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This will be used for a few security patches in later patches in this
patchset. The need to verify what kind of inode we are operating on in a
race-free way turns out to be quite a common pattern...
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Introduce and use iobail, xread, and xwrite wrappers so that we can
properly check read/write return value and call either bail or bailx on
error, with proper diagnostics (distinguishing failed read/write from a
short read/write).
This prevents the "Success" prefix in errors like:
failed to sync with stage-1: next state: Success
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add a few missing sane_kill calls where they make sense.
Remove one useless sane_kill of stage2_pid, as during SYNC_USERMAP stage2
is not yet started. It is harmless yet it makes the code slightly harder
to read.
Set the child pid to -1 upon receiving SYNC_CHILD_FINISH
to minimize the chances of killing an unrelated process.
When a child sends SYNC_CHILD_FINISH it is about to exit
(although theoretically it could be stuck during debug logging).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We use bail to report fatal errors, and bail always append %m
(aka strerror(errno)). In case an error condition did not set
errno, the log message will end up with ": Success" or an error
from a stale errno value. Either case is confusing for users.
Introduce bailx which is the same as bail except it does not
append %m, and use it where appropriate.
The naming follows libc's err(3) and errx(3).
PS we still use bail in a few cases after read or write, even
if that read/write did not return an error, because the code
does not distinguish between short read/write and error (-1).
This will be addressed by the next commit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since sane_kill after a failed read or write, but before reporting the
error from that read or write, it may change the errno value in case
kill(2) fails.
Save and restore the errno around the call to kill.
While at it,
- change the code to return early;
- don't return kill return value as no one is using it, and the errno
value no longer correlates.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The commit mentioned below has missed these changes.
Fixes: 17570625 ("Use for range over integers")
Signed-off-by: Ariel Otilibili <otilibil@eurecom.fr>
This is mostly to improve readability. While at it, make the script more
robust by adding -e option to shell. The exception is echo $pid which is
opportunistic and may fail depending on the order of pids in the file.
Also, remove the empty comment and a shellcheck annotation.
Fixes: c91fe9ae
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The "runc delete --force [paused container]" test case does not check
runc pause exit code, and if added, the test fails in rootless tests,
because:
- not all rootless tests have access to cgroups;
- rootless containers doesn't have default cgroups path.
To fix, add:
- setup for rootless case;
- require cgroups_freezer;
- runc pause exit code check.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In our bats tests, runc itself is a wrapper which calls bats run helper,
so using "run runc" is wrong as it results in calling run helper twice.
Fixes: 8d180e965
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commands that are not run via "run" helper (cat, mkdir, __runc)
do not set $status, so it makes no sense to check it.
Fixes: 94505a04, ed548376
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a bit opinionated, but some comments in integration tests do not
really help to understand the nature of the tests being performed by
stating something very obvious, like
# run busybox detached
runc run -d busybox
To make things worse, these not-so-helpful messages are being
copy/pasted over and over, and that is the main reason to remove them.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Remove the devicemapper driver mentions, and is it no longer
supported by docker (or podman).
2. Remove the test example -- we have plenty of real ones.
3. Add a link to (well written and extensive) bats documentation.
4. Fix capitalization in a sentence.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Bump bats to the version from Fedora 42 (used in "fedora" job), so we
have the same version everywhere.
This also fixes an issue introduced by commit d31e6b87 (which forgot to
bump bats in GHA CI), and adds a note to the yaml in order to avoid the
same issue in the future.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
All the new code appears in main (not in the release branches),
and we only want extra linter rules to apply to new code.
Disable lint-extra job if the PR is not to the main branch.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Factor startWithCgroupFD out of start to reduce the start complexity.
This also implements a more future-proof way of calling p.comm.closeChild.
Co-authored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The (*setns).start is supposed to close child fds once the child has
started, or upon an error. Commit 5af4dd4e6 added a bug -- child fds
are not closed if prepareCgroupFD fails.
Fix by adding a missing call to closeChild.
I'm not sure how to write a good test case for it. Found when working
on PR 4928 (and tested in there).
Fixes: 5af4dd4e6
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
strings.Title is deprecated since Go 1.18. Replace it with a simple
manual capitalization of the first character in criuNsToKey().
Signed-off-by: Osama Abdelkader <osama.abdelkader@gmail.com>
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now that we've updated to golang.org/x/sys@v0.37.0, CPUSet has a Fill
helper that does the equivalent to our underflow trick to make setting
all CPUs efficient.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This package was marked deprecated in commit 9b60a93cf3
("libcontainer/userns: migrate to github.com/moby/sys/userns"), which
was included in runc 1.2. Users have thus had a year to migrate to
github.com/moby/sys/userns and it's okay for us to remove this wrapper
package.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
ba0b5e26 ("libcontainer: remove all mount logic from nsexec") introduced
a request function that handles two tasks:
- the exchanges with the request and response channels
- the closing of the request channel.
From 1.21, the closing of the request channel may be done with
context.AfterFunc(). Moreover, context.AfterFunc() is guaranteed to run
once.
Link: https://pkg.go.dev/context#AfterFunc
Suggested-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Ariel Otilibili <otilibil@eurecom.fr>
This removes `mips64le` (no longer supported by the image / upstream in Debian Trixie+) and adds `riscv64`.
Signed-off-by: Tianon Gravi <admwiggin@gmail.com>
Set the process personality early to ensure it takes effect before
seccomp is initialized. If seccomp filters are applied first and they
block personality-related system calls (e.g., `personality(2)`),
subsequent attempts to set the personality will fail.
Signed-off-by: lifubang <lifubang@acmcoder.com>
The linux.intelRdt.enableMonitoring field enables the creation of
a per-container monitoring group. The monitoring group is removed when
the container is destroyed.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
The main benefit here is when we are using a systemd cgroup driver,
we actually ask systemd to add a PID, rather than doing it ourselves.
This way, we can add rootless exec PID to a cgroup.
This requires newer opencontainers/cgroups and coreos/go-systemd.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Remove cgroupPaths field from struct setnsProcess, because:
- we can get base cgroup paths from p.manager.GetPaths();
- we can get sub-cgroup paths from p.process.SubCgroupPaths.
But mostly because we are going to need separate cgroup paths when
adopting cgroups.AddPid.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes the following warning (seen on Fedora 42 and Ubuntu 24.04):
+ sudo chown -R rootless.rootless /home/rootless
chown: warning: '.' should be ':': ‘rootless.rootless’
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When a non–page-aligned value is written to memory.max, the kernel aligns it
down to the nearest page boundary. On systems with a page size greater
than 4K (e.g., 64K), this caused failures because the configured
memory.max value was not 64K aligned.
This patch fixes the issue by explicitly aligning the memory.max value
to 64K. Since 64K is also a multiple of 4K, the value is correctly
aligned on both 4K and 64K page size systems.
However, this approach will still fail on systems where the hardcoded
memory.max value is not aligned to the system page size.
Fixes: https://github.com/opencontainers/runc/issues/4841
Signed-off-by: Vishal Chourasia <vishalc@linux.ibm.com>
Signed-off-by: Donet Tom <donettom@linux.ibm.com>
Implement support for the linux.intelRdt.schemata field of the spec.
This allows management of the "schemata" file in the resctrl group in a
generic way.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
These sysctls are all per-userns (termed `ucounts` in the kernel code) are
settable with CAP_SYS_RESOURCE in the user namespace.
Signed-off-by: Tycho Andersen <tycho@tycho.pizza>
libcontainer/user was marked as deprecated in d9ea71bf96 ("deprecate
libcontainer/user") and users have had plenty of time to migrate to
github.com/moby/sys/user.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Modernize tool [1] basically ensures that the new language features and
packages are used across the code.
The reason to run it in CI is to ensure that
- PR authors use modern code;
- our code is modern whether we bump Go version in go.mod.
Shove it into go-fix job which already does a similar thing
but for 'go-fix' and rename the whole job to modernize.
[1]: https://pkg.go.dev/golang.org/x/tools/gopls/internal/analysis/modernize
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In these cases, omitempty doesn't really work so it is useless,
but omitzero actually works.
As a result, output of `runc events` may omit these fields if all they
contain are zeroes.
NOTE this might be a breaking change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now that Go 1.25 is out, let's switch to go 1.24.0 as a minimally
supported version, drop Go 1.23 and add Go 1.25 to CI matrix.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. In case runc binary file name is not runc, the test fails like
below. The fix is to get the binary name from $RUNC.
✗ runc command -h
(in test file tests/integration/help.bats, line 27)
`[[ ${lines[1]} =~ runc\ checkpoint+ ]]' failed
runc-go1.25.0-main checkpoint -h (status=0):
NAME:
runc-go1.25.0-main checkpoint - checkpoint a running container
2. Simplify the test by adding a loop for all commands. While at it, add
a loop for -h --help as well.
3. Add missing commands (create, ps, features).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The setup in selinux.bats assumes $RUNC binary name ends in runc, and
thus it fails when we run it like this:
sudo -E RUNC=$(pwd)/runc.patched bats tests/integration/selinux.bats
Fix is easy.
Fixes: b39781b06 ("tests/int: add selinux test case")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In certain deployments, it's possible for runc to be spawned by a
process with a restrictive cpumask (such as from a systemd unit with
CPUAffinity=... configured) which will be inherited by runc and thus the
container process by default.
The cpuset cgroup used to reconfigure the cpumask automatically for
joining processes, but kcommit da019032819a ("sched: Enforce user
requested affinity") changed this behaviour in Linux 6.2.
The solution is to try to emulate the expected behaviour by resetting
our cpumask to correspond with the configured cpuset (in the case of
"runc exec", if the user did not configure an alternative one). Normally
we would have to parse /proc/stat and /sys/fs/cgroup, but luckily
sched_setaffinity(2) will transparently convert an all-set cpumask (even
if it has more entries than the number of CPUs on the system) to the
correct value for our usecase.
For some reason, in our CI it seems that rootless --systemd-cgroup
results in the cpuset (presumably temporarily?) being configured such
that sched_setaffinity(2) will allow the full set of CPUs. For this
particular case, all we care about is that it is different to the
original set, so include some special-casing (but we should probably
investigate this further...).
Reported-by: ningmingxiao <ning.mingxiao@zte.com.cn>
Reported-by: Martin Sivak <msivak@redhat.com>
Reported-by: Peter Hunt <pehunt@redhat.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Sometimes we need to run runc through some wrapper (like nohup), but
because "__runc" and "runc" are bash functions in our test suite this
doesn't work trivially -- and you cannot just pass "$RUNC" because you
you need to set --root for rootless tests.
So create a setup_runc_cmdline helper which sets $RUNC_CMDLINE to the
beginning cmdline used by __runc (and switch __runc to use that).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
"runc" was a special wrapper around bats's "run" which output some very
useful diagnostic information to the bats log, but this was not usable
for other commands. So let's make it a more generic helper that we can
use for other commands.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Currently, criu package from opensuse build farm times out on GHA arm,
so let's only use criu-dev (i.e. compiled from source on CI machine).
Once this is fixed, this patch can be reverted.
Related to criu issue 2709.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since GHA now provides ARM, we can switch away from actuated.
Many thanks to @alexellis (@self-actuated) for being the sponsor of this
project.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
openSUSE has an unfortunate default udev setup which forcefully sets all
loop devices to use the "none" scheduler, even if you manually set it.
As this is a property of the host configuration (and udev is monitoring
from the host) we cannot really change this behaviour from inside our
test container.
So we should just skip the test in this (hopefully unusual) case.
Ideally tools running the test suite should disable this behaviour when
running our test suite.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If an error occurs during a test which sets up loopback devices, the
loopback device is not freed. Since most systems have very conservative
limits on the number of loopback devices, re-running a failing test
locally to debug it often ends up erroring out due to loopback device
exhaustion.
So let's just move the "losetup -d" to teardown, where it belongs.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Also, use GetPath() in Apply to get the resctrl group path, similar to
other methods of intelRdtManager.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
Error out --l3-cache-schema and --mem-bw-schema if the original
spec didn't specify intelRdt which also means that no CLOS (resctrl
group) was created for the container.
This prevents serious issues in this corner case.
First, a CLOS was created but the schemata of the CLOS was not
correctly updated. Confusingly, calling runc update twice
did the job: the first call created the resctrl group and the seccond
call was able to update the schemata. This issue would be relatively
easily fixable, though.
Second, more severe issue is that creating new CLOSes this way caused
them to be orphaned, not being removed when the container exists. This
is caused by runc not capturing the updated state (original spec was
intelRdt=nil -> no CLOS but after update this is not the case).
The most severe problem is that the update only move (or tried to move)
the original init process pid but all children escaped the update. Doing
this (i.e. migrating all processes of a container from CLOS to another
CLOS) reliably, race-free, would probably require freezing the
container.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
If intelRdt is specified in the spec, check that the resctrl fs is
actually mounted. Fixes e.g. the case where "intelRdt.closID" is
specified but runc silently ignores this if resctrl is not mounted.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
Prevent --l3-cache-schema from clearing the intel_rdt.memBwSchema state
and --mem-bw-schema clearing l3_cache_schema, respectively.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
While it does not make much sense practically, as runc is going to exit
soon and all fds will be closed anyway, various linters (including
SVACE) keep reporting this.
Let's make them happy.
Reported-by: Tigran Sogomonian <tsogomonian@astralinux.ru>
Reported-by: Mikhail Dmitrichenko <m.dmitrichenko222@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was added in 2ee9cbbd12 ("It's /proc/stat, not /proc/stats") with
no actual justification, and doesn't really make much sense on further
inspection:
* /proc/net is a symlink to "self/net", which means that /proc/net/dev
is a per-process file, and so overmounting it would only affect pid1.
Any other program that cares about /proc/net/dev would see their own
process's configuration, and unprivileged processes wouldn't be able
to see /proc/1/... data anyway.
In addition, the fact that this is a symlink means that runc will
deny the overmount because /proc/1/net/dev is not in the proc
overmount allowlist. This means that this has not worked for many
years, and probably never worked in the first place.
* /proc/self/net is already namespaced with network namespaces, so the
primary argument for allowing /proc overmounts (lxcfs-like masking of
procfs files to emulate namespacing for files that are not properly
namespaced for containers -- such as /proc/cpuinfo) is moot.
It goes without saying that lxcfs has never overmounted
/proc/self/net/... files, so the general "because lxcfs"
justification doesn't hold water either.
* The kernel has slowly been moving towards blocking overmounts in
/proc/self/. Linux 6.12 blocked overmounts for fd, fdinfo, and
map_files; future Linux versions will probably end up blocking
everything under /proc/self/.
Fixes: 2ee9cbbd12 ("It's /proc/stat, not /proc/stats")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Apparently, having a minor of 0 does not always mean it's the
whole device (not a partition):
=== /proc/partitions (using major: 259) ===
major minor #blocks name
8 16 78643200 sdb
8 17 77593583 sdb1
8 30 4096 sdb14
8 31 108544 sdb15
259 0 934912 sdb16
8 0 78643200 sda
8 1 78641152 sda1
Rewrite the test to not assume minor is 0, and use
lsblk -d to find out whole devices.
This fixes a test case which was added in commit 7696402da.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The dmem controller is added into kernel v6.13 and is now enabled in
Fedora 42 kernels. Yet, systemd is not aware of dmem.
This fixes the test case failure on Fedora.
For the initial test case, see commit 27515719.
For earlier commits similar to this one, see
commits 601cf582, 05272718, e83ca519.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since opencontainers/cgroups v0.0.2 (commit b206a015), all stuct
Resources fields are annotated with "omitempty" attribute.
As a result, the loaded configuration may have Resources == nil.
It is totally OK (rootless containers may have no resources configured)
except since commit 6c5441e5, cgroup v1 fs manager requires Resources to
be set in the call to NewManager (this is a cgroup v1 deficiency,
or maybe our implementation deficiency, or both).
To work around this, let's add code to ensure Resources is never nil
after loading from state.json.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The per-file deprecation in cgroup_deprecated.go is not working,
let's replace it.
Link to Hooks.Run in Hook.Run deprecation notice.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Implement support for passing Linux Network Devices to the container
network namespace.
The network device is passed during the creation of the container,
before the process is started.
It implements the logic defined in the OCI runtime specification.
Signed-off-by: Antonio Ojea <aojea@google.com>
In case there's a duplicate in the device list, the latter entry
overrides the former one.
So, we need to modify the last entry, not the first one. To do that,
use slices.Backward.
Amend the test case to test the fix.
Reported-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This support was missing from runc, and thus the example from the
podman-update wasn't working.
To fix, introduce a function to either update or insert new weights and iops.
Add integration tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of generating a list of tmpfs mount and have a special function
to check whether the path is in the list, let's go over the list of
mounts directly. This simplifies the code and improves readability.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since its code is now trivial, and it is only called from a single
place, it does not make sense to have it as a separate function.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It makes sense to ignore cgroup mounts much early in the code,
saving some time on unnecessary operations.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Replace the big "if !" block with the if block and continue,
simplifying the code flow.
2. Move comments closer to the code, improving readability.
This commit is best reviewed with --ignore-all-space or similar.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is to ensure that our CI is not rotting away even if there are no
new PRs or merges. This is especially useful for release branches
which tend to cease working over time due to some external reasons.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When we run CI not on a pull request, the commit job is skipped, as a
result, all-done is also skipped.
To allow all-done to succeed, modify the commit job to succeed for
non-PRs.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 874207492 neglects to update the exclusion rules when bumping Go
releases, and so we no longer exclude running on actuated with older Go
release, or running with criu-dev with older Go release.
Fixes: 874207492 ("CI: add Go 1.24, drop go1.22")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of providing systemd CPU quota value (CPUQuotaPerSec),
calculate it based on how opencontainers/cgroups/systemd handles
it (see addCPUQuota).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
To not accidentally break `go install`, let's add CI to check it. If in
the future we need those directives, we can remove the CI check.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We already have the indirect require for 1.17.3, that comes
opencontainers/cgroups[1]. That module requires that version as min, so
go can't use older versions. We can just remove the excludes.
There might be cases where people can use runc as a dependency and use
replace to override it (not sure, but probably). We were clear on what
our dependencies are, so we can sleep fine. In the unlikely case that
some project uses runc as a dependency and:
* Uses a replace for cilium v0.17.x but not the latest patch release (0.17.3 is fixed)
* they run with 32bits
* and hit this (that didn't happen always on CI)
* Ignore the changelog for 0.17.3 that mentions the buffer overflow on
32 bits platforms[2].
In that case, if we have a bug report, we can point them to the right
place. But 0.17.3 was released for some months now (most people probably
update) and 0.18.0 was released recently. I wouldn't worry about someone
hitting this in real life.
Also, the excludes directives prevent go install from working, so let's
just remove them.
[1]: https://github.com/opencontainers/cgroups/blob/9657f5a18b8d60a0f39fbb34d0cb7771e28e6278/go.mod#L6
[2]: https://github.com/cilium/ebpf/releases/tag/v0.17.3
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
For some reason, launchpad.net is frequently giving us Gateway Timeout.
Let's retry adding the ppa once to mitigate that.
(The alternative is not to install criu and thus run criu-related unit
tests on i386 -- this might actually be better).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This ensures that if runc is built without the provided Makefile, the
version is still properly set.
No change in the output.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of setting cli.App.Version in main, let's set up
cli.VersionPrinter. This way, we only get various versions
when needed.
Note it does not change the output of runc --version.
It changes the output of runc --help though, and I think it's for the
better.
Before this patch:
> $ runc help
> ...
> USAGE:
> runc [global options] command [command options] [arguments...]
>
> VERSION:
> 1.3.0-rc.1+dev
> commit: v1.3.0-rc.1-93-g932e8342
> spec: 1.2.1
> go: go1.24.2
> libseccomp: 2.5.5
>
> COMMANDS:
> checkpoint checkpoint a running container
> ...
After:
> $ runc help
> ...
> USAGE:
> runc [global options] command [command options] [arguments...]
>
> VERSION:
> 1.3.0-rc.1+dev
>
> COMMANDS:
> checkpoint checkpoint a running container
> ...
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
While debugging an issue involving failing mounts, I discovered that
just returning the plain mount error message when we are in the fallback
code for handling locked mounts leads to unnecessary confusion.
It also doesn't help that podman currently forcefully sets "rw" on
mounts, which means that rootless containers are likely to hit the
locked mounts issue fairly often.
So we should improve our error messages to explain why the mount is
failing in the locked flags case.
Fixes: 7c71a22705 ("rootfs: remove --no-mount-fallback and finally fix MS_REMOUNT")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
When reading mount errors, it is quite hard to make sense of mount flags
in their hex form. As this is the error path, the minor performance
impact of constructing a string is probably not worth hyper-optimising.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
For some reason, ssh-keygen is unable to write to /root even as root on
AlmaLinux 8:
# id
uid=0(root) gid=0(root) groups=0(root) context=system_u:system_r:initrc_t:s0
# id -Z
ls -ld /root
# ssh-keygen -t ecdsa -N "" -f /root/rootless.key || cat /var/log/audit/audit.log
Saving key "/root/rootless.key" failed: Permission denied
The audit.log shows:
> type=AVC msg=audit(1744834995.352:546): avc: denied { dac_override } for pid=13471 comm="ssh-keygen" capability=1 scontext=system_u:system_r:ssh_keygen_t:s0 tcontext=system_u:system_r:ssh_keygen_t:s0 tclass=capability permissive=0
> type=SYSCALL msg=audit(1744834995.352:546): arch=c000003e syscall=257 success=no exit=-13 a0=ffffff9c a1=5641c7587520 a2=241 a3=180 items=0 ppid=4978 pid=13471 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=(none) ses=4294967295 comm="ssh-keygen" exe="/usr/bin/ssh-keygen" subj=system_u:system_r:ssh_keygen_t:s0 key=(null)␝ARCH=x86_64 SYSCALL=openat AUID="unset" UID="root" GID="root" EUID="root" SUID="root" FSUID="root" EGID="root" SGID="root" FSGID="root"
A workaround is to use /root/.ssh directory instead of just /root.
While at it, let's unify rootless user and key setup into a single place.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We are seeing a ton on flakes on almalinux-8 CI job, all caused by criu
inability to freeze a cgroup. This was worked around in criu [1], but
obviously we can't rely on a distro vendor to update the package.
Let's use a copr (thanks to Adrian Reber!)
[1]: https://github.com/checkpoint-restore/criu/pull/2545
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The colon after "Error:" caused actionlint to report error on map in
context where map is not allowed.
Signed-off-by: Antti Kervinen <antti.kervinen@intel.com>
This issue was originally reported in podman PR 25792.
When calling runc pause/unpause for an ordinary user, podman do not
provide --systemd-cgroups option, and shouldUseRootlessCgroupManager
returns true. This results in a warning:
$ podman pause sleeper
WARN[0000] runc pause may fail if you don't have the full access to cgroups
sleeper
Actually, it does not make sense to call shouldUseRootlessCgroupManager
at this point, because we already know if we're rootless or not, from
the container state.json (same for systemd).
Also, busctl binary is not available either in this context, so
shouldUseRootlessCgroupManager would not work properly.
Finally, it doesn't really matter if we use systemd or not, because we
use fs/fs2 manager to freeze/unfreeze, and it will return something like
EPERM (or tell that cgroups is not configured, for a true rootless
container).
So, let's only print the warning after pause/unpause failed,
if the error returned looks like a permission error.
Same applies to "runc ps".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Because we have to set a default HOME env for the current container
user, so we should set it after we are in the jail of the container,
or else we'll use host's `/etc/passwd` to get a wrong HOME value.
Please see: #4688.
Signed-off-by: lifubang <lifubang@acmcoder.com>
That is,
> QF1008: could remove embedded field "Resources" from selector (staticcheck)
While occasionally useful, in other cases it actually decreases
readability, so let's disable it even for "extra" (i.e. "new code")
linters.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This tests the functionality added by commit cd96170c1
("Need to setup labeling of kernel keyrings."), for both
runc run and runc exec, with and without user namespace.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This package is to provide unix.* wrappers to ensure that:
- they retry on EINTR;
- a "rich" error is returned on failure.
A first such wrapper, Sendmsg, is introduced.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since v3.14, CRIU always restores processes into a time namespace to
prevent backward jumps of monotonic and boottime clocks. This change
updates the container configuration to ensure that `runc exec` launches
new processes within the container's time namespace.
Fixes#2610
Signed-off-by: Andrei Vagin <avagin@gmail.com>
The new configuration file was initially generated by golangci-lint
migrate, when tweaked to minimize and simplify.
golangci-lint v2 switches to a new version of staticcheck which shows
much more warnings. Some of them were fixed by a few previous commits,
and the rest of them are disabled.
In particular, ST1005 had to be disabled (an attempt to fix it was made
in https://github.com/opencontainers/runc/pull/3857 but it wasn't
merged).
Also, golangci-extra was modified to include ALL staticcheck linters.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
> libcontainer/intelrdt/cmt.go:5:1: ST1020: comment on exported function IsCMTEnabled should be of the form "IsCMTEnabled ..." (staticcheck)
> // Check if Intel RDT/CMT is enabled.
> ^
> libcontainer/intelrdt/intelrdt.go:419:1: ST1020: comment on exported function IsCATEnabled should be of the form "IsCATEnabled ..." (staticcheck)
> // Check if Intel RDT/CAT is enabled
> ^
> libcontainer/intelrdt/intelrdt.go:425:1: ST1020: comment on exported function IsMBAEnabled should be of the form "IsMBAEnabled ..." (staticcheck)
> // Check if Intel RDT/MBA is enabled
> ^
> libcontainer/intelrdt/intelrdt.go:446:1: ST1020: comment on exported method Apply should be of the form "Apply ..." (staticcheck)
> // Applies Intel RDT configuration to the process with the specified pid
> ^
> libcontainer/intelrdt/intelrdt.go:481:1: ST1020: comment on exported method Destroy should be of the form "Destroy ..." (staticcheck)
> // Destroys the Intel RDT container-specific 'container_id' group
> ^
> libcontainer/intelrdt/intelrdt.go:497:1: ST1020: comment on exported method GetPath should be of the form "GetPath ..." (staticcheck)
> // Returns Intel RDT path to save in a state file and to be able to
> ^
> libcontainer/intelrdt/intelrdt.go:506:1: ST1020: comment on exported method GetStats should be of the form "GetStats ..." (staticcheck)
> // Returns statistics for Intel RDT
> ^
> libcontainer/intelrdt/mbm.go:6:1: ST1020: comment on exported function IsMBMEnabled should be of the form "IsMBMEnabled ..." (staticcheck)
> // Check if Intel RDT/MBM is enabled.
> ^
> 8 issues:
> * staticcheck: 8
While at it, add missing periods.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I was pretty sure we have a linter for these but apparently we did not.
> libcontainer/capabilities/capabilities.go:108:1: ST1020: comment on exported method ApplyCaps should be of the form "ApplyCaps ..." (staticcheck)
> // Apply sets all the capabilities for the current process in the config.
> ^
>
>
> types/events.go:15:1: ST1021: comment on exported type Stats should be of the form "Stats ..." (with optional leading article) (staticcheck)
> // stats is the runc specific stats structure for stability when encoding and decoding stats.
> ^
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Like these:
> libcontainer/criu_linux.go:959:3: QF1001: could apply De Morgan's law (staticcheck)
> !(req.GetType() == criurpc.CriuReqType_FEATURE_CHECK ||
> ^
> libcontainer/rootfs_linux.go:360:19: QF1001: could apply De Morgan's law (staticcheck)
> if err == nil || !(errors.Is(err, unix.EPERM) || errors.Is(err, unix.EBUSY)) {
> ^
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
> notify_socket.go:44:24: ST1016: methods on the same type should have the same receiver name (seen 1x "n", 5x "s") (staticcheck)
> func (s *notifySocket) Close() error {
> ^
As reported by staticcheck from golangci-lint v2.0.0
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The errrolint linter can finally ignore errors from Close,
and it also ignores direct comparisons of errors from x/sys/unix.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We received several times issues that the repro steps are human readable
text with ambiguous instructions. That usually ends up in maintainers
asking questions so people provide clear steps.
Let's just make the issue template more clear in that regard.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This makes the state.json file 1303 bytes or almost 25% smaller (when
using the default spec, YMMV) by omitting default values.
Before: 5496 bytes
After: 4193 bytes
(With cgroups#9 applied, the new size is 3424, which is almost 40%
savings, compared to the original).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no behavior change, it is just more readable to use -1 to mean
don't touch this.
Please note that if the GID is not mapped in the userns, by using -1 for
that no error is returned. We just avoid dealing with it completely, as
we want here.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
1. There is no need to have -p option in mkdir here, since
/home/rootless was already created by useradd above.
2. When there is no -p, there is no need to suppress the shellcheck
warning (which looked like this):
> In script/setup_host_fedora.sh line 21:
> mkdir -m 0700 -p /home/rootless/.ssh
> ^-- SC2174 (warning): When used with -p, -m only applies to the deepest directory.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Those are no longer needed with shellcheck v0.10.0 (possibly with an
earlier version, too, but I am too lazy to check that).
While at it, fix a typo in the comment.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This addresses a TODO item added by commit 40f146841
("keyring: handle ENOSYS with keyctl(KEYCTL_JOIN_SESSION_KEYRING)"),
as we do have runc init logging working fine for quite some time.
While at it, fix a typo in a comment (standart -> standard).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It was a mistake to say that SELinux need to be in the enforcing mode
for these tests to run. It only needs to be enabled.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In a nutshell:
- use git-core instead of git;
- do not install weak deps;
- do not install docs.
This results in less packages to install:
- 25 instead of 72 for almalinux-8
- 24 instead of 90 for almalinux-9
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This way, given a recent Go and Linux version, pidfd_send_signal will
be used under the hood.
Keep unix.Signal and unix.SignalName for logging (it is way more
readable than what os.Signal.String() provides).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Because we should switch to unix.PidFDSendSignal in new kernels, it has
been supported in go runtime. We don't need to add fall back to
unix.Kill code here.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
- Unlike proprietary Vagrant, Lima remains to be an open source project
- GHA now natively supports nested virt on Linux runners
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
As per
- https://github.com/opencontainers/runtime-spec/pull/1253
- https://github.com/opencontainers/runtime-spec/pull/1261
CPU affinity can be set in two ways:
1. When creating/starting a container, in config.json's
Process.ExecCPUAffinity, which is when applied to all execs.
2. When running an exec, in process.json's CPUAffinity, which
applied to a given exec and overrides the value from (1).
Add some basic tests.
Note that older kernels (RHEL8, Ubuntu 20.04) change CPU affinity of a
process to that of a container's cgroup, as soon as it is moved to that
cgroup, while newer kernels (Ubuntu 24.04, Fedora 41) don't do that.
Because of the above,
- it's impossible to really test initial CPU affinity without adding
debug logging to libcontainer/nsenter;
- for older kernels, there can be a brief moment when exec's affinity
is different than either initial or final affinity being set;
- exec's final CPU affinity, if not specified, can be different
depending on the kernel, therefore we don't test it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is a big loop(at least 65 times) in `signal.Notify`, it costs as much
time as `runc init`, so we can call it in parallel ro reduce the container
start time. In a general test, it can be reduced about 38.70%.
Signed-off-by: lifubang <lifubang@acmcoder.com>
(cyphar: move signal channel definition inside goroutine)
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This removes libcontainer/cgroups packages and starts
using those from github.com/opencontainers/cgroups repo.
Mostly generated by:
git rm -f libcontainer/cgroups
find . -type f -name "*.go" -exec sed -i \
's|github.com/opencontainers/runc/libcontainer/cgroups|github.com/opencontainers/cgroups|g' \
{} +
go get github.com/opencontainers/cgroups@v0.0.1
make vendor
gofumpt -w .
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Prior to kernel Linux 5.5, F_SEAL_FUTURE_WRITE has a bug which maps
memory as shared between processes even if it is set as private. See
kernel commit 05d351102dbe ("mm, memfd: fix COW issue on MAP_PRIVATE and
F_SEAL_FUTURE_WRITE mappings") for more details.
According to the fcntl(2) man pages, F_SEAL_WRITE is enough:
> Furthermore, trying to create new shared, writable memory-mappings via
> mmap(2) will also fail with EPERM.
>
> Using the F_ADD_SEALS operation to set the F_SEAL_WRITE seal fails
> with EBUSY if any writable, shared mapping exists. Such mappings must
> be unmapped before you can add this seal.
F_SEAL_FUTURE_WRITE only makes sense if a read-write shared mapping in
one process should be read-only in another process. This is not case for
runc, especially not for the /proc/self/exe we are protecting.
Signed-off-by: Tomasz Duda <tomaszduda23@gmail.com>
(cyphar: improve the comment regarding F_SEAL_FUTURE_WRITE)
(cyphar: improve commit message)
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The "dmz" name was originally used because the libcontainer/dmz package
housed the runc-dmz binary, but since we removed it in commit
871057d863 ("drop runc-dmz solution according to overlay solution")
the name is an anachronism and we should just give it a more
self-explanatory name.
So, call it libcontainer/exeseal because the purpose of the package is
to provide tools to seal /proc/self/exe against attackers.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We should configure the process's timens offset only when we need to
create new time namespace, we shouldn't do it if we are joining an
existing time namespace. (#4635)
Signed-off-by: lifubang <lifubang@acmcoder.com>
This was needed for a test case only, but we can easily copy the data
needed.
The alternatives are:
- keep things as is (and have cgroups depend on
runc/libcontainer/specconv);
- remove this test case;
- move AllowedDevices to cgroups/devices/config.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead, we can just do filepath.Clean("/"+path) here.
While at it, add a comment telling why this is needed and important.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The code which determines inner cgroup path from cgroup config is
identical in fs and fs2 drivers, and it is using utils.CleanPath.
In preparation to move libcontainer/cgroups to a separate repo,
we have to get rid of libcontainer/utils dependency. So,
- copy the utils.CleanPath implementation to internal/path;
- consolidate the two innerPath implementations to internal/path.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The _defaultDirPath was only used for testing, and the test case
is quite easy to adopt to defaultDirPath.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We were using utils.ProcThreadSelf since commit 8e8b136c,
which provides two things:
1. locking the OS tread;
2. fallback to /proc/self/task/$TID when /proc/thread-self
is not available (kernel < 3.17).
Now, (1) is not needed since we only call readlink and not perform any
file data operation, and (2) is not needed here as this code is
only running when openat2 syscall is available, meaning kernel >= v5.6.
Also, check the error from readlink, so when it fails, we do not try to
enhance the error message.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is an announce that Ubuntu 20.04 will be removed in April,
and in March there will be a few "brown-out" dates/times when
it won't work.
This leaves us with no other options than to remove ubuntu-20.04
from the testing matrix.
As a result, cgroup v1 testing will only be done on AlmaLinux 8
running on CirrusCI. It is probably going to be sufficient for
the time being (until we deprecate cgroup v1).
If not, our options are
- run Ubuntu 20.04 (or other cgroup v1 distro) in a VM on GHA;
- switch to cirrus-ci.
[1]: https://github.com/actions/runner-images/issues/11101
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This simplifies the code flow and basically removes the last
filepath.Clean, which is not necessary in either case:
- for absolute path, single filepath.Clean is enough (as it is
guaranteed to remove all dot and dot-dot elements);
- for relative path, filepath.Rel calls Clean at the end
(which is even documented).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
On my machine, the --recv-keys steps to get upstream keys started
producing errors recently, and even setting a default keyserver in the
global gpg configuration doesn't seem to help:
+ gpg --homedir=/tmp/runc-sign-tmpkeyring.qm0IP6
--no-default-keyring --keyring=seccomp.keyring
--recv-keys 0x47A68FCE37C7D7024FD65E11356CE62C2B524099
gpg: keybox '/tmp/runc-sign-tmpkeyring.qm0IP6/seccomp.keyring' created
gpg: keyserver receive failed: No keyserver available
So just explicitly specify a reputable keyserver. Ideally we would use
an .onion-address keyserver to avoid potential targeted attacks but not
everybody runs a Tor proxy on their machine.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
A new libseccomp releases (v2.5.6 and v2.6.0) were cut last month.
Theoretically, we could use v2.6.0 but let's stay conservative for now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Also:
1. Change GO_VERSION to GO_VER_PREFIX, and move the "." from the jq
argument to the variable value. It allows to use something like
"1.25" to match "1.25rc" etc, but set to "1.24." for now to require
a released 1.24.x version.
2. Change PREFIX to URL_PREFIX.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Also, bump golangci-lint to v1.64 (v1.64.2 added Go 1.24 support).
NOTE we still use Go 1.23.x for official builds.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 770728e1 added Scheduler field into both Config and Process,
but forgot to add a mechanism to actually use Process.Scheduler.
As a result, runc exec does not set Process.Scheduler ever.
Fix it, and a test case (which fails before the fix).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit bfbd0305b added IOPriority field into both Config and Process,
but forgot to add a mechanism to actually use Process.IOPriority.
As a result, runc exec does not set Process.IOPriority ever.
Fix it, and a test case (which fails before the fix).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For all other properties that are available in both Config and Process,
the merging is performed by newInitConfig.
Let's do the same for Capabilities for the sake of code uniformity.
Also, thanks to the previous commit, we no longer have to make sure we
do not call capabilities.New(nil).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In runtime-spec, capabilities property is optional, but
libcontainer/capabilities panics when New(nil) is called.
Because of this, there's a kludge in finalizeNamespace to ensure
capabilities.New is not called with nil argument, and there's a
TestProcessEmptyCaps to ensure runc won't panic.
Let's fix this at the source, allowing libct/cap to work with nil
capabilities.
(The caller is fixed by the next commit.)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
They are passed in initConfig twice, so it does not make sense.
NB: the alternative to that would be to remove Config field from
initConfig, but it results in a much bigger patch and more maintenance
down the road.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is one of the dark corners of runc / libcontainer, so let's shed
some light on it.
initConfig is a structure which is filled in [mostly] by newInitConfig,
and one of its hidden aspects is it contains a process config which is
the result of merge between the container and the process configs.
Let's document how all this happens, where the fields are coming from,
which one has a preference, and how it all works.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Pass an argument as a pointer rather than copying the whole structure.
It was a pointer initially, but this has changed in commit b2d9d996
without giving a reason why.
2. The newProcess description was added by commit 9fac18329 (yes, the
very first one) and hasn't changed since. As of commit 29b139f7,
the part of it which says "and stdio from the current process"
is no longer valid.
Remove it, and while at it, rewrite the description entirely.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The rootuid and rootgid are only needed when detach and createTTY are
both false. We also call c.Config() twice, every time creating a copy
of struct Config.
Solve both issues by passing container pointer to setupIO, and get
rootuid/rootgid only when we need those.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Every time we call container.Config(), a new copy of
struct Config is created and returned, and we do it twice here.
Accessing container.config directly fixes this.
Fixes: 805b8c73d ("Do not create exec fifo in factory.Create")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Let's move some code from execProcess to newProcess, fixing the
following few issues:
1. container.State (which does quite a lot) is not needed --
we only need container.Config here.
2. utils.SearchLabels is not needed when "runc exec --process" is used.
3. Context.String("process") is called twice.
4. It is not very clear from the code why checking for
len(context.Args()) is performed. Move the check to just before
Args is used, to make it clear why.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Looking into old opened runc issues, I noticed #1663 is there without
any resolution, and wrote this simple test checking if we mangle hook's
argv[0] in any way.
Apparently we're good, but the test actually makes sense to have.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Avoid splitting mount data into []string if it does not contain
options we're interested in. This should result in slightly less
garbage to collect.
2. Use if / else if instead of continue, to make it clearer that
we're processing one option at a time.
3. Print the whole option as a sting in an error message; practically
this should not have any effect, it's just simpler.
4. Improve some comments.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.CutPrefix (available since Go 1.20) instead of
strings.HasPrefix and/or strings.TrimPrefix makes the code
a tad more straightforward.
No functional change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.HasPrefix with strings.TrimPrefix results in doing the
same thing (checking if prefix exists) twice. In this case, using
strings.TrimPrefix right away is sufficient.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. GetCgroupParamUint: drop strings.TrimSpace since it was already
done by GetCgroupParamString.
2. GetCgroupParamInt: use GetCgroupParamString, drop strings.TrimSpace.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It makes sense to report an error if a key or a value is empty,
as we don't expect anything like this.
Reported-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.CutPrefix (added in Go 1.20, see [1]) results in faster and
cleaner code with less allocations (as the code only allocates memory
for the value, and does it once).
While at it, improve the function documentation.
[1]: https://github.com/golang/go/issues/42537
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Cut (added in Go 1.18, see [1]) results in faster and
cleaner code with less allocations (as we're not using a slice).
Also, use switch in parseRdmaKV.
[1]: https://github.com/golang/go/issues/46336
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Cut (added in Go 1.18, see [1]) results in faster and
cleaner code with less allocations (as we're not using a slice).
This code is tested by TestStatCPUPSI.
[1]: https://github.com/golang/go/issues/46336
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Cut (added in Go 1.18, see [1]) results in faster and
cleaner code with less allocations (as we're not using a slice).
The code is tested by testCgroupResourcesUnified.
[1]: https://github.com/golang/go/issues/46336
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For cgroup v2, we always expect /proc/$PID/cgroup contents like this:
> 0::/user.slice/user-1000.slice/user@1000.service/app.slice/vte-spawn-f71c3fb8-519d-4e2d-b13e-9252594b1e05.scope
So, it does not make sense to parse it using strings.Split, we can just
cut the prefix and return the rest.
Code tested by TestParseCgroupFromReader.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Remove extra global constants that are only used in a single place and
make it harder to read the code.
Rename nanosecondsInSecond -> nsInSec.
This code is tested by unit tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Cut (added in Go 1.18, see [1]) results in faster and
cleaner code with less allocations (as we're not using a slice). This
also drops the check for extra dash (we're unlikely to get it from the
kernel anyway).
While at it, rename min/max -> from/to to avoid collision with Go
min/max builtins.
This code is tested by TestCPUSetStats* tests.
[1]: https://github.com/golang/go/issues/46336
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Document the function.
2. Add sanity checks for empty and repeated controllers.
Reported-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Cut (added in Go 1.18, see [1]) results in faster and
cleaner code with less allocations (as we're not using a slice).
This part of code is covered by tests in tests/integration/exec.bats.
[1]: https://github.com/golang/go/issues/46336
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Switch from github.com/moby/sys/user to Go stdlib os/user
(which has both libc-backed and pure Go implementations).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since the UID/GID/AdditonalGroups fields are now numeric,
we can address the following TODO item in the code (added
by commit d2f49696 back in 2016):
> TODO: We currently can't do
> this check earlier, but if libcontainer.Process.User was typesafe
> this might work.
Move the check to much earlier phase, when we're preparing
to start a process in a container.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This addresses the following TODO in the code (added back in 2015
by commit 845fc65e5):
> // TODO: fix libcontainer's API to better support uid/gid in a typesafe way.
Historically, libcontainer internally uses strings for user, group, and
additional (aka supplementary) groups.
Yet, runc receives those credentials as part of runtime-spec's process,
which uses integers for all of them (see [1], [2]).
What happens next is:
1. runc start/run/exec converts those credentials to strings (a User
string containing "UID:GID", and a []string for additional GIDs) and
passes those onto runc init.
2. runc init converts them back to int, in the most complicated way
possible (parsing container's /etc/passwd and /etc/group).
All this conversion and, especially, parsing is totally unnecessary,
but is performed on every container exec (and start).
The only benefit of all this is, a libcontainer user could use user and
group names instead of numeric IDs (but runc itself is not using this
feature, and we don't know if there are any other users of this).
Let's remove this back and forth translation, hopefully increasing
runc exec performance.
The only remaining need to parse /etc/passwd is to set HOME environment
variable for a specified UID, in case $HOME is not explicitly set in
process.Env. This can now be done right in prepareEnv, which simplifies
the code flow a lot. Alas, we can not use standard os/user.LookupId, as
it could cache host's /etc/passwd or the current user (even with the
osusergo tag).
PS Note that the structures being changed (initConfig and Process) are
never saved to disk as JSON by runc, so there is no compatibility issue
for runc users.
Still, this is a breaking change in libcontainer, but we never promised
that libcontainer API will be stable (and there's a special package
that can handle it -- github.com/moby/sys/user). Reflect this in
CHANGELOG.
For 3998.
[1]: https://github.com/opencontainers/runtime-spec/blob/v1.0.2/config.md#posix-platform-user
[2]: https://github.com/opencontainers/runtime-spec/blob/v1.0.2/specs-go/config.go#L86
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Historically, when HOME is not explicitly set in process.Env,
and UID to run as doesn't have a corresponding entry in container's
/etc/passwd, runc sets HOME=/ as a fallback.
Add the corresponding check, for the sake of backward compatibility
preservation.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes k3s cross-compilation on Windows, broken by commit
1912d5988b ("*: actually support joining a userns with a new
container").
[@kolyshkin: commit message]
Fixes: 1912d5988b
Signed-off-by: Brad Davidson <brad.davidson@rancher.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Similar to when SetAmbient() can fail, runc should be graceful about
ResetAmbient failing.
This functionality previously worked under gvisor, which doesn't
implement ambient capabilities atm. The hard error on reset broke gvisor
usage.
Signed-off-by: Evan Phoenix <evan@phx.io>
Use the old package name as an alias to minimize the patch.
No functional change; this just eliminates a bunch of deprecation
warnings.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently, libcontainer/devices contains two things:
1. Device-related configuration data structures and accompanying
methods. Those are used by runc itself, mostly by libct/cgroups.
2. A few functions (HostDevices, DeviceFromPath, GetDevices).
Those are not used by runc directly, but have some external users
(cri-o, microsoft/hcsshim), and they also have a few forks
(containerd/pkg/oci, podman/pkg/util).
This commit moves (1) to a new separate package, config (under
libcontainer/cgroups/devices), adding a backward-compatible aliases
(marked as deprecated so we will be able to remove those later).
Alas it's not possible to move this to libcontainer/cgroups directly
because some IDs (Type, Rule, Permissions) are too generic, and renaming
them (to DeviceType, DeviceRule, DevicePermissions) will break backward
compatibility (mostly due to Rule being embedded into Device).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This release includes a minor breaking API change that requires us to
rework the types of our wrappers, but there is no practical behaviour
change.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The previous logic caused runc to hang if CloseExecFrom returned an
error, as the defer waiting on logsDone never finished as the parent
process was never started (and it controls the closing of logsDone via
it's logsPipe).
This moves the defer to after we have started the parent, with means all
the logic related to managing the logsPipe should also be running.
Signed-off-by: Evan Phoenix <evan@phx.io>
The current implementation sets all the environment variables passed in
Process.Env in the current process, one by one, then uses os.Environ to
read those back.
As pointed out in [1], this is slow, as runc calls os.Setenv for every
variable, and there may be a few thousands of those. Looking into how
os.Setenv is implemented, it is indeed slow, especially when cgo is
enabled.
Looking into why it was implemented the way it is, I found commit
9744d72c and traced it to [2], which discusses the actual reasons.
It boils down to these two:
- HOME is not passed into container as it is set in setupUser by
os.Setenv and has no effect on config.Env;
- there is a need to deduplicate the environment variables.
Yet it was decided in [2] to not go ahead with this patch, but
later [3] was opened with the carry of this patch, and merged.
Now, from what I see:
1. Passing environment to exec is way faster than using os.Setenv and
os.Environ (tests show ~20x speed improvement in a simple Go test,
and ~3x improvement in real-world test, see below).
2. Setting environment variables in the runc context may result is some
ugly side effects (think GODEBUG, LD_PRELOAD, or _LIBCONTAINER_*).
3. Nothing in runtime spec says that the environment needs to be
deduplicated, or the order of preference (whether the first or the
last value of a variable with the same name is to be used). We should
stick to what we have in order to maintain backward compatibility.
So, this patch:
- switches to passing env directly to exec;
- adds deduplication mechanism to retain backward compatibility;
- takes care to set PATH from process.Env in the current process
(so that supplied PATH is used to find the binary to execute),
also to retain backward compatibility;
- adds HOME to process.Env if not set;
- ensures any StartContainer CommandHook entries with no environment
set explicitly are run with the same environment as before. Thanks
to @lifubang who noticed that peculiarity.
The benchmark added by the previous commit shows ~3x improvement:
│ before │ after │
│ sec/op │ sec/op vs base │
ExecInBigEnv-20 61.53m ± 1% 21.87m ± 16% -64.46% (p=0.000 n=10)
[1]: https://github.com/opencontainers/runc/pull/1983
[2]: https://github.com/docker-archive/libcontainer/pull/418
[3]: https://github.com/docker-archive/libcontainer/pull/432
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Make CommandHook.Command a pointer, which reduces the amount of data
being copied when using hooks, and allows to modify command hooks.
2. Add SetDefaultEnv, which is to be used by the next commit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is to ensure that changes in Process.Env handling won't affect
StartContainer hook.
Reported-by: lfbzhm <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a slight refactor of TestExecInEnvironment, making it more
strict wrt checking the exec output.
1. Explain why DEBUG is added twice to the env.
2. Reuse the execEnv for the check.
3. Make the check more strict -- instead of looking for substrings,
check line by line.
4. Add a check for extra environment variables.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Here's what it shows on my laptop (with -count 10 -benchtime 10s,
summarized by benchstat):
│ sec/op │
ExecTrue-20 8.477m ± 2%
ExecInBigEnv-20 61.53m ± 1%
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Assigning a multi-line value to a bash variable should not be so complex.
While at it, slightly reformat create_runtime_hook.
No functional change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Android mounts the v1 cpuset cgroup with the noprefix option.
As a result, runc first attempts to access cpuset files using the prefix format (e.g., cpuset.cpus).
If this fails, it falls back to accessing them without the prefix (e.g., cpus).
Once a successful access method is determined, it is cached and used for all subsequent operations.
Only the v1 cpuset cgroup is allowed to mount with noprefix. See kernel source:
https://github.com/torvalds/linux/blob/2e1b3cc9d7f790145a80cb705b168f05dab65df2/kernel/cgroup/cgroup-v1.c#L1070.
Cpuset cannot be mounted with and without prefix simultaneously.
Signed-off-by: Tomasz Duda <tomaszduda23@gmail.com>
This helper was added for runc-dmz in commit dac417174, but runc-dmz was
later removed in commit 871057d, which forgot to remove the helper.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is an internal implementation detail and should not be either
public or visible.
Amend setIOPriority to do own class conversion.
Fixes: bfbd0305 ("Add I/O priority")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For some reason, io priority is set in different places between runc
start/run and runc exec:
- for runc start/run, it is done in the middle of (*linuxStandardInit).Init,
close to the place where we exec runc init.
- for runc exec, it is done much earlier, in (*setnsProcess) start().
Let's move setIOPriority call for runc exec to (*linuxSetnsInit).Init,
so it is in the same logical place as for runc start/run.
Also, move the function itself to init_linux.go as it's part of init.
Should not have any visible effect, except part of runc init is run with
a different I/O priority.
While at it, rename setIOPriority to setupIOPriority, and make it accept
the whole *configs.Config, for uniformity with other similar functions.
Fixes: bfbd0305 ("Add I/O priority")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This code is not in libcontainer, meaning it is only used by a short lived
binary (runc start/run/exec). Unlike code in libcontainer (see
CreateLibcontainerConfig), here we don't have to care about copying the
structures supplied as input, meaning we can just reuse the pointers
directly.
Fixes: bfbd0305 ("Add I/O priority")
Fixes: 770728e1 ("Support `process.scheduler`")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
SpecState field of initConfig is only needed to run hooks that are
executed inside a container -- namely CreateContainer and
StartContainer.
If these hooks are not configured, there is no need to fill, marshal and
unmarshal SpecState.
While at it, inline updateSpecState as it is trivial and only has one user.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This allows to omit a call to c.currentOCIState (which can be somewhat
costly when there are many annotations) when the hooks of a given kind
won't be run.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Introduce a common parent struct `containerProcess`,
let both `initProcess` and `setnsProcess` are inherited
from it.
Signed-off-by: lfbzhm <lifubang@acmcoder.com>
Historically, our release cadence and support policy has been quite
ad-hoc, which has caused some strife and is not really becoming of a
project as widely used as ours. So, in an attempt to make our releases
more regular and to provide more guidance on our backport policy, add a
document outlining our policy.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
GHA shows a warning telling that "ubuntu-latest" is going to be switched
to ubuntu-24.04 soon. Let's specify the version explicitly (and switch
to 24.04 for this job ahead of github).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since v1.2.0 was released, a number of users complained that the removal
of tun/tap device access from the default device ruleset is causing a
regression in their workloads.
Additionally, it seems that some upper-level orchestration tools
(Docker Swarm, Kubernetes) makes it either impossible or cumbersome
to supply additional device rules.
While it's probably not quite right to have /dev/net/tun in a default
device list, it was there from the very beginning, and users rely on it.
Let's keep it there for the sake of backward compatibility.
This reverts commit 2ce40b6ad7.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We have quite a few external users of libcontainer/cgroups packages,
and they all have to depend on libcontainer/configs as well.
Let's move cgroup-related configuration to libcontainer/croups.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These:
> Error: libcontainer/cgroups/fs2/cpu.go:15:6: var-naming: func isCpuSet should be isCPUSet (revive)
> func isCpuSet(r *cgroups.Resources) bool {
> ^
> Error: libcontainer/cgroups/fs2/cpu.go:19:6: var-naming: func setCpu should be setCPU (revive)
> func setCpu(dirPath string, r *cgroups.Resources) error {
> ^
They are going to be shown after next commits because of linter-extra CI
job (which, due to major changes, now thinks it's a new code so extra
linters apply).
Fixing it beforehand.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of having every test helper binary in its own directory, let's
use /tests/cmd/_bin as a destination directory.
This allows for simpler setup/cleanup.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This allows to make a 17% smaller runc binary by not compiling in
checkpoint/restore support.
It turns out that google.golang.org/protobuf package, used by go-criu,
is quite big, and go linker can't drop unused stuff if reflection is
used anywhere in the code.
Currently there's no alternative to using protobuf in go-criu, and since
not all users use c/r, let's provide them an option for a smaller
binary.
For the reference, here's top10 biggest vendored packages, as reported
by gsa[1]:
$ gsa runc | grep vendor | head
│ 8.59% │ google.golang.org/protobuf │ 1.3 MB │ vendor │
│ 5.76% │ github.com/opencontainers/runc │ 865 kB │ vendor │
│ 4.05% │ github.com/cilium/ebpf │ 608 kB │ vendor │
│ 2.86% │ github.com/godbus/dbus/v5 │ 429 kB │ vendor │
│ 1.25% │ github.com/urfave/cli │ 188 kB │ vendor │
│ 0.90% │ github.com/vishvananda/netlink │ 135 kB │ vendor │
│ 0.59% │ github.com/sirupsen/logrus │ 89 kB │ vendor │
│ 0.56% │ github.com/checkpoint-restore/go-criu/v6 │ 84 kB │ vendor │
│ 0.51% │ golang.org/x/sys │ 76 kB │ vendor │
│ 0.47% │ github.com/seccomp/libseccomp-golang │ 71 kB │ vendor │
And here is a total binary size saving when `runc_nocriu` is used.
For non-stripped binaries:
$ gsa runc-cr runc-nocr | tail -3
│ -17.04% │ runc-cr │ 15 MB │ 12 MB │ -2.6 MB │
│ │ runc-nocr │ │ │ │
└─────────┴──────────────────────────────────────────┴──────────┴──────────┴─────────┘
And for stripped binaries:
│ -17.01% │ runc-cr-stripped │ 11 MB │ 8.8 MB │ -1.8 MB │
│ │ runc-nocr-stripped │ │ │ │
└─────────┴──────────────────────────────────────────┴──────────┴──────────┴─────────┘
[1]: https://github.com/Zxilly/go-size-analyzer
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 7f64fb47 made the main package, and runc/libcontainer's CriuOpts
depend on criu/rpc. This is not good; among the other things, it makes
it complicated to make c/r optional.
Let's switch CriuOpts.ManageCgroupsMode to a string (yes, it's an APIt
breaking change) and move the cgroup mode string parsing to
libcontainer.
While at it, let's better document ManageCgroupsMode.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes a regression in use of securejoin.MkdirAll, where multiple
runc processes racing to create the same mountpoint in a shared rootfs
would result in spurious EEXIST errors. In particular, this regression
caused issues with BuildKit.
Fixes: dd827f7b71 ("utils: switch to securejoin.MkdirAllHandle")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If we get an unexpected error here, it is probably because of a library
or kernel change that could cause our detection logic to be invalid. As
a result, these warnings should be louder so users have a chance to tell
us about them sooner (or so we might notice them before doing a release,
as happened with the 1.2.0 regression).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It is possible for LinkAttachProgram to return ErrNotSupported if
program attachment is not supported at all (which doesn't matter in this
case), but it seems possible that upstream will start returning
ErrNotSupported for BPF_F_REPLACE at some point so it's best to make
sure we don't cause additional regressions here.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In v0.13.0, cilium/ebpf stopped supporting setting BPF_F_REPLACE as an
explicit flag and instead requires us to use link.Anchor to specify
where the program should be attached.
Commit 216175a9ca ("Upgrade Cilium's eBPF library version to 0.16")
did update this correctly for the actual attaching logic, but when
checking for kernel support we still passed BPF_F_REPLACE. This would
result in a generic error being returned, which our feature-support
checking logic would treat as being an error the indicates that
BPF_F_REPLACE *is* supported, resulting in a regression on pre-5.6
kernels.
It turns out that our debug logging saying that this unexpected error
was happening was being output as a result of this change, but nobody
noticed...
Fixes: 216175a9ca ("Upgrade Cilium's eBPF library version to 0.16")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
A map which is created in func init is only used by capSlice, which is
only used by New, which is only used by runc init. Switch to lazy init
to slightly save on startup time.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If we are running with privileges and are asked to join an externally
created user namespaces as well as some other namespace that was *not*
created underneath said user namespace, the approach we added in commit
2cd9c31b99 ("nsenter: guarantee correct user namespace ordering")
doesn't work.
While in theory you would want all externally created namespaces to be
sane, it seems that some tools really do create unrelated namespaces and
ask us to join them. Luckily we can just loosely copy what nsenter(1)
appears to do -- we first try to join any namespaces we can (with host
root privileges), then we join any user namespaces, and then we join any
remaining namespaces (now with the user namespace's privileges).
Note that we *do not* have to try to join namespaces after we create our
own user namespace. Namespace permissions are based purely on the owning
user namespace (not the rootuid) so we will not have access to any extra
namespaces once we unshare(CLONE_NEWUSER) (in fact we will not be able
to setns(2) to anything!).
Fixes: 2cd9c31b99 ("nsenter: guarantee correct user namespace ordering")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This is basically a no-op change because runc already disallows this,
but it will be needed in future patches when we have to track what
namespaces have already been joined.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Arguably these docs should live elsewhere (especially if we plan to
remove memfd-bind in the future), but for now this is the only place
that fully explains this issue.
Suggested-by: Rodrigo Campos <rodrigoca@microsoft.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If the sub-cgroup RemovePath has failed for any reason, return the
error right away. This way, we don't have to check for err != nil
before retrying rmdir.
This is a cosmetic change and should not change any functionality.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
An issue with runc 1.2.0 was reported to buildkit, in which
runc delete returns with an error, with the log saying:
> unable to destroy container: unable to remove container's cgroup: open /sys/fs/cgroup/snschvixiy3s74w74fjantrdg: no such file or directory
Apparently, what happens is runc is running with no cgroup access
(because /sys/fs/cgroup is mounted read-only). In this case error to
create a cgroup path (in runc create/run) is ignored, but cgroup removal
(in runc delete) is not.
This is caused by commit d3d7f7d, which changes the cgroup removal
logic in RemovePath. In the current code, if the initial rmdir has
failed (in this case with EROFS), but the subsequent os.ReadDir returns
ENOENT, it is returned (instead of being ignored -- as the path does not
exist and so there is nothing to remove).
Here is the minimal fix for the issue.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Daniel and Qiang are stepping down as runc maintainers,
not being able to contribute as much as they used to.
Thank you for all the hard work!
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If /run/runc and /usr/bin are on different filesystems, overlayfs may
enable the xino feature which results in the following log message:
kernel: overlayfs: "xino" feature enabled using 3 upper inode bits.
Each time we have to protect /proc/self/exe. So disable xino to remove
the log message (we don't care about the inode numbers of the files
anyway).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In Fedora 41, dnf5 is used and it does not have dnf shell. Let's use
old dnf update; dnf install instead. It is two transactions instead
of one, but dnf5 is faster.
While at it:
- add `--setopt=tsflags=nodocs` as we don't need docs in CI;
- change golang-go to golang as this is a new rpm name;
- remove gcc as it is now required by golang-bin;
- remove container-selinux, criu, fuse-sshfs, iptables from rpms
as they are already installed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is a typo in the comment (ClonedBinary should be CloneBinary), and
the code has changed a bit since then, and it makes more sense to refer
to CloneSelfExe now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
While this is used by the majority of upper container runtimes, it was
not needed for runc itself. Since commit 515f09f7 runc uses overlay,
too, so let's add a check for this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Because we have the overlay solution, we can drop runc-dmz binary
solution since it has too many limitations.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Containerd pre-creates userns and netns before calling runc, which
results in the current code not working when SELinux is enabled,
resulting in the following error:
> runc create failed: unable to start container process: error during
container init: error mounting "mqueue" to rootfs at "/dev/mqueue":
setxattr /path/to/rootfs/dev/mqueue: operation not permitted
The solution is to become root in the user namespace right after
we join it.
Fixes#4466.
Co-authored-by: Wei Fu <fuweid89@gmail.com>
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Co-authored-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
This allows to do
runc update $ID --memory=-1 --memory-swap=$VAL
for cgroup v2, i.e. set memory to unlimited and swap to a specific
value.
This was not possible because ConvertMemorySwapToCgroupV2Value rejected
memory=-1 ("unlimited"). In a hindsight, it was a mistake, because if
memory limit is unlimited, we should treat memory+swap limit as just swap
limit.
Revise the unit test; add description to each case.
Fixes: c86be8a2 ("cgroupv2: fix setting MemorySwap")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Improve readability of ConvertMemorySwapToCgroupV2Value by switching
from a bunch of if statements to a switch, and adding a comment
describing each case.
No functional change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since Go 1.19, the same functionality is there in os/exec package.
As we require go 1.22 now, there's no need to have this.
This basically reverts commit 9258eac0 ("libct/start: use execabs for
newuidmap lookup").
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This reverts commit 65a1074c75.
We needed [1] because when we removed the bindfd logic in [2] we had not
yet moved the binary cloning logic to Go and thus it was necessary to
increase the memory limit in CI because the clone was happening after
joining the cgroup. However, [3] finally moved that code to Go and thus
the cloning is now done outside of the container's cgroup and thus is no
longer accounted as part of the container's memory usage at any point.
Now we can properly support running a simple container with lower memory
usage as we did before.
[1]: commit 65a1074c75 ("increase memory.max in cgroups.bats")
[2]: commit b999376fb2 ("nsenter: cloned_binary: remove bindfd logic entirely")
[3]: commit 0e9a3358f8 ("nsexec: migrate memfd /proc/self/exe logic to Go code")
Signed-off-by: lfbzhm <lifubang@acmcoder.com>
[cyphar: fixed commit messages]
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
There is a race situation when we are opening a file, if there is a
small fd was closed at that time, maybe it will be reused by safeExe.
Because of Go stdlib fds shuffling bug, if the fd of safeExe is too
small, go stdlib will dup3 it to another fd, or dup3 a other fd to this
fd, then it will cause the fd type cmd.Path refers to a random path,
and it can lead to an error "permission denied" when starting the process.
Please see #4294 and <https://github.com/golang/go/issues/61751>.
So we should not use the original fd of safeExe, but use the fd after
shuffled by Go stdlib. Because Go stdlib will guarantee this fd refers to
the correct file.
Signed-off-by: lfbzhm <lifubang@acmcoder.com>
Commit b999376fb2 ("nsenter: cloned_binary: remove bindfd logic
entirely") removed the read-only bind-mount logic from our cloned binary
code because it wasn't really safe because a container with
CAP_SYS_ADMIN could remove the MS_RDONLY bit and get write access to
/proc/self/exe (even with user namespaces this could've been an issue
because it's not clear if the flags are locked).
However, copying a binary does seem to have a minor performance impact.
The only way to have no performance impact would be for the kernel to
block these write attempts, but barring that we could try to reduce the
overhead by coming up with a mount that cannot have it's read-only bits
cleared.
The "simplest" solution is to create a temporary overlayfs using
fsopen(2) which uses the directory where runc exists as a lowerdir,
ensuring that the container cannot access the underlying file -- and we
don't have to do any copies.
While fsopen(2) is not free because mount namespace cloning is usually
expensive (and so it seems like the difference would be marginal), some
basic performance testing seems to indicate there is a ~60% improvement
doing it this way and that it has effectively no overhead even when
compared to just using /proc/self/exe directly:
% hyperfine --warmup 50 \
> "./runc-noclone run -b bundle ctr" \
> "./runc-overlayfs run -b bundle ctr" \
> "./runc-memfd run -b bundle ctr"
Benchmark 1: ./runc-noclone run -b bundle ctr
Time (mean ± σ): 13.7 ms ± 0.9 ms [User: 6.0 ms, System: 10.9 ms]
Range (min … max): 11.3 ms … 16.1 ms 184 runs
Benchmark 2: ./runc-overlayfs run -b bundle ctr
Time (mean ± σ): 13.9 ms ± 0.9 ms [User: 6.2 ms, System: 10.8 ms]
Range (min … max): 11.8 ms … 16.0 ms 180 runs
Benchmark 3: ./runc-memfd run -b bundle ctr
Time (mean ± σ): 22.6 ms ± 1.3 ms [User: 5.7 ms, System: 20.7 ms]
Range (min … max): 19.9 ms … 26.5 ms 114 runs
Summary
./runc-noclone run -b bundle ctr ran
1.01 ± 0.09 times faster than ./runc-overlayfs run -b bundle ctr
1.65 ± 0.15 times faster than ./runc-memfd run -b bundle ctr
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This test case is frequently hanging recently. Might be caused
by a recent kernel update from 5.14.0-427.33.1.el9_4.x86_64 to
5.14.0-427.37.1.el9_4.x86_64.
Could not reproduce locally.
Let's skip it for now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The issue quoted is now fixed, so add some information about the fixed
kernel version, and remove links to older discussions about idmapped
mounts security.
We can actually remove all of it for now, but let's keep it. Change
the skip message to say which kernel is required.
Amends commit b460dc39.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This initWaiter logic was introduced by commit 4ecff8d9, but since the logic of
/proc/self/exe was moved out of runc init in commit 0e9a335, this
seems unnecessary to have initWaiter.
Remove it.
This essentially reverts commit 4ecff8d9.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
While we did set +x when "sealing" regular temporary files, the "is
executable" checks were done before then and would thus fail, causing
the fallback to not work properly.
So just set +x after we create the file. We already have a O_RDWR handle
open when we do the chmod so we won't get permission issues when writing
to the file.
Fixes: e089db3b4a ("dmz: add fallbacks to handle noexec for O_TMPFILE and mktemp()")
Signed-off-by: lifubang <lifubang@acmcoder.com>
We are not really interested in the capabilities of the current process,
so there is no need to load those.
This results in some performance improvement since now the capability
package don't have to parse /proc/self/status.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The userns package was moved to the moby/sys/userns module
at commit 3778ae603c.
This patch deprecates the old location, and adds it as an alias
for the moby/sys/userns package.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Let's point to the relevant README directly in the systemd unit file,
as it is hard to find in the whole nine yards of the runc repo.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The example of starting memfd-bind via systemd in README did not work
for me (Fedora 40, systemd 255):
# systemctl status memfd-bind@/usr/bin/runc
Invalid unit name "memfd-bind@/usr/bin/runc" escaped as "memfd-bind@-usr-bin-runc" (maybe you should use systemd-escape?).
○ memfd-bind@-usr-bin-runc.service
Loaded: bad-setting (Reason: Unit memfd-bind@-usr-bin-runc.service has a bad unit file setting.)
Active: inactive (dead)
Docs: https://github.com/opencontainers/runc
So, let's use systemd-escape -p ("path") in the README example,
and use %f in the systemd unit file to prepend the slash to the
filename.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When preparing to mount container root, we need to make its parent mount
private (i.e. disable propagation), otherwise the new in-container
mounts are leaked to the host.
To find a parent mount, we use to read mountinfo and find the longest
entry which can be a parent of the container root directory.
Unfortunately, due to kernel bug in all Linux kernels older than v5.8
(see [1], [2]), sometimes mountinfo can't be read in its entirety. In
this case, getParentMount may occasionally return a wrong parent mount.
As a result, we do not change the mount propagation to private, and
container mounts are leaked.
Alas, we can not fix the kernel, and reading mountinfo a few times to
ensure its consistency (like it's done in, say, Kubernetes) does not
look like a good solution for performance reasons.
Fortunately, we don't need mountinfo. Let's just traverse the directory
tree, trying to remount it private until we find a mount point (any
error other than EINVAL means we just found it).
Fixes issue 2404.
[1]: https://github.com/kolyshkin/procfs-test
[2]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=9f6c61f96f2d97cbb5f
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Initially, this was a commit to switch from strings.Fields to
strings.SplitN in getCpuUsageBreakdown, since strings.Fields
was probably slower than strings.SplitN in some old Go versions.
Afterwards, strings.Cut was also considered for potential
speed improvements.
After writing a benchmark test, we learned that:
- strings.Fields performance is now adequate;
- strings.SplitN is slower than strings.Fields;
- strings.Cut had <5% performance gain from strings.Fields;
So, remove the TODO and keep the benchmark test.
Signed-off-by: Stavros Panakakis <stavrospanakakis@gmail.com>
The example is too long since it lists too many capabilities.
Simplify it, leaving only two capabilities.
Also, remove ambient capabilities from the set. Inheritable capabilities
were removed earlier by commit 98fe566c, but ambient capabilities can't
be raised without inheritable ones.
Fixes: 98fe566c
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 98fe566c removed inheritable capabilities from the example spec
(used by runc spec) and from the libcontainer/integration test config,
but neglected to also remove ambient capabilities.
An ambient capability could only be set if the same inheritable
capability is set, so as a result of the above change ambient
capabilities were not set (but due to a bug in gocapability package,
those errors are never reported).
Once we start using a library with the fix [1], that bug will become
apparent (both bats-based and libct/int tests will fail).
[1]: https://github.com/kolyshkin/capability/pull/3
Fixes: 98fe566c ("runc: do not set inheritable capabilities")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 98fe566c removed setting inheritable capabilities from runc exec
--cap, but neglected to also remove ambient capabilities.
An ambient capability could only be set if the same inheritable
capability is set, so as a result of the above change ambient
capabilities were not set (but due to a bug in gocapability package,
those errors are never reported).
Once we start using a library with the fix [1], that bug will become
apparent. Alas, we do not have any tests for runc exec --cap, so add
one.
Yet, if some inheritable bits are already set from spec, let's set
ambient to avoid a possible regression. Add a test case for that, too.
[1]: https://github.com/kolyshkin/capability/pull/3
Fixes: 98fe566c ("runc: do not set inheritable capabilities")
Co-authored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
bats-core/bats-action installs a few support libraries by default which are not used by
runc. Disable the installation, which will remove /usr/bin/tar: Permission denied errors.
Signed-off-by: Akhil Mohan <akhilerm@gmail.com>
For some rootless container, runc has no access to cgroup,
But the container is still running. So we should return the
`ErrNotRunning` and `ErrCgroupNotExist` error seperatlly.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Go since 1.21 allows to set a "toolchain" that specifies the minimum Go
toolchain to use when working in runc. In contrast to the go line,
toolchain does not impose a requirement on other modules[1][2].
As documented in the 1.2.0-rc.1 release notes, 1.22.4 is needed for the
nsenter package. Let's suggest this with the toolchain version.
[1]: https://go.dev/doc/toolchain
[2]: https://go.dev/blog/toolchain
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
mig4/setup-bats is now unmaintained(last commit in Sep 2021).
bats-core/bats-action can be used as a replacement maintained
by the bats-core team.
Signed-off-by: Akhil Mohan <akhilerm@gmail.com>
Shared pid namespace means `runc kill` (or `runc delete -f`) have to
kill all container processes, not just init. To do so, it needs a cgroup
to read the PIDs from.
If there is no cgroup, processes will be leaked, and so such
configuration is bad and should not be allowed. To keep backward
compatibility, though, let's merely warn about this for now.
Alas, the only way to know if cgroup access is available is by returning
an error from Manager.Apply. Amend fs cgroup managers to do so (systemd
doesn't need it, since v1 can't work with rootless, and cgroup v2 does
not have a special rootless case).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This aids in failed test analysis by allowing to distinguish the output
of various commands being run as part of the test case from the output
of teardown command like runc delete.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In these cases, this is exactly what we want to find out.
Slightly improves performance and readability.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It turns out that the suid and sgid mode bits are silently ignored by
Linux (though the sticky bit is honoured), and some users are requesting
mode bits that are ignored. While returning an error (as securejoin
does) makes some sense, this is a regression.
Ref: https://github.com/cyphar/filepath-securejoin/issues/23
Fixes: dd827f7b71 ("utils: switch to securejoin.MkdirAllHandle")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We reintroduced this once already because it is quite easy to miss this
subtle aspect of proc mounting. The recent migration to
securejoin.MkdirAllInRoot could have also inadvertently reintroduced
this (though it didn't).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It is available since Go 1.21 and is defined during compile time
(i.e. based on GOARCH during build).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
`signalAllProcesses()` depends on the cgroup and is expected to fail
when runc is running in rootless without an access to the cgroup.
When `RootlessCgroups` is set to `true`, runc just ignores the error
from `signalAllProcesses` and may leak some processes running.
(See the comments in PR 4395)
In the future, runc should walk the process tree to avoid such a leak.
Note that `RootlessCgroups` is a misnomer; it is set to `false` despite
the name when cgroup v2 delegation is configured.
This is expected to be renamed in a separate commit.
Fix issue 4394
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
filepath-securejoin has a bunch of extra hardening features and is very
well-tested, so we should use it instead of our own homebrew solution.
A lot of rootfs_linux.go callers pass a SecureJoin'd path, which means
we need to keep the wrapper helpers in utils, but at least the core
logic is no longer in runc. In future we will want to remove this dodgy
logic and just use file handles for everything (using libpathrs,
ideally).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While we use SecureJoin to try to make all of our target paths inside
the container safe, SecureJoin is not safe against an attacker than can
change the path after we "resolve" it.
os.MkdirAll can inadvertently follow symlinks and thus an attacker could
end up tricking runc into creating empty directories on the host (note
that the container doesn't get access to these directories, and the host
just sees empty directories). However, this could potentially cause DoS
issues by (for instance) creating a directory in a conf.d directory for
a daemon that doesn't handle subdirectories properly.
In addition, the handling for creating file bind-mounts did a plain
open(O_CREAT) on the SecureJoin'd path, which is even more obviously
unsafe (luckily we didn't use O_TRUNC, or this bug could've allowed an
attacker to cause data loss...). Regardless of the symlink issue,
opening an untrusted file could result in a DoS if the file is a hung
tty or some other "nasty" file. We can use mknodat to safely create a
regular file without opening anything anyway (O_CREAT|O_EXCL would also
work but it makes the logic a bit more complicated, and we don't want to
open the file for any particular reason anyway).
libpathrs[1] is the long-term solution for these kinds of problems, but
for now we can patch this particular issue by creating a more restricted
MkdirAll that refuses to resolve symlinks and does the creation using
file descriptors. This is loosely based on a more secure version that
filepath-securejoin now has[2] and will be added to libpathrs soon[3].
[1]: https://github.com/openSUSE/libpathrs
[2]: https://github.com/cyphar/filepath-securejoin/releases/tag/v0.3.0
[3]: https://github.com/openSUSE/libpathrs/issues/10
Fixes: CVE-2024-45310
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We recently switched VERSION to be read from env vars (#4270). This
broke several projects, as they were building runc and using a `VERSION`
env var for, e.g. the containerd version.
When fixing that in #4370, we discussed to consider doing the same for
these variables too
(https://github.com/opencontainers/runc/pull/4370#pullrequestreview-2240030944).
Let's stop reading them from env vars, as it is very easy to do it by
mistake (e.g. compile runc and define a COMMIT env var, not to override
the commit shown in `runc --version`) and users that want can still
override them if they want to. For example, with:
make EXTRA_BUILDTAGS=runc_nodmz
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Add this new make variable so users can specify build information
without modifying the runc version nor the source code.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
The following commands are moved from `contrib/cmd` to `tests/cmd`:
- fs-idmap
- pidfd-kill
- recvtty
- remap-rootfs
- sd-helper
- seccompagent
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
This reverts commit 9d9273c926.
This commit broke the build for several other projects (see comments
here: https://github.com/opencontainers/runc/pull/4270, after the merge)
and we don't really need this to be able to set the version without
changing the file.
With this commit reverted, we can still run:
make VERSION="1.2.3"
and it just works. It doesn't take it from an env variable, but that is
what broke all the other projects (VERSION is just too generic as an env
var, especially for a project like runc that is embedded in many
others).
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
The warnings fixed were:
libcontainer/configs/config_test.go:205:12: printf: non-constant format string in call to (*testing.common).Errorf (govet)
t.Errorf(fmt.Sprintf("Expected error to not occur but it was %+v", err))
^
libcontainer/cgroups/fs/blkio_test.go:481:13: printf: non-constant format string in call to (*testing.common).Errorf (govet)
t.Errorf(fmt.Sprintf("test case '%s' failed unexpectedly: %s", testCase.desc, err))
^
libcontainer/cgroups/fs/blkio_test.go:595:13: printf: non-constant format string in call to (*testing.common).Errorf (govet)
t.Errorf(fmt.Sprintf("test case '%s' failed unexpectedly: %s", testCase.desc, err))
^
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
- drop Go 1.21;
- add Go 1.23;
- for a few jobs that were using Go 1.21, switch to 1.22;
Also, bump go to 1.22 in go.mod.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Rename current -> got, expected -> want.
2. check_cgroup_value: add file name to output.
3. Improve functions description.
This is mostly to simplify debugging test failures.
Example output before:
current 500000 !? 500
After:
cpu.max.burst: got 500000, want 500
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In all the three cases, we check that the program returned non-zero exit
code. This can be done in a much simpler manner.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The logic for how we create mountpoints is spread over each mountpoint
preparation function, when in reality the behaviour is pretty uniform
with only a handful of exceptions. So just move it all to one function
that is easier to understand.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Add a CI job to ensure go fix produces no result. Quoting
`go doc cmd/fix`:
> Fix finds Go programs that use old APIs and rewrites them to use newer
> ones. After you update to a new Go release, fix helps make the
> necessary changes to your programs.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Remove CentOS 7 as it is EOL.
Add back RHEL 8 clone (CentOS Stream 8 was removed by commit
40bb9c468e).
Switch from CentOS Stream 9 to Almalinux 9.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Remove changes that are already reflected in v1.1.13 changelog:
- rlimit_nofile fix;
- rt_period vs rt_runtime fix;
- gpg vs keyboxd fix;
- nsexec debug log fix;
- fips faking;
- vagrant Fedora 39 bump;
- golangci-lint bump;
- x/net bump;
- centos stream 8 removal;
- codespell ci fixes.
Compact some of the entries that are related (e.g. about actuated-ci).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now that we dropped support for go < 1.21, we can use this; moving
the sync.once out of the runningInUserNS() implementation would also
allow for it to be more easily tested if we'd decide to.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The fuzzer for this only runs on Linux; rename the file to be Linux-only
so that we don't have to stub out the uidMapInUserNS function.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This was a poor decision on my side; 4316df8b53
moved this utility to a separate package, and split the exported function
from the implementation (and stubs). Out of convenience, I used an alias
for the latter part, but there's two downsides to that;
- `RunningInUserNS` being an exported var means that (technically) it can
be replaced by other code; perhaps that's a "feature", but not one we
intended it to be used for.
- `RunningInUserNS` being implemented through a var / alias means it's
also documented as such on [pkg.go.dev], which is confusing.
This patch changes it to a regular function, acting as a wrapper for
the underlying implementations. While at it, also slightly touching
up the GoDoc to describe its functionality / behavior.
[pkg.go.dev]: https://pkg.go.dev/github.com/opencontainers/runc@v1.1.13/libcontainer/userns#RunningInUserNS
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Commit 4316df8b53 isolated RunningInUserNS
to a separate package to make it easier to consume without bringing in
additional dependencies, and with the potential to move it separate in
a similar fashion as libcontainer/user was moved to a separate module
in commit ca32014adb. While RunningInUserNS
is fairly trivial to implement, it (or variants of this utility) is used
in many codebases, and moving to a separate module could consolidate
those implementations, as well as making it easier to consume without
large dependency trees (when being a package as part of a larger code
base).
Commit 1912d5988b and follow-ups introduced
cgo code into the userns package, and code introduced in those commits
are not intended for external use, therefore complicating the potential
of moving the userns package separate.
This commit moves the new code to a separate package; some of this code
was included in v1.1.11 and up, but I could not find external consumers
of `GetUserNamespaceMappings` and `IsSameMapping`. The `Mapping` and
`Handles` types (added in ba0b5e2698) only
exist in main and in non-stable releases (v1.2.0-rc.x), so don't need
an alias / deprecation.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Since Michael Crosby is stepping down, and we don't want to nominate
someone else to be a chief maintainer, let's remove the position.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I talked to Michael, he says he is stepping down as a maintainer, being
busy with other stuff.
Thank you for all the hard work that you did!
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
An error from strconv.Atoi already contains the text it fails to parse.
Because of that, errors look way too verbose, e.g.:
[root@kir-rhat runc-tst]# ./runc exec --user 1:1:1 2345 true
ERRO[0000] exec failed: parsing 1:1 as int for gid failed: strconv.Atoi: parsing "1:1": invalid syntax
With this patch, the error looks like this now:
[root@kir-rhat runc]# ./runc exec --user 1:1:1 2345 true
ERRO[0000] exec failed: bad gid: strconv.Atoi: parsing "1:1": invalid syntax
Still not awesome, but better.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Do not ask for the same option value twice.
2. For tty, we always want false, unless specified, and this is what
GetBool gets us.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
By definition, every container has only 1 init (i.e. PID 1) process.
Apparently, libcontainer API supported running more than 1 init, and
at least one tests mistakenly used it.
Let's not allow that, erroring out if we already have init. Doing
otherwise _probably_ results in some confusion inside the library.
Fix two cases in libct/int which ran two inits inside a container.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. The code to call c.exec from c.Run was initially added by commit
3aacff695. At the time, there was a lock in c.Run. That lock was
removed by commit bd3c4f84, which resulted in part of c.Run executing
without the lock.
2. All the Start/Run/Exec calls were a mere wrappers for start/run/exec
adding a lock, but some more code crept into Start at some point,
e.g. by commits 805b8c73 and 108ee85b8. Since the reason mentioned in
commit 805b8c73 is no longer true after refactoring, we can fix this.
Fix both issues by moving code out of wrappers, and adding locking into
c.Run.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case file already exists, mknod(2) will return EEXIST.
This os.Stat call was (inadvertently?) added by commit 805b8c73.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There's too much logic here figuring out which CPUs to use. Runc is a
low level tool and is not supposed to be that "smart". What's worse,
this logic is executed on every exec, making it slower. Some of the
logic in (*setnsProcess).start is executed even if no annotation is set,
thus making ALL execs slow.
Also, this should be a property of a process, rather than annotation.
The plan is to rework this.
This reverts commit afc23e3397.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Get the image from Fedora directly.
Also, remove the comment about cgroup v2 as it is also tested on Ubuntu
22.04 now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It has been pointed out that some controllers can not accept multiple
lines of output at once. In particular, io.max can only set one device
at a time.
Practically, the only multi-line resource values we can get come from
unified.* -- let's write those line by line.
Add a test case.
Reported-by: Tao Shen <shentaoskyking@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I am nominating @rata for the role of runc maintainer.
He is pretty active in the project, did some substantial work
and is helping with PR review and releases.
As noted in MAINTAINERS_GUIDE.md, we have a week to vote, and need to
get 66% of current maintainers' votes.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It is not needed since Go 1.20 (which was released in February 2023 and
is no longer supported since February 2024).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Go 1.20 was released in February 2023 and is no longer supported since
February 2024. Time to move on.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now when Go 1.22.4 is out it should no longer be a problem.
Leave Go 1.21 for CentOS testing (CentOS 7 and 8 have older glibc)
and Dockerfile (Debian 11 have older glibc).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Go 1.23 includes a fix (https://go.dev/cl/587919) so go1.23.x can be
used. This fix is also backported to 1.22.4, so go1.22.x can also be
used (when x >= 4). Finally, for glibc >= 2.32 it doesn't really matter.
Add a note about Go 1.22.x > 1.22.4 to README as well.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There are cgroup v2 systems out there that do not have cgroup swap enabled,
and this test will probably fail in there.
Move it to a separate case, guarded with requires cgroups_swap.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When swap is being disabled (as set to 0), or set to max, ignore
non-existent memory.swap.max cgroup file.
If swap is being set explicitly to some value, do return an error like
before.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Add "-maxdepth 2" to not dive too deep into cgroup hierarchy.
2. Add "-type f" to look for a regular file.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I used script/keyring_validate.sh, which gave me this error:
> [*] User cyphar in runc.keyring is not a maintainer!
Apparently, when gnupg 2.4.1+ sees a fresh install (i.e. no ~/.gnupg
directory), it configures itself to use keyboxd instead of keyring
files, and when just silently ignores options like --keyring and
--no-default-keyring, working with keyboxd all the time.
The only way I found to make it not use keyboxd is to set --homedir.
Let's do that when we explicitly want a separate keyring.
Similar change is made to script/release_key.sh.
Also, change "--import --import-options=show-only" to "--show-keys"
which is a shortcut. When using this, there is no need to protect
the default keyring since this command does not read or modify it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test case fails when there's a single CPU. Fix this by adding
"require smp".
While at it, document the test case and add a FIXME to maybe remove
this test later.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Go 1.23 tightens access to internal symbols, and even puts runc into
"hall of shame" for using an internal symbol (recently added by commit
da68c8e3). So, while not impossible, it becomes harder to access those
internal symbols, and it is a bad idea in general.
Since Go 1.23 includes https://go.dev/cl/588076, we can clean the
internal rlimit cache by setting the RLIMIT_NOFILE for ourselves,
essentially disabling the rlimit cache.
Once Go 1.22 is no longer supported, we will remove the go:linkname hack.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test panics if userns is detected (such as when run in a rootless
docker container) because SetV1 does nothing in this case.
We could fix the panic, but it doesn't make sense to run the test at
all.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The issue is the same as in commit 1b2adcf but for RT scheduler;
the fix is also the same.
Test case by ls-ggg.
Co-authored-by: ls-ggg <335814617@qq.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I tried to fix it, but it looks like older CentOS 7 kernel is the
ultimate reason why it doesn't work.
So, remove FIXME and add some explanation.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since we're now testing on ARM, the test case fails when trying to do
pre-dump since MemTrack is not available.
Skip the pre-dump part if so.
This also reverts part of commit 3f4a73d6 as it is no longer needed
(now, instead of skipping the whole test, we're just skipping the
pre-dump).
[Review with --ignore-all-space]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit c77aaa3f the tail of criu log is printed by runc, so
there's no need to do the same thing in tests.
This also fixes a test failure on ARM where showLog fails (because
there's no log file) and thus the conditional t.Skip is not called.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
criu check --feature userns also tests for the /proc/self/ns/user
presense, so remove the redundant check, and simplify the error message.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since criu 2.12, rpcOpts is not needed when checking criu features.
As we requires criu >= 3.0 in Checkpoint, we can remove rpcOpts.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As reported in issue #4195, the new version(since 1.19) of go runtime
will cache rlimit-nofile. Before executing execve, the rlimit-nofile
of the process will be restored with the cache. In runc, this will
cause the rlimit-nofile set by the parent process for the container
to become invalid. It can be solved by clearing the cache.
Signed-off-by: ls-ggg <335814617@qq.com>
(cherry picked from commit f9f8abf310)
Signed-off-by: lifubang <lifubang@acmcoder.com>
Note that github-actions output format is deprecated and no longer supported,
and it is also no longer needed since setup-go problem matcher already
handles default golangci-lint output format ("colored-line-number").
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If CPU burst knob is non-existent, the current implementation (added in
commit e1584831) still tries to set it again after setting the new CPU
quota, which is useless (and we have to ignore ENOENT again).
Fix this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Earlier, commit fce8dd4d already increased this timeout from 1 to 5 seconds.
Yet, I just saw this timeout being hit in actuated-arm CI.
Increase the timeout again, this time from 5 to 50 (100 * 0.5) seconds.
Also, use wait_pids_gone, and improve some comments.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
They were failing earlier but are working now.
This includes a fix to criu repo path assignment so it works for actuated case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. A few tests use "criu check --feature" to check for a specific
feature. Let's generalize it.
2. Fix "checkpoint --pre-dump and restore" test to require memory
tracking (which is missing on ARM).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit c77aaa3f the tail of criu.log is printed by runc, so
there's no need to do the same thing in tests.
Related to 3711, 3816.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
kcmp is often missing: https://man7.org/linux/man-pages/man2/kcmp.2.html
> Before Linux 5.12, this system call is available only if the
> kernel is configured with CONFIG_CHECKPOINT_RESTORE, since the
> original purpose of the system call was for the
> checkpoint/restore in user space (CRIU) feature. (The
> alternative to this system call would have been to expose
> suitable process information via the proc(5) filesystem; this was
> deemed to be unsuitable for security reasons.) Since Linux 5.12,
> this system call is also available if the kernel is configured
> with CONFIG_KCMP.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Commit b6967fa84c moved the functionality of managing cgroup devices
into a separate package, and decoupled libcontainer/cgroups from it.
Yet, some software (e.g. cadvisor) may need to use libcontainer package,
which imports libcontainer/cgroups/devices, thus making it impossible to
use libcontainer without bringing in cgroup/devices dependency.
In fact, we only need to manage devices in runc binary, so move the
import to main.go.
The need to import libct/cg/dev in order to manage devices is already
documented in libcontainer/cgroups, but let's
- update that documentation;
- add a similar note to libcontainer/cgroups/systemd;
- add a note to libct README.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This handles a corner case when joining a container having all
the processes running exclusively on isolated CPU cores to force
the kernel to schedule runc process on the first CPU core within the
cgroups cpuset.
The introduction of the kernel commit
46a87b3851f0d6eb05e6d83d5c5a30df0eca8f76 has affected this deterministic
scheduling behavior by distributing tasks across CPU cores within the
cgroups cpuset. Some intensive real-time application are relying on this
deterministic behavior and use the first CPU core to run a slow thread
while other CPU cores are fully used by real-time threads with SCHED_FIFO
policy. Such applications prevents runc process from joining a container
when the runc process is randomly scheduled on a CPU core owned by a
real-time thread.
Introduces isolated CPU affinity transition OCI runtime annotation
org.opencontainers.runc.exec.isolated-cpu-affinity-transition to restore
the behavior during runc exec.
Fix issue with kernel >= 6.2 not resetting CPU affinity for container processes.
Signed-off-by: Cédric Clerget <cedric.clerget@gmail.com>
The motivation behind this change is to provide a flexible mechanism for
containers within a Kubernetes cluster to opt out of FIPS mode when necessary.
This change enables apps to simulate FIPS mode being enabled or disabled for testing
purposes. Users can control whether apps believe FIPS mode is on or off by manipulating
`/proc/sys/crypto/fips_enabled`.
Signed-off-by: Sohan Kunkerkar <sohank2602@gmail.com>
When we run this:
mount --bind -o remount,diratime,strictatime "$DIR"
It fails in debian testing, when it is the second time we call this
function in the same bats test (i.e. when $DIR is defined already).
strace shows this syscall failing:
mount_setattr(3, "", AT_EMPTY_PATH, {attr_set=MOUNT_ATTR_NOSUID|MOUNT_ATTR_NODEV|MOUNT_ATTR_NOEXEC|MOUNT_ATTR_NOATIME|MOUNT_ATTR_STRICTATIME, attr_clr=MOUNT_ATTR_RDONLY|MOUNT_ATTR_NOATIME|MOUNT_ATTR_STRICTATIME|MOUNT_ATTR_NODIRATIME|0x40, propagation=0 /* MS_??? */, userns_fd=0}, 32) = -1 EINVAL (Invalid argument)
Note it has `MOUNT_ATTR_NOATIME` and `MOUNT_ATTR_STRICTATIME` which
probably causes it to return EINVAL.
This patch simply adds atime to the options, so the mount command now
works and fixes most of the tests in debian testing.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
relatime is not shown on some debian systems. Let's check that no other
setting that removes the relatime effect is set, as that should be
enough too.
For more info, see the issue linked in the comments.
Signed-off-by: Rodrigo Campos <rodrigo@sdfg.com.ar>
Prior to this commit, commands like `runc update --cpuset-cpus=1`
were implying to set cpu burst to "0" (which does not mean "leave it as is").
This was failing when the kernel does not support cpu burst:
`openat2 /sys/fs/cgroup/runc-cgroups-integration-test/test-cgroup-22167/cpu.max.burst: no such file or directory`
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
In issue #4210, if we don't provide `--cpu-burst` in `runc update`,
the value of cpu burst will always set to 0.
Signed-off-by: lifubang <lifubang@acmcoder.com>
Aleksa Sarai (5):
VERSION: back to development
VERSION: release v1.2.0-rc.1
changelog: update to include all new changes since 1.1.0
changelog: sync changelog entries up to runc 1.1.12
changelog: mention key breaking changes for mount options
LGTMs: lifubang AkihiroSuda kolyshkin cyphar
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
[ cyphar: restructuring and removal of outdated or incorrect info ]
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
1. Sort the list of matrix excludes in the order of matrix,
add comments explaining why we disable some jobs.
2. Exclude some jobs:
- runc_nodmz && go 1.20.x
- runc_nodmz && -race
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We will almost certainly need to eventually rework nsenter to:
1. Figure out a way to make pthread_self() not break after nsenter runs
(probably not possible, because the core issue is likely that we are
ignoring the rules of signal-safety(7)); or
2. Do an other re-exec of /proc/self/exe to execute the Go half of
"runc init" -- after we've done the nsenter setup. This would reset
all of the process state and ensure we have a clean glibc state for
Go, but it would make runc slower...
For now, just block Go 1.22 builds to avoid having broken runcs floating
around until we resolve the issue. It seems possible for musl to also
have an issue, but it appears to work and so for now just block glibc
builds.
Note that this will only block builds for anything that uses nsenter --
so users of our (internal) libcontainer libraries should be fine. Only
users that are starting containers using nsenter to actually start
containers will see the error (which is precisely what we want).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It turns out that on ppc64le (at least), Docker doesn't include any
architectures in the list of allowed architectures. libseccomp
interprets this as "just include the default architecture" but patchbpf
would return a no-op ENOSYS stub, which would lead to the exact issues
that commit 7a8d7162f9 ("seccomp: prepend -ENOSYS stub to all
filters") fixed for other architectures.
So, just always include the running architecture in the list. There's
no real downside.
Ref: https://bugzilla.suse.com/show_bug.cgi?id=1192051#c6
Reported-by: Fabian Vogt <fvogt@suse.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Calling the Linux AUDIT_* architecture constants "native" leads to
confusing code when we are getting the actual native architecture of the
running system.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The referenced issue was fixed in `github.com/urfave/cli` v1.22.6. We
can now remove the dependabot ignore for this package.
Signed-off-by: SuperQ <superq@gmail.com>
Now that runc-dmz is opt-in, we no longer need to try to detect whether
SELinux would cause issues for us. We can also remove the
special-purpose build-tag we added.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If it is compiled, the user needs to opt-in with this env variable to
use it.
While we are there, remove the RUNC_DMZ=legacy as that is now the
default.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
It takes some time for the kernel to kill the process (and remove its
PID from cgroup.procs). To ensure we don't have flakes from reading
cgroup.procs right after the kill, check and wait for processes to
actually be gone.
Fixes: 4163
Reported-by: lifubang@acmcoder.com
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
* Simple error correction of a spelling mistake which was
introduced at commit b8f75f3
Signed-off-by: Sjoerd van Leent <sjoerd.van.leent@alliander.com>
This expression is specific to GNU awk (gawk), so if someone has other version
of awk installed, this won't work and it's not easy to see why.
Explicitly requiring gawk here is better.
Revert "tests/int/helpers: gawk -> awk"
This reverts commit 4e65118d02.
Reported-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was added by commit 9c444070 (to use LONG_MAX and INT_MAX) but the
code was later removed by commit ba0b5e26.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We close the logfd before execve so there's no need to special case it.
In addition, it turns out that (*os.File).Fd() doesn't handle the case
where the file was closed and so it seems suspect to use that kind of
check.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Given the core issue in GHSA-xr7r-f8xq-vfvv was that we were unknowingly
leaking file descriptors to "runc init", it seems prudent to make sure
we proactively prevent this in the future. The solution is to simply
mark all non-stdio file descriptors as O_CLOEXEC before we spawn "runc
init".
For libcontainer library users, this could result in unrelated files
being marked as O_CLOEXEC -- however (for the same reason we are doing
this for runc), for security reasons those files should've been marked
as O_CLOEXEC anyway.
Fixes: GHSA-xr7r-f8xq-vfvv CVE-2024-21626
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We auto-close this file descriptor in the final exec step, but it's
probably a good idea to not possibly leak the file descriptor to "runc
init" (we've had issues like this in the past) especially since it is a
directory handle from the host mount namespace.
In practice, on runc 1.1 this does leak to "runc init" but on main the
handle has a low enough file descriptor that it gets clobbered by the
ForkExec of "runc init".
OPEN_TREE_CLONE would let us protect this handle even further, but the
performance impact of creating an anonymous mount namespace is probably
not worth it.
Also, switch to using an *os.File for the handle so if it goes out of
scope during setup (i.e. an error occurs during setup) it will get
cleaned up by the GC.
Fixes: GHSA-xr7r-f8xq-vfvv CVE-2024-21626
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If we leak a file descriptor referencing the host filesystem, an
attacker could use a /proc/self/fd magic-link as the source for execve
to execute a host binary in the container. This would allow the binary
itself (or a process inside the container in the 'runc exec' case) to
write to a host binary, leading to a container escape.
The simple solution is to make sure we close all file descriptors
immediately before the execve(2) step. Doing this earlier can lead to very
serious issues in Go (as file descriptors can be reused, any (*os.File)
reference could start silently operating on a different file) so we have
to do it as late as possible.
Unfortunately, there are some Go runtime file descriptors that we must
not close (otherwise the Go scheduler panics randomly). The only way of
being sure which file descriptors cannot be closed is to sneakily
go:linkname the runtime internal "internal/poll.IsPollDescriptor"
function. This is almost certainly not recommended but there isn't any
other way to be absolutely sure, while also closing any other possible
files.
In addition, we can keep the logrus forwarding logfd open because you
cannot execve a pipe and the contents of the pipe are so restricted
(JSON-encoded in a format we pick) that it seems unlikely you could even
construct shellcode. Closing the logfd causes issues if there is an
error returned from execve.
In mainline runc, runc-dmz protects us against this attack because the
intermediate execve(2) closes all of the O_CLOEXEC internal runc file
descriptors and thus runc-dmz cannot access them to attack the host.
Fixes: GHSA-xr7r-f8xq-vfvv CVE-2024-21626
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If a file descriptor of a directory in the host's mount namespace is
leaked to runc init, a malicious config.json could use /proc/self/fd/...
as a working directory to allow for host filesystem access after the
container runs. This can also be exploited by a container process if it
knows that an administrator will use "runc exec --cwd" and the target
--cwd (the attacker can change that cwd to be a symlink pointing to
/proc/self/fd/... and wait for the process to exec and then snoop on
/proc/$pid/cwd to get access to the host). The former issue can lead to
a critical vulnerability in Docker and Kubernetes, while the latter is a
container breakout.
We can (ab)use the fact that getcwd(2) on Linux detects this exact case,
and getcwd(3) and Go's Getwd() return an error as a result. Thus, if we
just do os.Getwd() after chdir we can easily detect this case and error
out.
In runc 1.1, a /sys/fs/cgroup handle happens to be leaked to "runc
init", making this exploitable. On runc main it just so happens that the
leaked /sys/fs/cgroup gets clobbered and thus this is only consistently
exploitable for runc 1.1.
Fixes: GHSA-xr7r-f8xq-vfvv CVE-2024-21626
Co-developed-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
[refactored the implementation and added more comments]
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While it doesn't make much of a practical difference, it seems far more
reasonable to use os.NewFile to wrap all of our passed file descriptors
to make sure they're tracked by the Go runtime and that we don't
double-close them.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This error code is using functions that are present in nolibc too.
When using nolibc, the error is printed like:
exec /runc.armel: errno=8
When using libc, as its perror() implementation translates the errno to
a message, it is printed like:
exec /runc.armel: exec format error
Note that when using libc, the error is printed in the same way as
before.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
It seems that newer podman versions need the kernel comment flag too.
By podman run, iptables using -m comment in the iptables-command to add the corresponding network rules.
Signed-off-by: Christian Happ <Christian.Happ@jumo.net>
This gives a more clear error message when idmap mounts are not
supported on the source filesystem. For example, a k8s user will see
this now in kubectl describe pod:
Warning Failed 2s (x2 over 4s) kubelet, 127.0.0.1 Error: failed to create containerd task: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: failed to fulfil mount request: failed to set MOUNT_ATTR_IDMAP on /var/lib/kubelet/pods/f037a704-742c-40fe-8dbf-17ed9225c4df/volumes/kubernetes.io~empty-dir/hugepage: invalid argument (maybe the source filesystem doesn't support idmap mounts on this kernel?): unknown
This gives a hint on where to look at.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Looking at the code, I found out that kernel_lt function was actually
doing le ("less than or equal") rather than lt ("less than") operation.
Let's fix this and do exactly what the name says.
A bigger issue is, the function use was not consistent (some uses
implied "less than or equal").
To fix the usage, find out all relevant kernel commits and kernel
versions that have them (the latter is done using "git describe
--contains $sha"), and fix the wrong cases. While at it, add references
to all these kernel commits for the future generations of
check-config.sh hackers.
Also, add kernel_ge function which is the opposite of kernel_lt,
and document both.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
...when the stdout is not a terminal, and also when NO_COLOR environment
variable is set to any non-empty value (as per no-color.org).
Co-authored-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kernel commit e55b9f96860f (which made its way into Linux v6.1-rc1)
removes CONFIG_MEMCG_SWAP entirely, so there's no sense to check for in
on newer kernels.
Make the check conditional.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using ints for all of our mapping structures means that a 32-bit binary
errors out when trying to parse /proc/self/*id_map:
failed to cache mappings for userns: failed to parse uid_map of userns /proc/1/ns/user:
parsing id map failed: invalid format in line " 0 0 4294967295": integer overflow on token 4294967295
This issue was unearthed by commit 1912d5988b ("*: actually support
joining a userns with a new container") but the underlying issue has
been present since the docker/libcontainer days.
In theory, switching to uint32 (to match the spec) instead of int64
would also work, but keeping everything signed seems much less
error-prone. It's also important to note that a mapping might be too
large for an int on 32-bit, so we detect this during the mapping.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Our previous implementation of idmapped mounts and bind-mount sources
would open all of the source paths before we did any mounts, meaning
that mounts using sources from inside the container rootfs would not be
correct.
This has been fixed with the new on-demand system, and so add some
regression tests.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
ridmap indicates that the id mapping should be applied recursively (only
really relevant for rbind mount entries), and idmap indicates that it
should not be applied recursively (the default). If no mappings are
specified for the mount, we use the userns configuration of the
container. This matches the behaviour in the currently-unreleased
runtime-spec.
This includes a minor change to the state.json serialisation format, but
because there has been no released version of runc with commit
fbf183c6f8 ("Add uid and gid mappings to mounts"), we can safely make
this change without affecting running containers. Doing it this way
makes it much easier to handle m.IsIDMapped() and indicating that a
mapping has been specified.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If a user specifies a configuration like "rro, rrw", we should have
similar behaviour to "ro, rw" where we clear the previous flags so that
the last specified flag takes precendence.
Fixes: 382eba4354 ("Support recursive mount attrs ("rro", "rnosuid", "rnodev", ...)")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Our previous test for whether we can mount on top of /proc incorrectly
assumed that it would only be called with bind-mount sources. This meant
that having a non bind-mount entry for a pseudo-filesystem (like
overlayfs) with a dummy source set to /proc on the host would let you
bypass the check, which could easily lead to security issues.
In addition, the check should be applied more uniformly to all mount
types, so fix that as well. And add some tests for some of the tricky
cases to make sure we protect against them properly.
Fixes: 331692baa7 ("Only allow proc mount if it is procfs")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
With the idmap work, we will have a tainted Go thread in our
thread-group that has a different mount namespace to the other threads.
It seems that (due to some bad luck) the Go scheduler tends to make this
thread the thread-group leader in our tests, which results in very
baffling failures where /proc/self/mountinfo produces gibberish results.
In order to avoid this, switch to using /proc/thread-self for everything
that is thread-local. This primarily includes switching all file
descriptor paths (CLONE_FS), all of the places that check the current
cgroup (technically we never will run a single runc thread in a separate
cgroup, but better to be safe than sorry), and the aforementioned
mountinfo code. We don't need to do anything for the following because
the results we need aren't thread-local:
* Checks that certain namespaces are supported by stat(2)ing
/proc/self/ns/...
* /proc/self/exe and /proc/self/cmdline are not thread-local.
* While threads can be in different cgroups, we do not do this for the
runc binary (or libcontainer) and thus we do not need to switch to
the thread-local version of /proc/self/cgroups.
* All of the CLONE_NEWUSER files are not thread-local because you
cannot set the usernamespace of a single thread (setns(CLONE_NEWUSER)
is blocked for multi-threaded programs).
Note that we have to use runtime.LockOSThread when we have an open
handle to a tid-specific procfs file that we are operating on multiple
times. Go can reschedule us such that we are running on a different
thread and then kill the original thread (causing -ENOENT or similarly
confusing errors). This is not strictly necessary for most usages of
/proc/thread-self (such as using /proc/thread-self/fd/$n directly) since
only operating on the actual inodes associated with the tid requires
this locking, but because of the pre-3.17 fallback for CentOS, we have
to do this in most cases.
In addition, CentOS's kernel is too old for /proc/thread-self, which
requires us to emulate it -- however in rootfs_linux.go, we are in the
container pid namespace but /proc is the host's procfs. This leads to
the incredibly frustrating situation where there is no way (on pre-4.1
Linux) to figure out which /proc/self/task/... entry refers to the
current tid. We can just use /proc/self in this case.
Yes this is all pretty ugly. I also wish it wasn't necessary.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The primary change is a switch to using /proc/thread-self, which is
needed for when we add a CLONE_FS thread to runc.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
With the rework of nsexec.c to handle MOUNT_ATTR_IDMAP in our Go code we
can now handle arbitrary mappings without issue, so remove the primary
artificial limit of mappings (must use the same mapping as the
container's userns) and add some tests.
We still only support idmap mounts for bind-mounts because configuring
mappings for other filesystems would require switching our entire mount
machinery to the new mount API. The current design would easily allow
for this but we would need to convert new mount options entirely to the
fsopen/fsconfig/fsmount API. This can be done in the future.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
With open_tree(OPEN_TREE_CLONE), it is possible to implement both the
id-mapped mounts and bind-mount source file descriptor logic entirely in
Go without requiring any complicated handling from nsexec.
However, implementing it the naive way (do the OPEN_TREE_CLONE in the
host namespace before the rootfs is set up -- which is what the existing
implementation did) exposes issues in how mount ordering (in particular
when handling mount sources from inside the container rootfs, but also
in relation to mount propagation) was handled for idmapped mounts and
bind-mount sources. In order to solve this problem completely, it is
necessary to spawn a thread which joins the container mount namespace
and provides mountfds when requested by the rootfs setup code (ensuring
that the mount order and mount propagation of the source of the
bind-mount are handled correctly). While the need to join the mount
namespace leads to other complicated (such as with the usage of
/proc/self -- fixed in a later patch) the resulting code is still
reasonable and is the only real way to solve the issue.
This allows us to reduce the amount of C code we have in nsexec, as well
as simplifying a whole host of places that were made more complicated
with the addition of id-mapped mounts and the bind sourcefd logic.
Because we join the container namespace, we can continue to use regular
O_PATH file descriptors for non-id-mapped bind-mount sources (which
means we don't have to raise the kernel requirement for that case).
In addition, we can easily add support for id-mappings that don't match
the container's user namespace. The approach taken here is to use Go's
officially supported mechanism for spawning a process in a user
namespace, but (ab)use PTRACE_TRACEME to avoid actually having to exec a
different process. The most efficient way to implement this would be to
do clone() in cgo directly to run a function that just does
kill(getpid(), SIGSTOP) -- we can always switch to that if it turns out
this approach is too slow. It should be noted that the included
micro-benchmark seems to indicate this is Fast Enough(TM):
goos: linux
goarch: amd64
pkg: github.com/opencontainers/runc/libcontainer/userns
cpu: Intel(R) Core(TM) i5-10210U CPU @ 1.60GHz
BenchmarkSpawnProc
BenchmarkSpawnProc-8 1670 770065 ns/op
Fixes: fda12ab101 ("Support idmap mounts on volumes")
Fixes: 9c444070ec ("Open bind mount sources from the host userns")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It turns out that the error added in commit 09822c3da8 ("configs:
disallow ambiguous userns and timens configurations") causes issues with
containerd and CRIO because they pass both userns mappings and a userns
path.
These configurations are broken, but to avoid the regression in this one
case, output a warning to tell the user that the configuration is
incorrect but we will continue to use it if and only if the configured
mappings are identical to the mappings of the provided namespace.
Fixes: 09822c3da8 ("configs: disallow ambiguous userns and timens configurations")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Previously, all of our userns tests worked around the remapping issue by
creating the paths that runc would attempt to create (like /proc).
However, this isn't really accurate to how real userns containers are
created, so it's much better to actually remap the rootfs.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Given we've had several bugs in this behaviour that have now been fixed,
add an integration test that makes sure that you can start a container
that joins all of the namespaces of a second container.
The only namespace we do not join is the mount namespace, because
joining a namespace that has been pivot_root'd leads to a bunch of
errors. In principle, removing everything from config.json that requires
a mount _should_ work, but the root.path configuration is mandatory and
we cannot just ignore setting up the rootfs in the namespace joining
scenario (if the user has configured a different rootfs, we need to use
it or error out, and there's no reasonable way of checking if if the
rootfs paths are the same that doesn't result in spaghetti logic).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The owner of /proc/self/timens_offsets doesn't change after creating a
userns, meaning that we need to request stage-0 to write our timens
mappings for us. Before this patch, attempting to use timens with a
proper userns resulted in:
FATA[0000] nsexec-1[18564]: failed to update /proc/self/timens_offsets: Permission denied
FATA[0000] nsexec-0[18562]: failed to sync with stage-1: next state: Success
ERRO[0000] runc run failed: unable to start container process: can't get final child's PID from pipe: EOF
Fixes: ebc2e7c435 ("Support time namespace")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
For userns and timens, the mappings (and offsets, respectively) cannot
be changed after the namespace is first configured. Thus, configuring a
container with a namespace path to join means that you cannot also
provide configuration for said namespace. Previously we would silently
ignore the configuration (and just join the provided path), but we
really should be returning an error (especially when you consider that
the configuration userns mappings are used quite a bit in runc with the
assumption that they are the correct mapping for the userns -- but in
this case they are not).
In the case of userns, the mappings are also required if you _do not_
specify a path, while in the case of the time namespace you can have a
container with a timens but no mappings specified.
It should be noted that the case checking that the user has not
specified a userns path and a userns mapping needs to be handled in
specconv (as opposed to the configuration validator) because with this
patchset we now cache the mappings of path-based userns configurations
and thus the validator can't be sure whether the mapping is a cached
mapping or a user-specified one. So we do the validation in specconv,
and thus the test for this needs to be an integration test.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
If a user has misconfigured their userns mappings, they need to know
which id specifically is not mapped. There's no need to be vague.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While we do cache the mappings when using userns paths, there's no need
to do this in this particular case, since we are in the namespace and
set[ug]id() give unambiguous EINVAL error codes if the id is unmapped.
This appears to also be the only code which does Host[UG]ID calculations
from inside "runc init".
Ref: 1a5fdc1c5f ("init: support setting -u with rootless containers")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Our handling for name space paths with user namespaces has been broken
for a long time. In particular, the need to parse /proc/self/*id_map in
quite a few places meant that we would treat userns configurations that
had a namespace path as if they were a userns configuration without
mappings, resulting in errors.
The primary issue was down to the id translation helper functions, which
could only handle configurations that had explicit mappings. Obviously,
when joining a user namespace we need to map the ids but figuring out
the correct mapping is non-trivial in comparison.
In order to get the mapping, you need to read /proc/<pid>/*id_map of a
process inside the userns -- while most userns paths will be of the form
/proc/<pid>/ns/user (and we have a fast-path for this case), this is not
guaranteed and thus it is necessary to spawn a process inside the
container and read its /proc/<pid>/*id_map files in the general case.
As Go does not allow us spawn a subprocess into a target userns,
we have to use CGo to fork a sub-process which does the setns(2). To be
honest, this is a little dodgy in regards to POSIX signal-safety(7) but
since we do no allocations and we are executing in the forked context
from a Go program (not a C program), it should be okay. The other
alternative would be to do an expensive re-exec (a-la nsexec which would
make several other bits of runc more complicated), or to use nsenter(1)
which might not exist on the system and is less than ideal.
Because we need to logically remap users quite a few times in runc
(including in "runc init", where joining the namespace is not feasable),
we cache the mapping inside the libcontainer config struct. A future
patch will make sure that we stop allow invalid user configurations
where a mapping is specified as well as a userns path to join.
Finally, add an integration test to make sure we don't regress this again.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We had ignore lines for these warnings, but it turns out this is most
likely a bug in shellcheck and we can work around it by moving the
helper function definition before any of the functions that use the
helper function. Hopefully it'll be fixed soon.
Ref: https://github.com/koalaman/shellcheck/issues/2873
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Kir Kolyshkin (8):
libct: replace runType with hasInit
libct: Signal: slight refactor
runc kill: fix sending KILL to non-pidns container
runc delete -f: fix for no pidns + no init case
libct/cg: improve cgroup removal logic
runc delete: do not ignore error from destroy
runc delete, container.Destroy: kill all processes
libct: Destroy: don't proceed in case of errors
LGTMs: lifubang cyphar
For some reason, container destroy operation removes container's state
directory even if cgroup removal fails (and then still returns an
error). It has been that way since commit 5c246d038f, which added
cgroup removal.
This is problematic because once the container state dir is removed, we
no longer know container's cgroup and thus can't remove it.
Let's return the error early and fail if cgroup can't be removed.
Same for other operations: do not proceed if we fail.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
(For a container with no private PID namespace, that is).
When runc delete (or container.Destroy) is called on a stopped
container without private PID namespace and there are processes
in its cgroup, kill those.
Add a test case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If container.Destroy() has failed, runc destroy still return 0, which is
wrong and can result in other issues down the line.
Let's always return error from destroy in runc delete.
For runc checkpoint and runc run, we still treat it as a warning.
Co-authored-by: Zhang Tianyang <burning9699@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The current code is only doing retries in RemovePaths, which is only
used for cgroup v1 (cgroup v2 uses RemovePath, which makes no retries).
Let's remove all retry logic and logging from RemovePaths, together
with:
- os.Stat check from RemovePaths (its usage probably made sense before
commit 19be8e5ba5 but not after);
- error/warning logging from RemovePaths (this was added by commit
19be8e5ba5 in 2020 and so far we've seen no errors other
than EBUSY, so reporting the actual error proved to be useless).
Add the retry logic to rmdir, and the second retry bool argument.
Decrease the initial delay and increase the number of retries from the
old implementation so it can take up to ~1 sec before returning EBUSY
(was about 0.3 sec).
Hopefully, as a result, we'll have less "failed to remove cgroup paths"
errors.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit f8ad20f moved the kill logic from container destroy to container
kill (which is the right thing to do).
Alas, it broke the use case of doing "runc delete -f" for a container
which does not have its own private PID namespace, when its init process
is gone. In this case, some processes may still be running, and runc
delete -f should kill them (the same way as "runc kill" does).
It does not do that because the container status is "stopped" (as runc
considers the container with no init process as stopped), and so we only
call "destroy" (which was doing the killing before).
The fix is easy: if --force is set, call killContainer no matter what.
Add a test case, similar to the one in the previous commit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit f8ad20f made it impossible to kill leftover processes in a
stopped container that does not have its own PID namespace. In other
words, if a container init is gone, it is no longer possible to use
`runc kill` to kill the leftover processes.
Fix this by moving the check if container init exists to after the
special case of handling the container without own PID namespace.
While at it, fix the minor issue introduced by commit 9583b3d:
if signalAllProcesses is used, there is no need to thaw the
container (as freeze/thaw is either done in signalAllProcesses already,
or not needed at all).
Also, make signalAllProcesses return an error early if the container
cgroup does not exist (as it relies on it to do its job). This way, the
error message returned is more generic and easier to understand
("container not running" instead of "can't open file").
Finally, add a test case.
Fixes: f8ad20f
Fixes: 9583b3d
Co-authored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The semantics of runType is slightly complicated, and the only place
where we need to distinguish between Created and Running is
refreshState.
Replace runType with simpler hasInit, simplifying its users (except the
refreshState, which now figures out on its own whether the container is
Created or Running).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The container manager like containerd-shim can't use cgroup.kill feature or
freeze all the processes in cgroup to terminate the exec init process.
It's unsafe to call kill(2) since the pid can be recycled. It's good to
provide the pidfd of init process through the pidfd-socket. It's similar to
the console-socket. With the pidfd, the container manager like containerd-shim
can send the signal to target process safely.
And for the standard init process, we can have polling support to get
exit event instead of blocking on wait4.
Signed-off-by: Wei Fu <fuweid89@gmail.com>
It now comes with container-selinux 2:2.224.0-1.module_el8+712+4cd1bd69,
so we only need the kludge for CentOS 7 (which, I guess, is the sole
reason why we have this kludge at all).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Because runc-dmz is not checked into Git, go get will end up creating a
copy of libcontainer/dmz with no runc-dmz binary, which causes external
libcontainer users to have compilation errors.
Unfortunately, we cannot get go:embed to just ignore that there are no
files matching the provided pattern, so instead we need to create a
dummy file that matches the go:embed (which we check into Git and so go
get _will_ copy) and switch to embed.FS.
This is a little bit uglier, but at least it will fix external
libcontainer users.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Kir Kolyshkin (3):
Add dmz-vs-selinux kludge and a way to disable it
README: fix reference to memfd-bind
tests/int: add selinux test case
LGTMs: AkihiroSuda cyphar
Bind-mounts cannot have any filesystem-specific "data" arguments,
because the kernel ignores the data argument for MS_BIND and
MS_BIND|MS_REMOUNT and we cannot safely try to override the flags
because those would affect mounts on the host (these flags affect the
superblock).
It should be noted that there are cases where the filesystem-specified
flags will also be ignored for non-bind-mounts but those are kernel
quirks and there's no real way for us to work around them. And users
wouldn't get any real benefit from us adding guardrails to existing
kernel behaviour.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Add a workaround for a problem of older container-selinux not allowing
runc to use dmz feature. If runc sees that SELinux is in enforced mode
and the container's SELinux label is set, it disables dmz.
Add a build tag, runc_dmz_selinux_nocompat, which disables the workaround.
Newer distros that ship container-selinux >= 2.224.0 (currently CentOS
Stream 8 and 9, RHEL 8 and 9, and Fedora 38+) may build runc with this
build tag set to benefit from dmz working with SELinux.
Document the build tag in the top-level and libct/dmz READMEs.
Use the build tag in our CI builds for CentOS Stream 9 and Fedora 38,
as they already has container-selinux 2.224.0 available in updates.
Add a TODO to use the build tag for CentOS Stream 8 once it has
container-selinux updated.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a test case to demonstrate the selinux vs dmz issue.
The issue is, runc calls selinux.SetExecLabel and then execs the
runc-dmz binary, but the execve is denied by selinux:
> type=PROCTITLE msg=audit(10/05/2023 22:54:07.911:10904) : proctitle=/tmp/bats-run-sGk2sn/runc.Ql243q/bundle/runc init
> type=SYSCALL msg=audit(10/05/2023 22:54:07.911:10904) : arch=x86_64 syscall=execveat success=no exit=EACCES(Permission denied) a0=0x6 a1=0xc0000b90fa a2=0xc0000a26a0 a3=0xc000024660 items=0 ppid=105316 pid=105327 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=pts0 ses=8 comm=runc:[2:INIT] exe=/tmp/bats-run-sGk2sn/runc.Ql243q/bundle/runc subj=unconfined_u:unconfined_r:container_runtime_t:s0-s0:c0.c1023 key=(null)
> type=AVC msg=audit(10/05/2023 22:54:07.911:10904) : avc: denied { entrypoint } for pid=105327 comm=runc:[2:INIT] path=/memfd:runc_cloned:runc-dmz (deleted) dev="tmpfs" ino=2341 scontext=system_u:system_r:container_t:s0:c4,c5 tcontext=unconfined_u:object_r:container_runtime_tmpfs_t:s0 tclass=file permissive=0
Once that error is fixed (by adding a selinux rule that enables it), we
see one more error, also related to executing a file on tmpfs.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
CentOS 7 is showing its age and we'd rather skip some tests on it than
find out why they are flaky.
Add internal/testutil package, and move the generalized version of
SkipOnCentOS7 from libcontainer/cgroups/devices to there.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This prevents potential exploit of using "../" in cgroups.OpenFile
(as well as other methods that use OpenFile) to read or write to
other cgroups.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently, sometimes a short-lived "runc run" produces result with \r
and sometimes without. As a result, we have an occasional failure of
"runc run with tmpfs perms" test.
The solution (to the flaky test) is to use the first line of the output
(like many other tests do).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This field reports swap-only usage. For cgroupv1, `Usage` and `Failcnt`
are set by subtracting memory usage from memory+swap usage. For cgroupv2,
`Usage`, `Limit`, and `MaxUsage` are set. This commit also export `MaxUsage`
of memory under cgroupv2 mode, using `memory.peak` introduced in kernel 5.19.
Signed-off-by: Heran Yang <heran55@126.com>
The original reasoning for this option was to avoid having mount options
be overwritten by runc. However, adding command-line arguments has
historically been a bad idea because it forces strict-runc-compatible
OCI runtimes to copy out-of-spec features directly from runc and these
flags are usually quite difficult to enable by users when using runc
through several layers of engines and orchestrators.
A far more preferable solution is to have a heuristic which detects
whether copying the original mount's mount options would override an
explicit mount option specified by the user. In this case, we should
return an error. You only end up in this path in the userns case, if you
have a bind-mount source with locked flags.
During the course of writing this patch, I discovered that several
aspects of our handling of flags for bind-mounts left much to be
desired. We have completely botched the handling of explicitly cleared
flags since commit 97f5ee4e6a ("Only remount if requested flags differ
from current"), with our behaviour only becoming increasingly more weird
with 50105de1d8 ("Fix failure with rw bind mount of a ro fuse") and
da780e4d27 ("Fix bind mounts of filesystems with certain options
set"). In short, we would only clear flags explicitly request by the
user purely by chance, in ways that it really should've been reported to
us by now. The most egregious is that mounts explicitly marked "rw" were
actually mounted "ro" if the bind-mount source was "ro" and no other
special flags were included. In addition, our handling of atime was
completely broken -- mostly due to how subtle the semantics of atime are
on Linux.
Unfortunately, while the runtime-spec requires us to implement
mount(8)'s behaviour, several aspects of the util-linux mount(8)'s
behaviour are broken and thus copying them makes little sense. Since the
runtime-spec behaviour for this case (should mount options for a "bind"
mount use the "mount --bind -o ..." or "mount --bind -o remount,..."
semantics? Is the fallback code we have for userns actually
spec-compliant?) and the mount(8) behaviour (see [1]) are not
well-defined, this commit simply fixes the most obvious aspects of the
behaviour that are broken while keeping the current spirit of the
implementation.
NOTE: The handling of atime in the base case is left for a future PR to
deal with. This means that the atime of the source mount will be
silently left alone unless the fallback path needs to be taken, and any
flags not explicitly set will be cleared in the base case. Whether we
should always be operating as "mount --bind -o remount,..." (where we
default to the original mount source flags) is a topic for a separate PR
and (probably) associated runtime-spec PR.
So, to resolve this:
* We store which flags were explicitly requested to be cleared by the
user, so that we can detect whether the userns fallback path would end
up setting a flag the user explicitly wished to clear. If so, we
return an error because we couldn't fulfil the configuration settings.
* Revert 97f5ee4e6a ("Only remount if requested flags differ from
current"), as missing flags do not mean we can skip MS_REMOUNT (in
fact, missing flags are how you indicate a flag needs to be cleared
with mount(2)). The original purpose of the patch was to fix the
userns issue, but as mentioned above the correct mechanism is to do a
fallback mount that copies the lockable flags from statfs(2).
* Improve handling of atime in the fallback case by:
- Correctly handling the returned flags in statfs(2).
- Implement the MNT_LOCK_ATIME checks in our code to ensure we
produce errors rather than silently producing incorrect atime
mounts.
* Improve the tests so we correctly detect all of these contingencies,
including a general "bind-mount atime handling" test to ensure that
the behaviour described here is accurate.
This change also inlines the remount() function -- it was only ever used
for the bind-mount remount case, and its behaviour is very bind-mount
specific.
[1]: https://github.com/util-linux/util-linux/issues/2433
Reverts: 97f5ee4e6a ("Only remount if requested flags differ from current")
Fixes: 50105de1d8 ("Fix failure with rw bind mount of a ro fuse")
Fixes: da780e4d27 ("Fix bind mounts of filesystems with certain options set")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Function teardown assumes that every test case will call
setup_sshfs. Currently, this assumption is true, but once
a test case that won't call setup_sshfs is added (say because
it has some extra "requires" or "skip"), it will break bats,
so any bats invocation involving such a test case will end up
hanging after the last test case is run.
The reason is, we have set -u in helpers.bash to help catching the use
of undefined variables. In the above scenario, such a variable is DIR,
which is referenced in teardown but is only defined after a call to
setup_sshfs. As a result, bash that is running the teardown function
will exit upon seeing the first $DIR, and thus teardown_bundle won't be
run. This, in turn, results in a stale recvtty process, which inherits
bats' fd 3. Until that fd is closed, bats waits for test logs.
Long story short, the fix is to
- check if DIR is set before referencing it;
- unset it after unmount.
PS it is still not clear why there is no diagnostics about the failed
teardown.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
There is no point in showing the underlying error when path == "",
because it is ENOENT.
Revert the change done in commit e1584831b6.
Fixes: e1584831b6
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit e1584831b6 did two modifications to check_cgroup_value():
1. It now skips the test if the file is not found.
2. If the comparison failed, a second comparison, with value divided by 1000,
is performed.
These modifications were only needed for cpu.burst, but instead were done
in a generic function used from many cgroup tests. As a result, we can
no longer be sure about the test coverage (item 1) and the check being
correct (item 2) anymore. In fact, part of "update cgroup cpu limits"
test is currently skipped on CentOS 7 and 8 because of item 1.
To fix:
- replace item 1 with a new "cgroups_cpu_burst" argument for "requires",
and move the test to a separate case;
- replace item 2 with a local change in check_cpu_burst.
Fixes: e1584831b6
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit f34eb2c00 introduced a workaround to retry on EINTR due to changes in Go 1.14.
It was fixed in Go 1.15 [1], meaning a custom retry loop is no longer
necessary.
Keep the test case to avoid future regressions.
[1] https://github.com/golang/go/issues/38033
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since today, the URL from download.opensuse.org started returning a
HTTP 302 redirect, so -L option for curl is needed to follow it.
While at it, remove apt-key as per its man page recommendation:
> Note: Instead of using this command a keyring should be placed
> directly in the /etc/apt/trusted.gpg.d/ directory with a descriptive
> name and either "gpg" or "asc" as file extension.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When we added nolibc, we started using it unconditionally. But runc is
currently being compiled on more arches than supported by nolibc, like
MIPS.
Let's compile using stdlib if the arch we are compiling on is not
supported by nolibc.
If compilation is broken in some arch, just removing it from the
NOLIBC_GOARCHES variable should fix the compilation, as it will fallback
to use the C stdlib.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This adds support for hugetlb.<pagesize>.rsvd limiting and accounting.
The previous non-rsvd max/limit_in_bytes does not account for reserved
huge page memory, making it possible for a processes to reserve all the
huge page memory, without being able to allocate it (due to cgroup
restrictions).
In practice this makes it possible to successfully mmap more huge page
memory than allowed via the cgroup settings, but when using the memory
the process will get a SIGBUS and crash. This is bad for applications
trying to mmap at startup (and it succeeds), but the program crashes
when starting to use the memory. eg. postgres is doing this by default.
This also keeps writing to the old max/limit_in_bytes, for backward
compatibility.
More info can be found here: https://lkml.org/lkml/2020/2/3/1153
(commit message mostly written by Odin Ugedal)
Co-authored-by: Odin Ugedal <odin@ugedal.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since Go 1.16, [Go issue 1435][1] is solved, and the stdlib syscall
implementations work on Linux. While they are a bit more
flexible/heavier-weight than the implementations that were copied to
libcontainer/system (working across all threads), we compile with Cgo,
and using the libc wrappers should be just as suitable.
[1]: https://github.com/golang/go/issues/1435
Signed-off-by: Bjorn Neergaard <bjorn.neergaard@docker.com>
All the targets in the Makefile we have are phony (as we mostly rely on
go to figure out dependencies and whether to rebuild something), and
they have to be marked as such. We do that at the end of the file, and
the list is pretty long.
Instead, let's just add .PHONY before each target. That way it is easier
to spot any omissions.
Alternative solutions:
- add ".PHONY: %"; it won't work as wildcards are not recongized in
this context;
- add "MAKEFLAGS += --always-make".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Every `make` now produces something like this:
make[1]: Entering directory '/home/kir/go/src/github.com/opencontainers/runc'
readelf -h runc
Machine: Advanced Micro Devices X86-64
Flags: 0x0
readelf -h libcontainer/dmz/runc-dmz
Machine: Advanced Micro Devices X86-64
Flags: 0x0
runc-dmz architecture matches runc binary.
make[1]: Leaving directory '/home/kir/go/src/github.com/opencontainers/runc'
That is a bit too much. Let's make it less verbose.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Bump fileutils to v0.5.1, which fixes permissions of newly created directories
to not depend on the value of umask.
Add a test case which fails like this before the fix:
mounts.bats
✗ runc run [tmpcopyup]
(in test file tests/integration/mounts.bats, line 28)
`[[ "${lines[0]}" == *'drwxrwxrwx'* ]]' failed
runc spec (status=0):
runc run test_busybox (status=0):
drwxr-xr-x 2 root root 40 Oct 4 22:35 /dir1/dir2
Fixes 3991.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The name "root" (or "containerRoot") is confusing; one might think it is
the root of container's file system (the directory we chroot into).
Rename to stateDir for clarity.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Umask is problematic for Go programs as it affects other goroutines
(see [1] for more details).
Instead of using it, let's just prop up with Chmod.
Note this patch misses the MkdirAll call in createDeviceNode. Since the
runtime spec does not say anything about creating intermediary
directories for device nodes, let's assume that doing it via mkdir with
the current umask set is sufficient (if not, we have to reimplement
MkdirAll from scratch, with added call to os.Chmod).
[1] https://github.com/opencontainers/runc/pull/3563#discussion_r990293788
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
"armhf" means ARMv7 for Debian, ARMv6 for Raspbian.
ARMv6 is chosen here for compatibility.
https://wiki.debian.org/RaspberryPi
> Raspberry Pi OS builds a single image for all of the Raspberry families,
> so you will get an armhf 32-bit, hard floating-point system, but built
> for the ARMv6 ISA (with VFP2), unlike Debian's ARMv7 ISA (with VFP3)
> port.
Prior to this commit, the script was setting GOARM=6 for armel,
GOARM=7 for armhf.
Fix issue 4033
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
In reviewing PR 4024 ("libct/dmz: Reduce the binary size using nolibc"),
we noticed that we do not intend to actively support MIPS.
We do not intend to support i386 either.
This might be a breaking change for Debian, which has been officially
providing runc packages for `i386`, `mips64el` and `mipsel`:
https://packages.debian.org/bookworm/runc
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Otherwise it is shown when compiling, like this:
# We use the flags suggested in nolibc/nolibc.h, it makes the binary very small.
gcc -fno-asynchronous-unwind-tables -fno-ident -s -Os -nostdlib -lgcc -static -o runc-dmz _dmz.c
strip -gs runc-dmz
Having it before the target is equally clear and will not be shown while
compiling.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Linux repo has under `tools/include/nolibc` very simple include files
that we can use to generate very small binaries that don't depend on
libc.
To make things even better, since Linux 6.6 it supports all the
architectures we support in runc, which is just beautiful.
The runc-dmz binary on x86_64 before this patch (on my debian host) was
taking 636K, with this patch it takes only 8K.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We have different requirements for the initial configuration and
initWaiter pipe (just send netlink and JSON blobs with no complicated
handling needed for message coalescing) and the packet-based
synchronisation pipe.
Tests with switching everything to SOCK_SEQPACKET lead to endless issues
with runc hanging on start-up because random things would try to do
short reads (which SOCK_SEQPACKET will not allow and the Go stdlib
explicitly treats as a streaming source), so splitting it was the only
reasonable solution. Even doing somewhat dodgy tricks such as adding a
Read() wrapper which actually calls ReadPacket() and makes it seem like
a stream source doesn't work -- and is a bit too magical.
One upside is that doing it this way makes the difference between the
modes clearer -- INITPIPE is still used for initWaiter syncrhonisation
but aside from that all other synchronisation is done by SYNCPIPE.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The old name was quite confusing, and with the addition of the
procMountPlease sync message there are now multiple sync messages that
are related to "resuming" runc-init.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Due to the way capabilities have to be set by runc, capabilities need to
be included in the inheritable and ambient sets anyway. Otherwise, the
container process would not have the correct privileges. This test only
functioned because adding CAP_DAC_OVERRIDE to the inherited,
permissible, and bounding sets means that only "runc init" has these
capabilities -- everything other than the bounding set is cleared on the
first execve(). This breaks with runc-dmz, but the behaviour was broken
from the outset.
Docker appears to not handle this properly at all (the logic for
capability sets changed with the introduction of ambient capabilities,
and while Docker was updated it seems the behaviour is still incorrect
for non-root users).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This really isn't ideal but it can be used to avoid the largest issues
with the memfd-based runc binary protection. There are several caveats
with using this tool, see the help page for the new binary for details.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The idea is to remove the need for cloning the entire runc binary by
replacing the final execve() call of the container process with an
execve() call to a clone of a small C binary which just does an execve()
of its arguments.
This provides similar protection against CVE-2019-5736 but without
requiring a >10MB binary copy for each "runc init". When compiled with
musl, runc-dmz is 13kB (though unfortunately with glibc, it is 1.1MB
which is still quite large).
It should be noted that there is still a window where the container
processes could get access to the host runc binary, but because we set
ourselves as non-dumpable the container would need CAP_SYS_PTRACE (which
is not enabled by default in Docker) in order to get around the
proc_fd_access_allowed() checks. In addition, since Linux 4.10[1] the
kernel blocks access entirely for user namespaced containers in this
scenario. For those cases we cannot use runc-dmz, but most containers
won't have this issue.
This new runc-dmz binary can be opted out of at compile time by setting
the "runc_nodmz" buildtag, and at runtime by setting the RUNC_DMZ=legacy
environment variable. In both cases, runc will fall back to the classic
/proc/self/exe-based cloning trick. If /proc/self/exe is already a
sealed memfd (namely if the user is using contrib/cmd/memfd-bind to
create a persistent sealed memfd for runc), neither runc-dmz nor
/proc/self/exe cloning will be used because they are not necessary.
[1]: https://github.com/torvalds/linux/commit/bfedb589252c01fa505ac9f6f2a3d5d68d707ef4
Co-authored-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: lifubang <lifubang@acmcoder.com>
[cyphar: address various review nits]
[cyphar: fix runc-dmz cross-compilation]
[cyphar: embed runc-dmz into runc binary and clone in Go code]
[cyphar: make runc-dmz optional, with fallback to /proc/self/exe cloning]
[cyphar: do not use runc-dmz when the container has certain privs]
Co-authored-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Previously, if /var/run was mounted noexec, our cloned binary logic
would not work if memfd_create(2) was not available because we would try
to exec a binary that is on a noexec filesystem.
We cannot guarantee there will be an executable filesystem on the system
(other than mounting one ourselves, which would cause a bunch of other
headaches) but we can at least try the obvious options (/tmp, /bin, and
/). If none of these work, we will have to fail.
Reported-by: lifubang <lifubang@acmcoder.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This allow us to remove the amount of C code in runc quite
substantially, as well as removing a whole execve(2) from the nsexec
path because we no longer spawn "runc init" only to re-exec "runc init"
after doing the clone.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We need these to match the Makefile detection of the right gcc for
runc-dmz, as well as making sure that everything builds properly for our
cross-i386 tests. While we're at it, add x86 to the list of build
targets for release builds (presumably nobody will use it, but since we
do test builds of this anyway it probably won't hurt).
In addition, clean up the handling of the native architecture build by
treating it the same as any other build (ensuring that building runc
from a different platform will work the same way regardless of the
native architecture). In practice, the build works the same way as
before.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The runtime spec now allows relative mount dst paths, so remove the
comment saying we will switch this to an error later and change the
error messages to reflect that.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
From the Go specification:
"1. For a nil slice, the number of iterations is 0." [1]
Therefore, an additional nil check for before the loop is unnecessary.
[1]: https://go.dev/ref/spec#For_range
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
Burstable CFS controller is introduced in Linux 5.14. This helps with
parallel workloads that might be bursty. They can get throttled even
when their average utilization is under quota. And they may be latency
sensitive at the same time so that throttling them is undesired.
This feature borrows time now against the future underrun, at the cost
of increased interference against the other system users, by introducing
cfs_burst_us into CFS bandwidth control to enact the cap on unused
bandwidth accumulation, which will then used additionally for burst.
The patch adds the support/control for CFS bandwidth burst.
runtime-spec: https://github.com/opencontainers/runtime-spec/pull/1120
Co-authored-by: Akihiro Suda <suda.kyoto@gmail.com>
Co-authored-by: Nadeshiko Manju <me@manjusaka.me>
Signed-off-by: Kailun Qin <kailun.qin@intel.com>
This makes libcontainer/userns self-dependent, largely returning to
the original implementation from lxc. The `uiMapInUserNS` is kept as
a separate function for unit-testing and fuzzing.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
If the container does not have own mount namespace configured (i.e. it
shares the mount namespace with the host), its "prestart" (obsoleted)
and "createRuntime" hooks are called twice, and its cgroups and Intel
RDT settings are also applied twice.
The code being removed was originally added by commit 2f2764984 ("Move
pre-start hooks after container mounts", Feb 17 2016). At that time,
the syncParentHooks() was called from setupRootfs(), which was only
used when the container config has mount namespace (NEWNS) enabled.
Later, commit 244c9fc426 ("*: console rewrite", Jun 4 2016) spli
the relevant part of setupRootfs() into prepareRootfs(). It was still
called conditionally (only if mount namespace was enabled).
Finally, commit 91ca331474 ("chroot when no mount namespaces is
provided", Jan 25 2018) removed the above condition, meaning
prepareRootfs(), and thus syncParentHooks(), is now called for any
container.
Meaning, the special case for when mount namespace is not enabled is no
longer needed.
Remove it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This file contains a test that tests .so hooks. It has a complicated
setup and teardown, and has special requirements (root and no_systemd).
Rename it to hooks_so.bats, so we can add hooks.bats for hooks tests
that do not have such complicated setup and requirements.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When a hook has failed, the error message looks like this:
> error running hook: error running hook #1: exit status 1, stdout: ...
The two problems here are:
1. it is impossible to know what kind of hook it was;
2. "error running hook" stuttering;
Change that to
> error running createContainer hook #1: exit status 1, stdout: ...
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
golangci-lint v1.54.2 comes with errorlint v1.4.4, which contains
the fix [1] whitelisting all errno comparisons for errors coming from
x/sys/unix.
Thus, these annotations are no longer necessary. Hooray!
[1] https://github.com/polyfloyd/go-errorlint/pull/47
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Usually errorlint allows io.EOF comparison (based on a whitelist of
functions that can return bare io.EOF), thus there is no need for nolint
annotation.
In this very case, though, the need for nolint is caused by issue with
errorlint, which fails to see where err is coming from.
Refer to the issue so when it is fixed we can remove the annotation.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The default timeout is 360 minutes, which is way long for these jobs.
If the CI (or a test) has stuck, we'd better know about it earlier than
in 6 hours.
Set the timeouts for some [relatively] long running jobs conservatively:
- test and release jobs usually take ~10 minutes;
- lint job takes 1 minute (but can be a few times slower when we switch
Go or golangci-lint version);
- cross-386 job takes about 2 minutes;
- the rest is seconds (and I am lazy to set timeouts everywhere).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In the runc state JSON we always use snake_case. This is a no-op change,
but it will cause any existing container state files to be incorrectly
parsed. Luckily, commit fbf183c6f8 ("Add uid and gid mappings to
mounts") has never been in a runc release so we can change this before a
1.2.z release.
Fixes: fbf183c6f8 ("Add uid and gid mappings to mounts")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The original implementation of cgroupns had additional synchronisation
to "ensure" that the process is in the correct cgroup before unsharing
the cgroupns. This behaviour was actually never necessary, and after
commit 5110bd2fc0 ("nsenter: remove cgroupns sync mechanism") there is
no synchronisation at all, meaning that CLONE_NEWCGROUP should not get
any special treatment.
Fixes: 5110bd2fc0 ("nsenter: remove cgroupns sync mechanism")
Fixes: df3fa115f9 ("Add support for cgroup namespace")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The kernel ignores these arguments, and passing them can lead to
confusing error messages (the old source is irrelevant for MS_REMOUNT),
as well as causing issues for a future patch where we switch to
move_mount(2).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
*os.File is correctly tracked by the garbage collector, and there's no
need to use raw file descriptors for this code.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This includes quite a few cleanups and improvements to the way we do
synchronisation. The core behaviour is unchanged, but switching to
embedding json.RawMessage into the synchronisation structure will allow
us to do more complicated synchronisation operations in future patches.
The file descriptor passing through the synchronisation system feature
will be used as part of the idmapped-mount and bind-mount-source
features when switching that code to use the new mount API outside of
nsexec.c.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The code in this function became quite complicated and not entirely
correct over time. As a result, if an error is returned from parseSync,
it might end up stuck waiting for the child to finish.
1. Let's not wait() for the child twice. We already do it in the
defer statement (call p.terminate()) when we are returning an error.
2. Remove sentResume and sentRun since we do not want to check if
these were sent or not. Instead, introduce and check seenProcReady, as
procReady is always expected from runc init.
3. Eliminate the possibility to wrap nil as an error.
4. Make sure we always call shutdown on the sync socket, and do not let
shutdown error shadow the ierr.
This fixes the issue of stuck `runc runc` with the optimization patch
(sending procSeccompDone earlier) applied.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These are not exhaustive, but at least confirm that the feature is not
obviously broken (we correctly set the time offsets).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Fix up a few things that were flagged in the review of the original
timens PR, namely around error handling and validation.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The idea was to make them strict on dest path from the beginning for
idmap mounts, as runc would do that for all mounts in the future. But
that is causing too many problems.
For now, let's just allow relative paths for idmap mounts too. It just
seems safer.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This reverts commit 881e92a3fd and adjust
the code so the idmap validations are strict.
We now only throw a warning and the container is started just fine.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This commit does two things:
1. Consolidate StartInitialization calling logic into Init().
2. Fix init error handling logic.
The main issues at hand are:
- the "unable to convert _LIBCONTAINER_INITPIPE" error from
StartInitialization is never shown;
- errors from WriteSync and WriteJSON are never shown;
- the StartInit calling code is triplicated;
- using panic is questionable.
Generally, our goals are:
- if there's any error, do our best to show it;
- but only show each error once;
- simplify the code, unify init implementations.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently, logrus is used from the Go part of runc init, mostly for a
few debug messages (see setns_init_linux.go and standard_init_linux.go),
and a single warning (see rootfs_linux.go).
This means logrus is part of init implementation, and thus, its setup
belongs to StartInitialization().
Move the code there. As a nice side effect, now we don't have to convert
_LIBCONTAINER_LOGPIPE twice.
Note that since this initialization is now also called from libct/int
tests, which do not set _LIBCONTAINER_LOGLEVEL, let's make
_LIBCONTAINER_LOGLEVEL optional.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
With the new vm.memfd_noexec sysctl, we need to make sure we explicitly
request MFD_EXEC, otherwise an admin could inadvertently break
containers in a somewhat-annoying-to-debug fashion.
It should be noted that vm.memfd_noexec=2 is broken on Linux 6.4
(MFD_EXEC works even in the most restrictive mode) and the most severe
breakage is going to be fixed in Linux 6.6[1].
[1]: https://lore.kernel.org/20230705063315.3680666-2-jeffxu@google.com/
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While the ro-bind-mount trick did eliminate the memory overhead of
copying the runc binary for each "runc init" invocation, on machines
with very significant container churn, creating a temporary mount
namespace on every container invocation can trigger severe lock
contention on namespace_sem that makes containers fail to spawn.
The only reason we added bindfd in commit 16612d74de ("nsenter:
cloned_binary: try to ro-bind /proc/self/exe before copying") was due to
a Kubernetes e2e test failure where they had a ridiculously small memory
limit. It seems incredibly unlikely that real workloads are running
without 10MB to spare for the very short time that runc is interacting
with the container.
In addition, since the original cloned_binary implementation, cgroupv2
is now almost universally used on modern systems. Unlike cgroupv1, the
cgroupv2 memcg implementation does not migrate memory usage when
processes change cgroups (even cgroupv1 only did this if you had
memory.move_charge_at_immigrate enabled). In addition, because we do the
/proc/self/exe clone before synchronising the bootstrap data read, we
are guaranteed to do the clone before "runc init" is moved into the
container cgroup -- meaning that the memory used by the /proc/self/exe
clone is charged against the root cgroup, and thus container workloads
should not be affected at all with memfd cloning.
The long-term fix for this problem is to block the /proc/self/exe
re-opening attack entirely in-kernel, which is something I'm working
on[1]. Though it should also be noted that because the memfd is
completely separate to the host binary, even attacks like Dirty COW
against the runc binary can be defended against with the memfd approach.
Of course, once we have in-kernel protection against the /proc/self/exe
re-opening attack, we won't have that protection anymore...
[1]: https://lwn.net/Articles/934460/
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
As we now log the log file name in logCriuErrors.
While at it, there is no need to use var.String() with %s as it is done
by the runtime.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When criu fails, it does not give us much context to understand what
was the cause of an error -- for that, we need to take a look into its
log file.
This is somewhat complicated to do (as you can see in parts of
checkpoint.bats removed by this commit), and not very user-friendly.
Add a function to find and log errors from criu logs, together with some
preceding context, in case either checkpoint or restore has failed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test fails in CI sometimes with the following error:
> `testcontainer test_update stopped' failed
Give OOM killer some time to do its job.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A few cases relied on the fact that systemd is used, and thus
/sys/fs/cgroup/user.slice is available.
Guess what, in case of "make unittest" it might not be.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Split the test into two -- for fs and systemd cgroup managers, and only
run the second one if systemd is available.
Prevents the following failure during `make unittest`:
> === RUN TestNilResources
> manager_test.go:27: systemd not running on this host, cannot use systemd cgroups manager
> --- FAIL: TestNilResources (0.22s)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For "make integration", the tests are run inside a Docker/Podman
container. Problem is, if cgroup v2 is used, the in-container
/sys/fs/cgroup/cgroup.subtree_control is empty.
The added script, used as Docker entrypoint, moves the current process
into a sub-cgroup, and then adds all controllers in top-level
cgroup.subtree_control.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
"time" namespace was introduced in Linux v5.6
support new time namespace to set boottime and monotonic time offset
Example runtime spec
"timeOffsets": {
"monotonic": {
"secs": 172800,
"nanosecs": 0
},
"boottime": {
"secs": 604800,
"nanosecs": 0
}
}
Signed-off-by: Chethan Suresh <chethan.suresh@sony.com>
We can't call log.Fatalf() and defer functions, as the former doesn't
call any defers. Let's just move the code to a new function and call
os.Exit() only in main, when all defer executed.
Now that all the code is one function, we only print twice to stderr. It
is simpler to just print to stderr instead of logging and having also
the timestamp we don't really want.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Let's just rely on the lookup performed to find the sleep binary.
This didn't cause any issues as far as I know, I just saw this while
doing other cleanups.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We don't really need to check if AT_RECURSIVE is possible here. We just
want to check if we can idmap the src, it doesn't matter other nested
mounts.
While we are there, allow relative paths too.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This aligns v2 usage calculations more closely with v1.
Current node-level reporting for v1 vs v2 on the same
machine under similar load may differ by ~250-750Mi.
Also return usage as combined swap + memory usage, aligned
with v1 and non-root v2 cgroups.
`mem_cgroup_usage` in the kernel counts NR_FILE_PAGES
+ NR_ANON_MAPPED + `nr_swap_pages` (if swap enabled) [^0].
Using total - free results in higher "usage" numbers.
This is likely due to various types of reclaimable
memory technically counted as in use (e.g. inactive anon).
See also https://github.com/kubernetes/kubernetes/issues/118916 for more context
[^0]: https://github.com/torvalds/linux/blob/06c2afb862f9da8dc5efa4b6076a0e48c3fbaaa5/mm/memcontrol.c#L3673-L3680
Signed-off-by: Alexander Eldeib <alexeldeib@gmail.com>
This is what we should do, although in practice this probably won't be a
big issue as the parent also exits.
While we are there, instead of waiting for the child to finish, kill it
if we did everything we wanted to do.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
close_range(2) is far more efficient than a readdir of /proc/self/fd and
then doing a syscall for each file descriptor.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
While testing this with old kernel versions and kernels that don't
support idmap mounts for some of the filesystems used by a container, I
realized we can throw a more clear errors.
Let's make it clear which syscall we are using, when it is not supported
and when if the fs doesn't support idmap mounts, which path it is.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Since systemd v240 (commit 8e8b5d2e6d91180a), one can use
/dev/{char,block}-MAJOR syntax to specify that all MAJOR:*
devices are allowed.
Use it, if available, since it's more straightforward, plus
we can skip somewhat expensive parsing of /proc/devices.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently bind mounts of filesystems with nodev, nosuid, noexec,
noatime, relatime, strictatime, nodiratime options set fail in rootless
mode if the same options are not set for the bind mount.
For ro filesystems this was resolved by #2570 by remounting again
with ro set.
Follow the same approach for nodev, nosuid, noexec, noatime, relatime,
strictatime, nodiratime but allow to revert back to the old behaviour
via the new `--no-mount-fallback` command line option.
Add a testcase to verify that bind mounts of filesystems with nodev,
nosuid, noexec, noatime options set work in rootless mode.
Add a testcase that mounts a nodev, nosuid, noexec, noatime filesystem
with a ro flag.
Add two further testcases that ensure that the above testcases would
fail if the `--no-mount-fallback` command line option is set.
* contrib/completions/bash/runc:
Add `--no-mount-fallback` command line option for bash completion.
* create.go:
Add `--no-mount-fallback` command line option.
* restore.go:
Add `--no-mount-fallback` command line option.
* run.go:
Add `--no-mount-fallback` command line option.
* libcontainer/configs/config.go:
Add `NoMountFallback` field to the `Config` struct to store
the command line option value.
* libcontainer/specconv/spec_linux.go:
Add `NoMountFallback` field to the `CreateOpts` struct to store
the command line option value and store it in the libcontainer
config.
* utils_linux.go:
Store the command line option value in the `CreateOpts` struct.
* libcontainer/rootfs_linux.go:
In case that `--no-mount-fallback` is not set try to remount the
bind filesystem again with the options nodev, nosuid, noexec,
noatime, relatime, strictatime or nodiratime if they are set on
the source filesystem.
* tests/integration/mounts_sshfs.bats:
Add testcases and rework sshfs setup to allow specifying
different mount options depending on the test case.
Signed-off-by: Ruediger Pluem <ruediger.pluem@vodafone.com>
We cannot have both srcFD and idMapFD set at the same time.
So, we can simplify this struct to only have one field which is used a srcFD
most of the time and as idMapFD when we do an id map mount.
Signed-off-by: Francis Laniel <flaniel@linux.microsoft.com>
Previously to this commit, we used a string for srcFD as /proc/self/fd/NN.
This commit modified to this behavior, so srcFD is only an *int and the full path
is constructed in mountViaFDs() if srcFD is different than nil.
Signed-off-by: Francis Laniel <flaniel@linux.microsoft.com>
runc ps requires cgroup, but all the tests but one required root. Let's
fix this.
1. Add rootless cgroup requirement to setup() to avoid repetition.
2. Add set_cgroups_path to setup() for rootless containers because
there is no default cgroup path.
3. Modify output checks to use $output rather than $lines because in case
of rootless the first line of output contains the following warning:
> runc ps may fail if you don't have the full access to cgroups
4. While at it, move the common part of every test (creating the
container and making sure it's running) to setup().
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This commit adds support for idmap mounts as specified in the runtime-spec.
We open the idmap source paths and call mount_setattr() in runc PARENT,
as we need privileges in the init userns for that, and then sends the
fds to the child process. For this fd passing we use the same mechanism
used in other parts of thecode, the _LIBCONTAINER_ env vars.
The mount is finished (unix.MoveMount) from go code, inside the userns,
so we reuse all the prepareBindMount() security checks and the remount
logic for some flags too.
This commit only supports idmap mounts when userns are used AND the mappings
are the same specified for the userns mapping. This limitation is to
simplify the initial implementation, as all our users so far only need
this, and we can avoid sending over netlink the mappings, creating a
userns with this custom mapping, etc. Future PRs will remove this
limitation.
Co-authored-by: Francis Laniel <flaniel@linux.microsoft.com>
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We will need to pass more slices of fds to these functions in future
patches. Let's add a struct that just contains them all, instead of
adding lot of parameters to these functions.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We will add code that uses this function in future patches. So let's
just split it to a new function now and reuse it later.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Let's move the code to send mount sources to a generic function. Future
patches will use it for idmap sources too.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We add idmap.h with the needed includes and defines in case the system
headers don't have the definition for the idmap syscalls we need.
Future patches will use these helpers.
Co-authored-by: Francis Laniel <flaniel@linux.microsoft.com>
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
We just added this helper for other parts of the code, let's switch this
function to use the helper too.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This was a warning already and it was requested to make this an error
while we will add validation of idmap mounts:
https://github.com/opencontainers/runc/pull/3717#discussion_r1154705318
I've also tested a k8s cluster and the config.json generated by
containerd didn't use any relative paths. I tested one pod, so it was
definitely not an extensive test.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
Processes can watch /proc/self/mounts or /mountinfo, and the kernel
will notify them whenever the namespace's mount table is modified. The
notified process still needs to read and parse the mountinfo to
determine what changed once notified. Many such processes, including
udisksd and SystemD < v248, make no attempt to rate-limit their
mountinfo notifications. This tends to not be a problem on many systems,
where mount tables are small and mounting and unmounting is uncommon.
Every runC exec which successfully uses the try_bindfd container-escape
mitigation performs two mount()s and one umount() in the host's mount
namespace, causing any mount-watching processes to wake up and parse the
mountinfo file three times in a row. Consequently, using 'exec' health
checks on containers has a larger-than-expected impact on system load
when such mount-watching daemons are running. Furthermore, the size of
the mount table in the host's mount namespace tends to be proportional
to the number of OCI containers as a unique mount is required for the
rootfs of each container. Therefore, on systems with mount-watching
processes, the system load increases *quadratically* with the number of
running containers which use health checks!
Prevent runC from incidentally modifying the host's mount namespace for
container-escape mitigations by setting up the mitigation in a temporary
mount namespace.
Signed-off-by: Cory Snider <csnider@mirantis.com>
Modify receive_fd() and send_fd() so they can be more readily reused in
cloned_binary.c. Change receive_fd() to have a single responsibility:
receiving and returning a single file descriptor over a UNIX domain
socket. Make send_fd() useable in precarious execution contexts such as
a clone(CLONE_VFORK|CLONE_VM) "thread" where allocating heap memory or
calling exit() would be dangerous.
Signed-off-by: Cory Snider <csnider@mirantis.com>
I am nominating @lifubang for the role of runc maintainer.
He provided a number of valuable contributions to runc, and demonstrated
both deep technical expertise and the long term commitment to the
project.
As noted in MAINTAINERS_GUIDE.md, we have a week to vote, and need to
get 66% of current maintainers' votes.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The passing run (with the fix) looks like this:
----
delete.bats
✓ runc delete removes failed systemd unit [4556]
runc spec (status=0):
runc run -d --console-socket /tmp/bats-run-B08vu1/runc.lbQwU5/tty/sock test-failed-unit (status=0):
Warning: The unit file, source configuration file or drop-ins of runc-cgroups-integration-test-12869.scope changed on disk. Run 'systemctl daemon-reload' to reload units.
× runc-cgroups-integration-test-12869.scope - libcontainer container integration-test-12869
Loaded: loaded (/run/systemd/transient/runc-cgroups-integration-test-12869.scope; transient)
Transient: yes
Drop-In: /run/systemd/transient/runc-cgroups-integration-test-12869.scope.d
└─50-DevicePolicy.conf, 50-DeviceAllow.conf
Active: failed (Result: timeout) since Tue 2023-06-13 14:41:38 PDT; 751ms ago
Duration: 2.144s
CPU: 8ms
Jun 13 14:41:34 kir-rhat systemd[1]: Started runc-cgroups-integration-test-12869.scope - libcontainer container integration-test-12869.
Jun 13 14:41:37 kir-rhat systemd[1]: runc-cgroups-integration-test-12869.scope: Scope reached runtime time limit. Stopping.
Jun 13 14:41:38 kir-rhat systemd[1]: runc-cgroups-integration-test-12869.scope: Stopping timed out. Killing.
Jun 13 14:41:38 kir-rhat systemd[1]: runc-cgroups-integration-test-12869.scope: Killing process 1107438 (sleep) with signal SIGKILL.
Jun 13 14:41:38 kir-rhat systemd[1]: runc-cgroups-integration-test-12869.scope: Failed with result 'timeout'.
runc delete test-failed-unit (status=0):
Unit runc-cgroups-integration-test-12869.scope could not be found.
----
Before the fix, the test was failing like this:
----
delete.bats
✗ runc delete removes failed systemd unit
(in test file tests/integration/delete.bats, line 194)
`run -4 systemctl status "$SD_UNIT_NAME"' failed, expected exit code 4, got 3
....
----
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
runc delete is supposed to remove all the container's artefacts.
In case systemd cgroup driver is used, and the systemd unit has failed
(e.g. oom-killed), systemd won't remove the unit (that is, unless the
"CollectMode: inactive-or-failed" property is set).
Call reset-failed from manager.Destroy so the failed unit will be
removed during "runc delete".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Sometimes we call resetFailedUnit as a cleanup measure, and we don't
care if it fails or not. So, move error reporting to its callers, and
ignore error in cases we don't really expect it to succeed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no such thing as linux.resources.memorySwap (the mem+swap is
set as linux.resources.memory.swap).
As it is not used in this test anyway, remove it.
Fixes: 4929c05ad1
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Replace a panic with a warning, unless it's ENOENT and we're running in
a user namespace. In the latter case, do the same as before, i.e. report
the error but using a Debug logging level.
This prevents software that uses libcontainer from panicking in
some exotic setups.
This will also print a warning on some very old systems which does not
use /sys/fs/cgroup for cgroup mount point. My bet is such systems no
longer exist.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When a directory already exists (or after a container is restarted) the
perms of the directory being mounted to were being used even when a
different permission is set on the tmpfs mount options.
This prepends the original directory perms to the mount options.
If the perms were already set in the mount opts then those perms will
win.
This eliminates the need to perform a chmod after mount entirely.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Remove some old exceptions (no longer needed), add a new one
(codespell 2.2.5 flags "(mis)features" in docs/terminal.md).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We used to skip testing rootless integration tests for systemd, because
in case of cgroup v1 it does not support user delegation.
Since we added ubuntu 22.04 to the testing matrix, we can actually test
rootless+systemd there (with the proper systemd setup).
Fixes: 953e1cc485
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We read output from the following files if they exists:
- cpu.pressure
- memory.pressure
- io.pressure
Each are in format:
```
some avg10=0.00 avg60=0.00 avg300=0.00 total=0
full avg10=0.00 avg60=0.00 avg300=0.00 total=0
```
Signed-off-by: Daniel Dao <dqminh89@gmail.com>
Co-authored-by: Sandor Szücs <sandor.szuecs@zalando.de>
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently, developer.gnome.org/documentation no longer hosts the
documentation we used to refer to. Link to docs.gtk.org instead.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In code we have frozen the cgroup to avoid the processes get
an occasional "permission denied" error, while the systemd's application of device
rules is done disruptively. When the processes in the container can not
be frozen over 2 seconds (which defined in fs/freezer.go),
we still update the cgroup which resulting the container get an occasional
"permission denied" error in some cases.
Return error directly without updating cgroup, when freeze fails.
Fixes: #3803
Signed-off-by: Zoe <hi@zoe.im>
Kir Kolyshkin (7):
libct: implement support for cgroup.kill
runc kill: drop -a option
libct: move killing logic to container.Signal
libct: fix shared pidns detection
libct: signalAllProcesses: remove child reaping
tests/int/kill: add kill -a with host pidns test
tests/rootless.sh: drop set -x
LGTMs: AkihiroSuda cyphar
Closes#3825
We were not running localrootlessintegration test on CentOS Stream 9
because of some failures fixed by previous commits.
Enable rootless integration with both systemd and fs drivers.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Rename CGROUP_PATH to CGROUP_V2_PATH so it is more clear that it can
only be used for CGROUP_V2, and to resolve ambiguity with CGROUP_PATH
variable used in tests/rootless.sh.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Filter out rdma controller since systemd is unable to delegate it.
Similar to commits 05272718f4 and 601cf5825f.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As of previous commit, this is implied in a particular scenario. In
fact, this is the one and only scenario that justifies the use of -a.
Drop the option from the documentation. For backward compatibility, do
recognize it, and retain the feature of ignoring the "container is
stopped" error when set.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
By default, the container has its own PID namespace, and killing (with
SIGKILL) its init process from the parent PID namespace also kills all
the other processes.
Obviously, it does not work that way when the container is sharing its
PID namespace with the host or another container, since init is no
longer special (it's not PID 1). In this case, killing container's init
will result in a bunch of other processes left running (and thus the
inability to remove the cgroup).
The solution to the above problem is killing all the container
processes, not just init.
The problem with the current implementation is, the killing logic is
implemented in libcontainer's initProcess.wait, and thus only available
to libcontainer users, but not the runc kill command (which uses
nonChildProcess.kill and does not use wait at all). So, some workarounds
exist:
- func destroy(c *Container) calls signalAllProcesses;
- runc kill implements -a flag.
This code became very tangled over time. Let's simplify things by moving
the killing all processes from initProcess.wait to container.Signal,
and documents the new behavior.
In essence, this also makes `runc kill` to automatically kill all container
processes when the container does not have its own PID namespace.
Document that as well.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When someone is using libcontainer to start and kill containers from a
long lived process (i.e. the same process creates and removes the
container), initProcess.wait method is used, which has a kludge to work
around killing containers that do not have their own PID namespace.
The code that checks for own PID namespace is not entirely correct.
To be exact, it does not set sharePidns flag when the host/caller PID
namespace is implicitly used. As a result, the above mentioned kludge
does not work.
Fix the issue, add a test case (which fails without the fix).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There are two very distinct usage scenarios for signalAllProcesses:
* when used from the runc binary ("runc kill" command), the processes
that it kills are not the children of "runc kill", and so calling
wait(2) on each process is totally useless, as it will return ECHLD;
* when used from a program that have created the container (such as
libcontainer/integration test suite), that program can and should call
wait(2), not the signalling code.
So, the child reaping code is totally useless in the first case, and
should be implemented by the program using libcontainer in the second
case. I was not able to track down how this code was added, my best
guess is it happened when this code was part of dockerd, which did not
have a proper child reaper implemented at that time.
Remove it, and add a proper documentation piece.
Change the integration test accordingly.
PS the first attempt to disable the child reaping code in
signalAllProcesses was made in commit bb912eb00c, which used a
questionable heuristic to figure out whether wait(2) should be called.
This heuristic worked for a particular use case, but is not correct in
general.
While at it:
- simplify signalAllProcesses to use unix.Kill;
- document (container).Signal.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is roughly the same as TestPIDHostInitProcessWait in libct/int,
except that here we use separate processes to create and to kill a
container, so the processes inside a container are not children of "runc kill", and
also we hit different codepaths (nonChildProcess.signal rather than
initProcess.signal).
One other thing is, rootless is also tested.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It seems that set -x was temporarily added as a debug measure, but
slipped into the final commit.
Remove it, for the sake of test logs brevity.
Fixes: 9f656dbb11
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test is occasionally failing on CS9.
The test case always takes about 7 seconds on my laptop (decreasing
memory, using a different memory eater in shell etc. doesn't help).
Increase the number of iterations from 10 to 30 to make sure we don't
see any flakes.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Fix some missing punctuation, use proper case.
2. Remove "runc init" (previously removed from "runc --help" by commit
7a0302f0d7).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As of today, vagrant stopped working, my best guess is due to bad
caching. Here's an excerpt from logs:
...
vagrant plugin install vagrant-libvirt
Installing the 'vagrant-libvirt' plugin. This can take a few minutes...
Building native extensions. This could take a while...
Installed the plugin 'vagrant-libvirt (0.12.1)'!
...
uname -s ; cat Vagrantfile.$DISTRO
Linux
...
Downloaded 481Mb in 4.096201s.
Cache hit for vagrant-8be35383dc00f23d080ff00b2a724c938d650254861f26b67624c28e3fe5e6ae!
...
Vagrant failed to initialize at a very early stage:
The plugins failed to initialize correctly. This may be due to manual
modifications made within the Vagrant home directory.
...
Error message given during initialization: Unable to resolve dependency:
user requested 'vagrant-libvirt (= 0.12.0)'
...
The problem is, vagrant cache overwrites newer plugin with an older one.
Let's only cache the downloaded image.
Also, change the cache fingerprint script (remove "Linux").
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently, TestInit sets up logrus, and init uses it to log an error
from StartInitialization(). This is solely used by TestExecInError
to check that error returned from StartInitialization is the one it
expects.
Note that the very same error is communicated to the runc init parent
and is ultimately returned by container.Run(), so checking what
StartInitialization returned is redundant.
Remove logrus setup and use from TestMain/init.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a cosmetic change to improve code readability, making it easier
to distinguish between a local error and the error being returned.
While at it, rename e to err (it was originally called e to not clash
with returned error named err) and ee to err2.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of having newContainerInit return an interface, and let its
caller call Init(), it is easier to call Init directly.
Do that, and rename newContainerInit to containerInit.
I think it makes the code more readable and straightforward.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit e3cf217cf1 actions/setup-go@v4 uses caching
implicitly, and olangci/golangci-lint-action also uses caching.
These two caches clash, resulting in multiple warnings in CI logs.
The official golangci-lint-action solution is to disable caching
for setup-go job (see [1]). Do the same.
[1] https://github.com/golangci/golangci-lint-action/pull/704
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As of today, installing fails with
> libc6:i386 : Depends: libgcc-s1:i386 but it is not going to be installed
Add the package explicitly to work around that.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The first kludge is not needed since the switch to Ubuntu 22.04 in
commit 953e1cc48.
The second one is not needed since Go 1.20.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Adding fd field to mountConfig was not a good thing since mountConfig
contains data that is not specific to a particular mount, while fd is
a mount entry attribute.
Introduce mountEntry structure, which embeds configs.Mount and adds
srcFd to replace the removed mountConfig.fd.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Simplify mount call by removing the procfd argument, and use the new
mount() where procfd is not used. Now, the mount() arguments are the
same as for unix.Mount.
2. Introduce a new mountViaFDs function, which is similar to the old
mount(), except it can take procfd for both source and target.
The new arguments are called srcFD and dstFD.
3. Modify the mount error to show both srcFD and dstFD so it's clear
which one is used for which purpose. This fixes the issue of having
a somewhat cryptic errors like this:
> mount /proc/self/fd/11:/sys/fs/cgroup/systemd (via /proc/self/fd/12), flags: 0x20502f: operation not permitted
(in which fd 11 is actually the source, and fd 12 is the target).
After this change, it looks like
> mount src=/proc/self/fd/11, dst=/sys/fs/cgroup/systemd, dstFD=/proc/self/fd/12, flags=0x20502f: operation not permitted
so it's clear that 12 is a destination fd.
4. Fix the mountViaFDs callers to use dstFD (rather than procfd) for the
variable name.
5. Use srcFD where mountFd is set.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When given an environment variable that is invalid, it's not a good idea
to output the contents in case they are supposed to be private (though
such a container wouldn't start anyway so it seems unlikely there's a
real way to use this to exfiltrate environment variables you didn't
already know).
Reported-by: Carl Henrik Lunde <chlunde@ifi.uio.no>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This field used to hold a string representation of log level (like
"debug" or "info"). Since commit 6c4a3b13d1 this is now a string
holding a numeric representation of log level (e.g. "4"). This was done
in preparation to commit f1b703fc45, and simplifies things in
a few places.
Let's document it.
Fixes: 6c4a3b13d1
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Use strings.TrimPrefix instead of fmt.Sscanf and simplify the code.
2. Add a test case and a benchmark.
The benchmark shows some improvement, compared to the old
implementation:
name old time/op new time/op delta
FindDeviceGroup-4 39.7µs ± 2% 26.8µs ± 2% -32.63% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
FindDeviceGroup-4 6.08kB ± 0% 4.23kB ± 0% -30.39% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
FindDeviceGroup-4 117 ± 0% 6 ± 0% -94.87% (p=0.008 n=5+5)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A version of vagrant available from the stock repos (2.2.19) is too old
and contains a bug that prevents downloading Fedora 38 image (see [1]).
Use packages from hashicorp repo, which currently has vagrant 2.3.4.
This resolves the problem of downloading the latest Fedora image.
Also, vagrant-libvirt plugin from Ubuntu repos is not working with
vagrant from hashicorp, so switch to using "vagrant plugin install".
The downside it, this takes extra 4 minutes or so in our CI, and I
am not sure how to cache it or speed it up.
[1] https://github.com/opencontainers/runc/pull/3835#issuecomment-1519321619
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 343951a22b added a call to os.Stat for the device path
when generating systemd device properties, to avoid systemd warning for
non-existing devices. The idea was, since systemd uses stat(2) to look
up device properties for a given path, it will fail anyway. In addition,
this allowed to suppress a warning like this from systemd:
> Couldn't stat device /dev/char/10:200
NOTE that this was done because:
- systemd could not add the rule anyway;
- runs puts its own set of rules on top of what systemd does.
Apparently, the above change broke some setups, resulting in inability
to use e.g. /dev/null inside a container. My guess is this is because
in cgroup v2 we add a second eBPF program, which is not used if the
first one (added by systemd) returns "access denied".
Next, commit 3b9582895b fixed that by adding a call to os.Stat for
"/sys/"+path (meaning, if "/dev/char/10:200" does not exist, we retry
with "/sys/dev/char/10:200", and if it exists, proceed with adding a
device rule with the original (non-"/sys") path).
How that second fix ever worked was a mystery, because the path we gave
to systemd still doesn't exist.
Well, I think now I know.
Since systemd v240 (commit 74c48bf5a8005f20) device access rules
specified as /dev/{block|char}/MM:mm are no longer looked up on the
filesystem, instead, if possible, those are parsed from the string.
So, we need to do different things, depending on systemd version:
- for systemd >= v240, use the /dev/{char,block}/MM:mm as is, without
doing stat() -- since systemd doesn't do stat() either;
- for older version, check if the path exists, and skip passing it on
to systemd otherwise.
- the check for /sys/dev/{block,char}/MM:mm is not needed in either
case.
Pass the systemd version to the function that generates the rules, and
fix it accordingly.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For a previous attempt to fix that (and added test cases), see commit
9087f2e827.
Alas, it's not always working because of cgroup directory TOCTOU.
To solve this and avoid the race, add an error _after_ the operation.
Implement it as a method that ignores the error that should be ignored.
Instead of currentStatus(), use faster runType(), since we are not
interested in Paused status here.
For Processes(), remove the pre-op check, and only use it after getting
an error, making the non-error path more straightforward.
For Signal(), add a second check after getting an error. The first check
is left as is because signalAllProcesses might print a warning if the
cgroup does not exist, and we'd like to avoid that.
This should fix an occasional failure like this one:
not ok 84 kill detached busybox
# (in test file tests/integration/kill.bats, line 27)
# `[ "$status" -eq 0 ]' failed
....
# runc kill test_busybox KILL (status=0):
# runc kill -a test_busybox 0 (status=1):
# time="2023-04-04T18:24:27Z" level=error msg="lstat /sys/fs/cgroup/devices/system.slice/runc-test_busybox.scope: no such file or directory"
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These checks ensure that all of the keys in the runc.keyring list are
actually the keys of the specified user and that the users themselves
are actually maintainers.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We need to make sure the release is being signed by a key that is
actually listed as a trusted signing key, and we also need to ask the
person cutting the release whether the list of trusted keys is
acceptable.
Also add some verification checks after a release is signed to make sure
everything was signed with the correct keys.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In order to allow any of the maintainers to cut releases for runc,
create a keyring file that distributions can use to verify that releases
are signed by one of the maintainers.
The format matches the gpg-offline format used by openSUSE packaging,
but it can be easily imported with "gpg --import" so any distribution
should be able to handle this keyring format wtihout issues.
Each key includes the GitHub handle of the associated user. There isn't
any way for this information to be automatically verified (outside of
using something like keybase.io) but since all changes of this file need
to be approved by maintainers this is okay for now.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
For test jobs, add ubuntu 22.04 into the matrix, so we can test of both
cgroup v1 and v2.
For validate jobs, just switch to ubuntu 22.04
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Do not use echo, as this results in lines like this:
...
echo "-----"
-----
...
2. Move "cat /proc/cpuinfo" to be the last one, as the output is usually
very long.
3. Add "go version" to CentOS jobs.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This variable is used in curl to download a go release, so we are using
the initial Go 1.19 release in Cirrus CI, not the latest Go 1.19.x
release.
From the CI perspective, it makes more sense to use the latest release.
Add some jq magic to extract the latest minor release information
from the download page, and use it.
This brings Cirrus CI jobs logic in line with all the others (GHA,
Dockerfile), where by 1.20 we actually mean "latest 1.20.x".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently, bash with set -e deliberately ignores non-zero return codes
from ! cmd, unless this is the last command. The workaround is to either
use "! cmd || false', "or run ! cmd". Choose the latter, and require
bash-core 1.5.0 (since this is when "run !" was added), replacing the
older check.
Alas I only learned this recently from the bash-core documentation.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This version is already used by Cirrus CI Fedora 37 job, but other CI
jobs are still using 1.3.0.
Bump it everywhere so we can enjoy new version features and fixes.
For one thing, I noticed that new bats is reporting error location
correctly.
We will also be able to use "run !" and "run -N" commands.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These functions were added in ancient times, facilitating the
docker-in-docker case when cgroup namespace was not available.
As pointed out in commit 2b28b3c276, using init 1 cgroup is not
correct because it won't work in case of host PID namespace.
The last user of GetInitCgroup was removed by commit
54e20217a8. GetInitCgroupPath was never used
as far as I can see, nor was I able to find any external users.
Remove both functions. Modify the comment in libct/cg/fs.subsysPath
to not refer to GetInitCgroupPath.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Systemd v252 (available in CentOS Stream 9 in our CI) added support
for setting cpu.idle (see [1]). The way it works is:
- if CPUWeight == 0, cpu.idle is set to 1;
- if CPUWeight != 0, cpu.idle is set to 0.
This commit implements setting cpu.idle in systemd cgroup driver via a
unit property. In case CPUIdle is set to non-zero value, the driver sets
adds CPUWeight=0 property, which will result in systemd setting cpu.idle
to 1.
Unfortunately, there's no way to set cpu.idle to 0 without also changing
the CPUWeight value, so the driver doesn't do anything if CPUIdle is
explicitly set to 0. This case is handled by the fs driver which is
always used as a followup to setting systemd unit properties.
Also, handle cpu.idle set via unified map. In case it is set to non-zero
value, add CPUWeight=0 property, and ignore cpu.weight (otherwise we'll
get two different CPUWeight properties set).
Add a unit test for new values in unified map, and an integration test case.
[1] https://github.com/systemd/systemd/pull/23299
[2] https://github.com/opencontainers/runc/issues/3786
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Values other than 1 or 0 are ignored by the kernel,
see sched_group_set_idle() in kernel/sched/fair.c
If the added test case ever fails, it means that the kernel now accepts
values other than 0 or 1, and runc needs to adopt to that.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Systemd v252 (available in CentOS Stream 9 in our CI) added support
for setting cpu.idle (see [1]). The way it works is:
- if CPUWeight == 0, cpu.idle is set to 1;
- if CPUWeight != 0, cpu.idle is set to 0.
This behavior breaks the existing test case, as described in [2].
To fix, skip the last check in the test case in case a newer systemd
is used.
Fixes: #3786
[1] https://github.com/systemd/systemd/pull/23299
[2] https://github.com/opencontainers/runc/issues/3786
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In code that checks that the resource name is in the for
Using strings.SplitN is an overkill in this case, resulting in
allocations and thus garbage to collect.
Using strings.IndexByte and checking that result is not less than 1
(meaning there is a period, and it is not the first character) is
sufficient here.
Fixes: 0cb8bf67a3
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit d08bc0c1b3 ("runc run: warn on non-empty cgroup") introduced
a warning when a container is started in a non-empty cgroup. Such
configuration has lots of issues.
In addition to that, such configuration is not possible at all when
using the systemd cgroup driver.
As planned, let's promote this warning to an error, and fix the test
case accordingly.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case a systemd unit fails (for example, timed out or OOM-killed),
systemd keeps the unit. This prevents starting a new container with
the same systemd unit name.
The fix is to call reset-failed in case UnitExists error is returned,
and retry once.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit d223e2adae ("Ignore error when starting transient unit
that already exists" modified the code handling errors from startUnit
to ignore UnitExists error.
Apparently it was done so that kubelet can create the same pod slice
over and over without hitting an error (see [1]).
While it works for a pod slice to ensure it exists, it is a gross bug
to ignore UnitExists when creating a container. In this case, the
container init PID won't be added to the systemd unit (and to the
required cgroup), and as a result the container will successfully
run in a current user cgroup, without any cgroup limits applied.
So, fix the code to only ignore UnitExists if we're not adding a process
to the systemd unit. This way, kubelet will keep working as is, but
runc will refuse to create containers which are not placed into a
requested cgroup.
[1] https://github.com/opencontainers/runc/pull/1124
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit 39914db679 this function is not used by runc (see
that commit to learn why this function is not that good).
I was not able to find any external users either.
Since it's not a good function, with no users, and it is rather trivial,
let's remove it right away (rather than mark as deprecated).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
... as a way to maybe catch some CHANGELOG.md bugs at the last moment.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
(cherry picked from commit 54cfb25d69)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Previously (see commit 91fa032da4) we found a few issues
using this check, but apparently the CHANGELOG.md is in UTF-8, and
the recently added quote is breaking this, so remove.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
(cherry picked from commit 7b3ac330f7)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes a bogus failure in "ro cgroup" test cases when running as
rootless.
The test finds the following mount that is not read-only:
> cgroup2 /sys/fs/cgroup/unified cgroup2 rw,nosuid,nodev,noexec,relatime,nsdelegate 0 0
This happens because:
1. runc spec --rootless adds an rbind /sys mounts, so we have all the
/sys/fs/cgroup/XXX mounts inside the container;
2. Those /sys/fs/cgroup/XXX mounts are shadowed by the /sys/fs/cgroup
tmpfs mount created by mountCgroupV1().
This means that this mount is shadowed, inaccessible, and it can not be
unshadowed, thus it should not be checked.
The fix is to check whether the directory exists, to exclude such
shadowed mounts.
NOTE that item 2 comes from commit ff692f289b60e19b3079cb; before it, we
had the whole hierarchy of host /sys/fs/cgroup visible (though not
writable -- because rootless) from inside of any rootless container.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Fix the following issues with the "ro /sys/fs/cgroup" test:
1. Disable bogus SC2016 warning from shellcheck.
2. Split the test into two -- with and without cgroupns. This is done
because not all systems have cgroupns available (so the "+cgroupns"
test will be skipped).
3. This splitting resulted in a few more bogus shellcheck warnings that
we have to suppress -- due to a known bug in shellcheck (see [1]).
4. s/mount/mounts/ in the test name, because in case of cgroup v1 there
are multiple mounts.
[1] https://github.com/koalaman/shellcheck/issues/2431
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a test case for issue reported as #3715. In short, even if a
(non-root) user that the container is run as does not have execute
permission bit set for the executable, it should still work in case runc
has the CAP_DAC_OVERRIDE capability set.
Note that since the upstream golang is also broken (see [1]), this test
will fail for Go 1.20 and 1.20.1 (fix is in Go 1.20.2 as per [2]).
[1] https://go.dev/issue/58552
[2] https://go-review.googlesource.com/c/go/+/469956
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Go 1.20.2 has an important fix to an issue described in [1].
Switch from using Go 1.19 from Dockerfile, which is used for release
binaries and some CI.
[1] https://github.com/golang/go/issues/58624
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The setRecAttr is only called for "bind" case, as cases end with a
return statement. Indeed, recursive mount attributes only make sense for
bind mounts.
Move the code to under case "bind" to improve readability. No change in
logic.
Fixes: 382eba4354
Reported-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Sometimes, the init process is not in the root cgroup.
This can be noted by GetInitPath, which already scrubs the path of `init.scope`.
This was encountered when trying to patch the Kubelet to handle systemd being in a separate cpuset
from root (to allow load balance disabling for containers). At present, there's no way to have libcontainer or runc
manage cgroups in a hierarchy outside of the one init is in (unless the path contains `init.scope`, which is limiting)
Signed-off-by: Peter Hunt <pehunt@redhat.com>
There is some kind of a race in CentOS 7 which sometimes result in one
of these tests failing like this:
systemd_test.go:136: mkdir /sys/fs/cgroup/hugetlb/system.slice/system-runc_test_pods.slice: no such file or directory
or
systemd_test.go:187: open /sys/fs/cgroup/cpuset/system.slice/system-runc_test_pods.slice/cpuset.mems: no such file or directory
As this is only happening on CentOS 7, let's skip this test on this
platform.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 3291d66b98 introduced a check for /proc and /sys, making sure
the destination (dest) is a directory (and not e.g. a symlink).
Later, a hunk from commit 0ca91f44f switched from using filepath.Join
to SecureJoin for dest. As SecureJoin follows and resolves symlinks,
the check whether dest is a symlink no longer works.
To fix, do the check without/before using SecureJoin.
Add integration tests to make sure we won't regress.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Older kernels may return EINVAL on unshare when a process is reading
runc's /proc/$PID/status or /proc/$PID/maps. This was fixed by kernel
commit 12c641ab8270f ("unshare: Unsharing a thread does not require
unsharing a vm") in Linuxt v4.3.
For CentOS 7, the fix was backported to CentOS 7.7 (kernel 3.10.0-1062).
To work around this kernel bug, let's retry on EINVAL a few times.
Reported-by: zzyyzte <zhang.yu58@zte.com.cn>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It was found that rootless runc makes `/sys/fs/cgroup` writable in following conditons:
1. when runc is executed inside the user namespace, and the config.json does not specify the cgroup namespace to be unshared
(e.g.., `(docker|podman|nerdctl) run --cgroupns=host`, with Rootless Docker/Podman/nerdctl)
2. or, when runc is executed outside the user namespace, and `/sys` is mounted with `rbind, ro`
(e.g., `runc spec --rootless`; this condition is very rare)
A container may gain the write access to user-owned cgroup hierarchy `/sys/fs/cgroup/user.slice/...` on the host.
Other users's cgroup hierarchies are not affected.
To fix the issue, this commit does:
1. Remount `/sys/fs/cgroup` to apply `MS_RDONLY` when it is being bind-mounted
2. Mask `/sys/fs/cgroup` when the bind source is unavailable
Fix CVE-2023-25809 (GHSA-m8cg-xc2p-r3fc)
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
The purpose of this test is to check that there are no extra file
descriptors left open after repeated calls to runContainer. In fact,
the first call to runContainer leaves a few file descriptors opened,
and this is by design.
Previously, this test relied on two things:
1. some other tests were run before it (and thus all such opened-once
file descriptors are already opened);
2. explicitly excluding fd opened to /sys/fs/cgroup.
Now, if we run this test separately, it will fail (because of 1 above).
The same may happen if the tests are run in a random order.
To fix this, add a container run before collection the initial fd list,
so those fds that are opened once are included and won't be reported.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
stage_2_pid is not yet assigned, so this kills the PID -1, but as
the sane_kill() wrapper is just a nop in that case. Just remove these
calls to kill stage_2_pid before it is cloned/assigned.
I've checked by executing the error paths that no binary is left by mistake.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
centos-9 unit test sometimes fails with:
=== RUN TestPodSkipDevicesUpdate
systemd_test.go:114: container stderr not empty: basename: missing operand
Try 'basename --help' for more information.
--- FAIL: TestPodSkipDevicesUpdate (0.11s)
I'm not sure why the container output is an error in basename. It seems
likely that the bashrc in that distro is kind of broken. Let's just run
a sleep command and forget about bash.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
With the updated git in golang:1.19-bullseye image, building fails with:
make -C /go/src/github.com/opencontainers/runc PKG_CONFIG_PATH=/opt/libseccomp/lib/pkgconfig COMMIT_NO= EXTRA_FLAGS=-a 'EXTRA_LDFLAGS=-w -s -buildid=' static
make[1]: Entering directory '/go/src/github.com/opencontainers/runc'
fatal: detected dubious ownership in repository at '/go/src/github.com/opencontainers/runc'
To add an exception for this directory, call:
git config --global --add safe.directory /go/src/github.com/opencontainers/runc
go build -trimpath -buildmode=pie -a -tags "seccomp urfave_cli_no_docs netgo osusergo" -ldflags "-X main.gitCommit= -X main.version=1.1.0+dev -linkmode external -extldflags --static-pie -w -s -buildid=" -o runc .
error obtaining VCS status: exit status 128
Use -buildvcs=false to disable VCS stamping.
This commit should fix it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In 18c4760a (libct: fixStdioPermissions: skip chown if not needed)
the check whether the STDIO file descriptors point to /dev/null was
removed which can cause /dev/null to change ownership e.g. when using
docker exec on a running container:
$ ls -l /dev/null
crw-rw-rw- 1 root root 1, 3 Aug 1 14:12 /dev/null
$ docker exec -u test 0ad6d3064e9d ls
$ ls -l /dev/null
crw-rw-rw- 1 test root 1, 3 Aug 1 14:12 /dev/null
Signed-off-by: Jaroslav Jindrak <dzejrou@gmail.com>
For the sake of developers who have LSP configured to auto-format the
code upon save (that would me with my new nvim setup), let's not
autoformat the C code when using clangd.
Initially I tried to write a set of rules for clang-format which is
identical to what we use (indent with a handful of options invoked
from cfmt target in Makefile), but it appears to be impossible.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
First, check if strdup() fails and error out.
While we are there, the else case was missing brackets, as we only need
to check ret in the else case. Fix that too
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
This is to check that tests/integration/get-images.sh is in sync
with tests/integration/bootstrap-get-images.sh.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
configs package can no longer be built on non-Linux OS, such as Darwin.
When running `GOOS=darwin go build` on the packge, we had the following
errors:
```
./configs/mount.go:34:16: undefined: unix.MountAttr
./configs/mount.go:47:22: undefined: unix.MS_BIND
```
Let's ensure that the linux specific bits are handled in mount_linux.go,
and introduce a _unsupported file, similar to how cgroups file is
handled within the package. This'll facilitate utilization of the pkg
for other projects that care about Darwin.
Signed-off-by: Eric Ernst <eric_ernst@apple.com>
Rewrite systemdVersionAtoi to not use regexp, and fix two issues:
1. It was returning 0 (rather than -1) for some errors.
2. The comment was saying that the input string is without quotes,
while in fact it is.
Note the new function, similar to the old one, works on input either
with or without quotes. Amend the test to add test cases without quotes.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A couple of test cases in delete.bats check that a particular cgroup
exists (or doesn't exist) using find. This is now resulting in errors
like these:
find: ‘/sys/fs/cgroup/blkio/azsec’: Permission denied
find: ‘/sys/fs/cgroup/blkio/azsec_clamav’: Permission denied
find: ‘/sys/fs/cgroup/cpu,cpuacct/azsec’: Permission denied
find: ‘/sys/fs/cgroup/cpu,cpuacct/azsec_clamav’: Permission denied
find: ‘/sys/fs/cgroup/memory/azsec’: Permission denied
find: ‘/sys/fs/cgroup/memory/azsec_clamav’: Permission denied
leading to test case failures.
Apparently, GHA runs something else on a test box, so we get this.
To fix, ignore non-zero exit code from find, and redirect its stderr
to /dev/null.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When doing a lazy checkpoint/restore, we should not restore into the
same cgroup, otherwise there is a race which result in occasional
killing of the restored container (GH #2760, #2924).
The fix is to use --manage-cgroup-mode=ignore, which allows to restore
into a different cgroup.
Note that since cgroupsPath is not set in config.json, the cgroup is
derived from the container name, so calling set_cgroups_path is not
needed.
For the previous (unsuccessful) attempt to fix this, as well as detailed
(and apparently correct) analysis, see commit 36fe3cc28c.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test checks that the container is restored into a different cgroup.
To do so, a user should
- use --manage-cgroups-mode ignore on both checkpoint and restore;
- change the cgroupsPath value in config.json before restoring.
The test does some checks to ensure that its logic is correct, and that
after the restore the old (original) cgroup does not exist, the new one
exists, and the container's init is in that new cgroup.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I don't want to implement it now, because this might result in some
new issues, but this is definitely something that is worth implementing.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
- add the new mode and document it;
- slightly improve the --help output;
- slightly simplify the parsing code.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Merge the logic of setPageServer, setManageCgroupsMode, and
setEmptyNsMask into criuOptions. This does three things:
1. Fixes ignoring --manage-cgroups-mode on restore;
2. Simplifies the code in checkpoint.go and restore.go;
3. Ensures issues like 1 won't happen again.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test (initially added by commit 58ea21daef and later amended in
commit 26dc55ef1a) currently has two major deficiencies:
1. All possible flag combinations, and their respective numeric values,
have to be explicitly listed. Currently we support 3 flags, so
there is only 2^3 - 1 = 7 combinations, but adding more flags will
become increasingly difficult (for example, 5 flags will result in
31 combinations).
2. The test requires kernel 4.17 (for SECCOMP_FILTER_FLAG_SPEC_ALLOW),
and not doing any tests when running on an older kernel. This, too,
will make it more difficult to add extra flags in the future.
Both issues can be solved by using runc features which now prints all
known and supported runc flags. We still have to hardcode the numeric
values of all flags, but most of the other work is coded now.
In particular:
* The test only uses supported flags, meaning it can be used with
older kernels, removing the limitation (2) above.
* The test calculates the powerset (all possible combinations) of
flags and their numeric values. This makes it easier to add more
flags, removing the limitation (1) above.
* The test will fail (in flags_value) if any new flags will be added
to runc but the test itself is not amended.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If no seccomps flags are set in OCI runtime spec (not even the empty
set), set SPEC_ALLOW as the default (if it's supported).
Otherwise, use the flags as they are set (that includes no flags for
empty seccomp.Flags array).
This mimics the crun behavior, and makes runc seccomp performance on par
with crun.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Amend runc features to print seccomp flags. Two set of flags are added:
* known flags are those that this version of runc is aware of;
* supported flags are those that can be set; normally, this is the same
set as known flags, but due to older version of kernel and/or
libseccomp, some known flags might be unsupported.
This commit also consolidates three different switch statements dealing
with flags into one, in func setFlag. A note is added to this function
telling what else to look for when adding new flags.
Unfortunately, it also adds a list of known flags, that should be
kept in sync with the switch statement.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Init State Error message was using the err variable instead of uerr, which has been fixed now.
The error message should not show "nil" now.
Signed-off-by: Vipul Newaskar <vipulnewaskar7@gmail.com>
If checkpointing has failed, the container is kept running. We do not
want to, and we can't remove it in such case.
Do not try to remove the container if there's an error from
checkpointing.
This avoids an unclear error message from destroy() saying "container
still running" or "container paused".
While at it, avoid using defer since it does not make a lot of sense
here.
Fixes: #3577
Reported-by: gosoon <tianfeiyu0@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is aimed at solving the problem of cgroup v2 memory controller
behavior which is not compatible with that of cgroup v1.
In cgroup v1, if the new memory limit being set is lower than the
current usage, setting the new limit fails.
In cgroup v2, same operation succeeds, and the container is OOM killed.
Introduce a new setting, memory.checkBeforeUpdate, and use it to mimic
cgroup v1 behavior.
Note that this is not 100% reliable because of TOCTOU, but this is the
best we can do.
Add some test cases.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Bump shfmt to v3.5.1. Release notes:
https://github.com/mvdan/sh/releases
2. Since shfmt v3.5.0, specifying -l bash (or -l bats) is no longer
necessary. Therefore, we can use shfmt to find all the files.
Add .editorconfig to ignore vendor subdirectory.
3. Use shfmt docker image, so that we don't have to install anything
explicitly. This greatly simplifies the shfmt CI job. Add
localshfmt target so developers can still use a local shfmt binary
when necessary.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In findLastSyscalls, we convert libseccomp.ArchNative to the real
libseccomp architecture, but archToNative already does that, so
this code is redundant.
Remove the redundant code, and move its comment to archToNative.
Fixes: 7a8d7162f
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This makes libseccomp produce a BPF which uses a binary tree for
syscalls (instead of linear set of if statements).
It does not make sense to enable binary tree for small set of rules,
so don't do that if we have less than 8 syscalls (the number is chosen
arbitrarily).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
On ARM, mkdirat(2) is used instead of mkdir(2), thus the seccomp rules
needs to be amended accordingly.
This is a change similar to one in commit e119db7a23, but but it
evaded the test case added in commit 58ea21dae as it took a long time to
merge, and we don't have ARM CI.
Fixes: 58ea21dae ("seccomp: add support for flags")
Reported-by: Ryan Phillips <rphillips@redhat.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This tag points to the latest v3 version (currently v3.0.11). Mainly
done to avoid cluttering git history with multiple minor v3 upgrades.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add a test case for an issue fixed by the previous commit.
The env should has more than 8 core CPU to meet the test requirement.
Signed-off-by: Chengen, Du <chengen.du@canonical.com>
Runc parses cpuset range to bits in the case of cgroup v2 + systemd as cgroup driver.
The byte order representation differs from systemd expectation, which will set
different cpuset range in systemd transient unit if the length of parsed byte array exceeds one.
# cat config.json
...
"resources": {
...
"cpu": {
"cpus": "10-23"
}
},
...
# runc --systemd-cgroup run test
# cat /run/systemd/transient/runc-test.scope.d/50-AllowedCPUs.conf
# This is a drop-in unit file extension, created via "systemctl set-property"
# or an equivalent operation. Do not edit.
[Scope]
AllowedCPUs=0-7 10-15
The cpuset.cpus in cgroup will also be set to wrong value after reloading systemd manager configuration.
# systemctl daemon-reload
# cat /sys/fs/cgroup/system.slice/runc-test.scope/cpuset.cpus
0-7,10-15
Signed-off-by: seyeongkim <seyeong.kim@canonical.com>
Signed-off-by: Chengen, Du <chengen.du@canonical.com>
This is a forward port of https://github.com/opencontainers/runc/pull/3620
The original code depended on the origin filesystem to have
/dev/{block,char} populated. This is done by udev normally and while is
very common non-containerized systemd installs, it's very easy to start
systemd in a container created by runc itself and not have
/dev/{block,char} populated. When this occurs, the following error
output is observed:
$ docker run hello-world
docker: Error response from daemon: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error reopening /dev/null inside container: open /dev/null: operation not permitted: unknown.
/dev/null can't be opened because it was not added to the
deviceAllowList, as there was no /dev/char directory. The change here
utilizes the fact that when sysfs in in use, there is a
/sys/dev/{block,char} that is kernel maintained that we can check.
Signed-off-by: Evan Phoenix <evan@phx.io>
It's more idiomatic Go to define interfaces on the receiver, and constructors to
return concrete types.
This patch changes various constructors to return a concrete type, with the
exceptions of NewWithPaths, which needs the abstraction as it switches between
implementations.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
It used to be enabled by default, but not as of last few weeks.
While at it, add rpm -q command to make sure all required RPMS were in
fact installed (at least CentOS 7 yum exits with 0 when some packages
requested are not available).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We use awk in other 9 or so places, and here it's gawk.
Since this is on Linux, most probably awk is gawk.
So s/gawk/awk/.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The situation when /sys/fs/cgroup/unified is not present normal and
should not result in anything on stderr. Suppress it.
Fixes: cc15b887a0
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
cgroup v2 requires CONFIG_CGROUP_BPF kernel option to be set
else runc can not start containers.
check-config.sh script checks if the CONFIG_CGROUP_BPF option
is set. The script checks if version of kernel is atleast
4.15 and cgroup v2 is being used before checking if the
CONFIG_CGROUP_BPF option is set.
Closes#3547
Signed-off-by: dharmicksai <dharmicksaik@gmail.com>
The v6.0.0 release of go-criu has deprecated the `rpc` package in favour
of the `crit` package. This commit provides the changes required to use
this version in runc.
Signed-off-by: Prajwal S N <prajwalnadig21@gmail.com>
Add a debug print of seccomp flags value, so the test can check
those (without using something like strace, that is).
Amend the flags setting test with the numeric values expected, and the
logic to check those.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 58ea21daef added support for seccomp flags such as
SPEC_ALLOW, but it does not work as expected, because since commit
7a8d7162f9 we do not use libseccomp-golang's Load(), but
handle flags separately in patchbfp.
This fixes setting SPEC_ALLOW flag.
Add a comment to not forget to amend filterFlags when adding new flags.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A regression reported for runc v1.1.3 says that "runc exec -t" fails
after doing "systemctl daemon-reload":
> exec failed: unable to start container process: open /dev/pts/0: operation not permitted: unknown
Apparently, with commit 7219387eb7 we are no longer adding
"DeviceAllow=char-pts rwm" rule (because os.Stat("char-pts") returns
ENOENT).
The bug can only be seen after "systemctl daemon-reload" because runc
also applies the same rules manually (by writing to devices.allow for
cgroup v1), and apparently reloading systemd leads to re-applying the
rules that systemd has (thus removing the char-pts access).
The fix is to do os.Stat only for "/dev" paths.
Also, emit a warning that the path was skipped. Since the original idea
was to emit less warnings, demote the level to debug.
Note this also fixes the issue of not adding "m" permission for block-*
and char-* devices.
A test case is added, which reliably fails before the fix
on both cgroup v1 and v2.
Fixes: https://github.com/opencontainers/runc/issues/3551
Fixes: 7219387eb7
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Recently released codespell 2.2 adds some more false positives,
such as:
./Makefile:78: ro ==> to, row, rob, rod, roe, rot
./Makefile:88: ro ==> to, row, rob, rod, roe, rot
./notify_socket.go:51: ro ==> to, row, rob, rod, roe, rot
./LICENSE:128: complies ==> compiles
./go.sum:59: BU ==> BY
./types/features/features.go:17: ro ==> to, row, rob, rod, roe, rot
./libcontainer/rootfs_linux.go:52: ro ==> to, row, rob, rod, roe, rot
./libcontainer/rootfs_linux.go:166: ro ==> to, row, rob, rod, roe, rot
....
./tests/integration/cgroup_delegation.bats:38: inh ==> in
...
To fix:
- exclude go.sum;
- add ro and complies to the list of ignored words;
- s/inh/inherit in cgroup_delegation.bats.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since Go 1.19, godoc recognizes lists, code blocks, headings etc. It
also reformats the sources making it more apparent that these features
are used.
Fix a few places where it misinterpreted the formatting (such as
indented vs unindented), and format the result using the gofumpt
from HEAD, which already incorporates gofmt 1.19 changes.
Some more fixes (and enhancements) might be required.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When golang 1.19 is used to build unit tests on 386, it fails like this:
sudo -E PATH="$PATH" -- make GOARCH=386 CGO_ENABLED=1 localunittest
<...>
go test -timeout 3m -tags "seccomp" -v ./...
<...>
# github.com/opencontainers/runc/libcontainer/capabilities.test
runtime/cgo(.text): unknown symbol __stack_chk_fail_local in pcrel
runtime/cgo(.text): unknown symbol __stack_chk_fail_local in pcrel
runtime/cgo(.text): unknown symbol __stack_chk_fail_local in pcrel
runtime/cgo(.text): unknown symbol __stack_chk_fail_local in pcrel
runtime/cgo(.text): unknown symbol __stack_chk_fail_local in pcrel
runtime/cgo(.text): relocation target __stack_chk_fail_local not defined
runtime/cgo(.text): relocation target __stack_chk_fail_local not defined
The fix is to add CGO_CFLAGS=-fno-stack-protector.
See also:
- https://github.com/docker-library/golang/pull/426
- https://go.dev/issue/52919
- https://go.dev/issue/54313
- https://go-review.googlesource.com/c/go/+/421935
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Unless the container's runtime config has intelRdt configuration set,
any checks for whether Intel RDT is supported or the resctrl filesystem
is mounted are a waste of time as, per the OCI Runtime Spec, "the
runtime MUST NOT manipulate any resctrl pseudo-filesystems." And in the
likely case where Intel RDT is supported by both the hardware and
kernel but the resctrl filesystem is not mounted, these checks can get
expensive as the intelrdt package needs to parse mountinfo to check
whether the filesystem has been mounted to a non-standard path.
Optimize for the common case of containers with no intelRdt
configuration by only performing the checks when the container has opted
in.
Signed-off-by: Cory Snider <csnider@mirantis.com>
The OCI runtime spec mandates "[i]f intelRdt is not set, the runtime
MUST NOT manipulate any resctrl pseudo-filesystems." Attempting to
delete files counts as manipulating, so stop doing that when the
container's RDT configuration is nil.
Signed-off-by: Cory Snider <csnider@mirantis.com>
The intelrdt package only needs to parse mountinfo to find the mount
point of the resctrl filesystem. Users are generally going to mount the
resctrl filesystem to the pre-created /sys/fs/resctrl directory, so
there is a common case where mountinfo parsing is not required. Optimize
for the common case with a fast path which checks both for the existence
of the /sys/fs/resctrl directory and whether the resctrl filesystem was
mounted to that path using a single statfs syscall.
Signed-off-by: Cory Snider <csnider@mirantis.com>
Reading /proc/cpuinfo is a surprisingly expensive operation. Since
kernel version 4.12 [1], opening /proc/cpuinfo on an x86 system can
block for around 20 milliseconds while the kernel samples the current
CPU frequency. There is a very recent patch [2] which gets rid of the
delay, but has yet to make it into the mainline kenel. Regardless,
kernels for which opening /proc/cpuinfo takes 20ms will continue to be
run in production for years to come. libcontainer only opens
/proc/cpuinfo to read the processor feature flags so all the delays to
get an accurate snapshot of the CPU frequency are just wasted time.
If we wanted to, we could interrogate the CPU features directly from
userspace using the `CPUID` instruction. However, Intel and AMD CPUs
have flags in different positions for their analogous sub-features and
there are CPU quirks [3] which would need to be accounted for. Some
Haswell server CPUs support RDT/CAT but are missing the `CPUID` flags
advertising their support; the kernel checks for support on that
processor family by probing the the hardware using privileged
RDMSR/WRMSR instructions [4]. This sort of probing could not be
implemented in userspace so it would not be possible to check for RDT
feature support in userspace without false negatives on some hardware
configurations.
It looks like libcontainer reads the CPU feature flags as a kind of
optimization so that it can skip checking whether the kernel supports an
RDT sub-feature if the hardware support is missing. As the kernel only
exposes subtrees in the `resctrl` filesystem for RDT sub-features with
hardware and kernel support, checking the CPU feature flags is redundant
from a correctness point of view. Remove the /proc/cpuinfo check as it
is an optimization which actually hurts performance.
[1]: https://unix.stackexchange.com/a/526679
[2]: https://lore.kernel.org/all/20220415161206.875029458@linutronix.de/
[3]: https://github.com/torvalds/linux/blob/7cf6a8a17f5b134b7e783c2d45c53298faef82a7/arch/x86/kernel/cpu/resctrl/core.c#L834-L851
[4]: https://github.com/torvalds/linux/blob/a6b450573b912316ad36262bfc70e7c3870c56d1/arch/x86/kernel/cpu/resctrl/core.c#L111-L153
Signed-off-by: Cory Snider <csnider@mirantis.com>
Currently only amd64 and arm64v8 tarball have been checked in testdata,
while busybox bundle is downloaded on fly, and supports multiple architectures.
To enable integration tests for more architectures, the hello world
bundle is replaced by busybox one.
Signed-off-by: Shengjing Zhu <zhsj@debian.org>
This is a forward-port of commit 91fa032da4 ("ci: add basic checks for
CHANGELOG.md"), plus whatever changes were made in release-1.1 branch
(up to v1.1.3).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When starting a new container, and the very last step of executing of a
user process fails (last lines of (*linuxStandardInit).Init), it is too
late to print a proper error since both the log pipe and the init pipe
are closed.
This is partially mitigated by using exec.LookPath() which is supposed
to say whether we will be able to execute or not. Alas, it fails to do
so when the binary to be executed resides on a filesystem mounted with
noexec flag.
A workaround would be to use access(2) with X_OK flag. Alas, it is not
working when runc itself is a setuid (or setgid) binary. In this case,
faccessat2(2) with AT_EACCESS can be used, but it is only available
since Linux v5.8.
So, use faccessat2(2) with AT_EACCESS if available. If not, fall back to
access(2) for non-setuid runc, and do nothing for setuid runc (as there
is nothing we can do). Note that this check if in addition to whatever
exec.LookPath does.
Fixes https://github.com/opencontainers/runc/issues/3520
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A new version of staticcheck (included into golangci-lint 1.46.2) gives
this new warning:
> libcontainer/factory_linux.go:230:59: SA9008: e refers to the result of a failed type assertion and is a zero value, not the value that was being type-asserted (staticcheck)
> err = fmt.Errorf("panic from initialization: %v, %s", e, debug.Stack())
> ^
> libcontainer/factory_linux.go:226:7: SA9008(related information): this is the variable being read (staticcheck)
> if e, ok := e.(error); ok {
> ^
Apparently, this is indeed a bug. Fix by using a different name for a
new variable, so we can access the old one under "else".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Due to a bug in commit 9c444070ec, when the user and mount namespaces
are used, and the bind mount is followed by the cgroup mount in the
spec, the cgroup is mounted using the bind mount's mount fd.
This can be reproduced with podman 4.1 (when configured to use runc):
$ podman run --uidmap 0:100:10000 quay.io/libpod/testimage:20210610 mount
Error: /home/kir/git/runc/runc: runc create failed: unable to start container process: error during container init: error mounting "cgroup" to rootfs at "/sys/fs/cgroup": mount /proc/self/fd/11:/sys/fs/cgroup/systemd (via /proc/self/fd/12), flags: 0x20502f: operation not permitted: OCI permission denied
or manually with the spec mounts containing something like this:
{
"destination": "/etc/resolv.conf",
"type": "bind",
"source": "/userdata/resolv.conf",
"options": [
"bind"
]
},
{
"destination": "/sys/fs/cgroup",
"type": "cgroup",
"source": "cgroup",
"options": [
"rprivate",
"nosuid",
"noexec",
"nodev",
"relatime",
"ro"
]
}
The issue was not found earlier since it requires using userns, and even then
mount fd is ignored by mountToRootfs, except for bind mounts, and all the bind
mounts have mountfd set, except for the case of cgroup v1's /sys/fs/cgroup
which is internally transformed into a bunch of bind mounts.
This is a minimal fix for the issue, suitable for backporting.
A test case is added which reproduces the issue without the fix applied.
Fixes: 9c444070ec ("Open bind mount sources from the host userns")
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
the struct blockIODevice is used in an exported struct but it is not itself exported rendering that type inaccessible to
outside projects
Signed-off-by: cdoern <cdoern@redhat.com>
systemd emits very loud warnings when the path specified doesn't exist
(which can be the case for some of our default rules). We don't need the
ruleset we give systemd to be completely accurate (we discard some kinds
of wildcard rules anyway) so we can safely skip adding these.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Remove upper bound in integer sanity check
to not restrict the number of socket-activated
sockets passed in.
Closes#3488
Signed-off-by: Erik Sjölund <erik.sjolund@gmail.com>
This removes the runc dependency on cpuguy83/md2man and
russross/blackfriday, which saves more than 400 KB (more than 300 KB
once stripped) from the binary.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Use env directive instead of adding to $GITHUB_ENV.
2. Use bash herefile to feed sha256sum instead of pipe to grep.
3. Fix the hardcoded checksum (it was missing the first character).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add checking of downloaded tarball checksum.
In case it doesn't match the hardcoded value, the error is like this:
libseccomp-2.5.4.tar.gz: FAILED
sha256sum: WARNING: 1 computed checksum did NOT match
In case the checksum for a particular version is not specified in the
script, the error will look like this:
./script/seccomp.sh: line 29: SECCOMP_SHA256[${ver}]: unbound variable
In case the the hardcoded value in the file is of wrong format/length,
we'll get:
sha256sum: 'standard input': no properly formatted SHA256 checksum lines found
In any of these cases, the script aborts (due to set -e).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This job is failing with "No space left on device" lately, and this
helps to fix it.
Besides, it seems that caching does not help to shorten execution times
(validate/release job succeeds in under 8 minutes now; ymmv).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Define sizeof(int) as a constant, and also return ENOSYS earlier in the
filter if it doesn't increase the number of instructions we generate
(this is a negligible performance improvement but it does make it easier
to understand the generated filter stub).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
On s390x, syscalls above 255 are multiplexed using the (now otherwise
unused) setup(2) syscall (syscall number 0). If the kernel supports the
syscall then it will correctly translate the syscall number such that
seccomp will correctly detect it -- however, for unknown syscalls the
syscall number remains unchanged. This can be verified by running the
following program under strace:
int main(void)
{
scmp_filter_ctx ctx = seccomp_init(SCMP_ACT_TRAP);
seccomp_load(ctx);
return syscall(439, AT_FDCWD, "asdf", X_OK, 0);
}
Which will then die with the following signal (on pre-5.8 kernels):
--- SIGSYS {si_signo=SIGSYS, si_code=SYS_SECCOMP,
si_call_addr=0x3ffb3006c22, si_syscall=__NR_setup,
si_arch=AUDIT_ARCH_S390X} ---
(Note that the si_syscall is __NR_setup, not __NR_faccessat2.)
As a result, the -ENOSYS handling we had previously did not work
completely correctly on s390x because any syscall not supported by the
kernel would be treated as syscall number 0 rather than the actual
syscall number.
Always returning -ENOSYS will not cause any issues because in all of the
cases where this multiplexing occurs, seccomp will see the remapped
syscall number -- and no userspace program will call setup(2)
intentionally (the syscall has not existed in Linux for decades and was
originally a hack used early in Linux init prior to spawning pid1 -- so
you will get -ENOSYS from the kernel anyway).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These are only used from inside the package, and we don't want them to
be public.
The only two methods left are Enable and Disable.
While at it, fix or suppress found lint-extra warnings.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This commit separates the functionality of setting cgroup device
rules out of libct/cgroups to libct/cgroups/devices package. This
package, if imported, sets the function variables in libct/cgroups and
libct/cgroups/systemd, so that a cgroup manager can use those to manage
devices. If those function variables are nil (when libct/cgroups/devices
are not imported), a cgroup manager returns the ErrDevicesUnsupported
in case any device rules are set in Resources.
It also consolidates the code from libct/cgroups/ebpf and
libct/cgroups/ebpf/devicefilter into libct/cgroups/devices.
Moved some tests in libct/cg/sd that require device management to
libct/sd/devices.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This moves the functionality related to devices, SkipDevices, and
SkipFreezeOnSet to a separate file, in preparation for the next commit.
No code changes.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It doesn't matter whether static or dynamic linking is used, runc
always needs libcontainer/nsenter, which is written in C and thus
requires cgo. Same is true for libcontainer/integration.
In addition, contrib/pkg/seccompagent also needs cgo (if seccomp build
tag is set), as it need to be linked against libseccomp C library.
By default, cgo is disabled when cross-compiling, meaning that
CGO_ENABLED=1 has to be set explicitly in such cases.
In all other cases (e.g. other contrib binaries) we do not need cgo.
Remove CGO_ENABLED=1 from GO_BUILD_STATIC (as it does not have anything
to do with static linking), and add it to all targets that require it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Set to empty value by default.
2. Assume Linux (remove GOOS check, since we do not support other OSes).
3. Instead of using a "not-supported" list, use a "supported" list
(as Go release notes usually say which platforms are supported).
As of today, -buildmode=pie is supported for:
* linux/386, linux/amd64, linux/arm, linux/arm64, and linux/ppc64le
(since Go 1.6, see https://tip.golang.org/doc/go1.6#compiler)
* linux/s390x (since Go 1.7, which adds the initial port)
* linux/riscv64 (since Go 1.16, see
https://tip.golang.org/doc/go1.16#riscv)
NOTE this does not mean we support these architectures; it is merely
a way to see if -buildmode=pie can be used.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
LDFLAGS_COMMON are used from two places, so it makes sense to dedup.
LDFLAGS_STATIC is a preparation for the next commit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
All we need is gcc, libc-dev, and binutils. In addition to that,
crossbuild-essential installs g++, libstdc++-dev, and a bunch of perl
packages and libraries which we do not need.
This should speed up image building, as well as make it smaller.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Dockerfile used to install libseccomp-dev packages for different
architectures. This is no longer true since commit f30244ee1b, which
changed to cross-compiling libseccomp (so we can get a static library
to link against).
Thus, adding extra architectures is no longer needed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We do not use all the files from scripts, only seccomp.sh and lib.sh.
This prevents unneeded rebuild of the image if e.g.
scripts/release_build.sh has changed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Do not set inheritable capabilities in runc spec, runc exec --cap,
and in libcontainer integration tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
With the updated git in Fedora 35, we can't build it via sudo:
ssh default 'sudo -i make -C /vagrant localunittest'
make: Entering directory '/vagrant'
fatal: unsafe repository ('/vagrant' is owned by someone else)
To add an exception for this directory, call:
git config --global --add safe.directory /vagrant
go build -trimpath "-buildmode=pie" -tags "seccomp" -ldflags "-X main.gitCommit= -X main.version=1.1.0+dev " -o runc .
error obtaining VCS status: exit status 128
Use -buildvcs=false to disable VCS stamping.
make: Leaving directory '/vagrant'
This commit should fix this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Looking through git blame, this was added by commit 9fac18329
aka "Initial commit of runc binary", most probably by mistake.
Obviously, a container should not have access to tun/tap device, unless
it is explicitly specified in configuration.
Now, removing this might create a compatibility issue, but I see no
other choice.
Aside from the obvious misconfiguration, this should also fix the
annoying
> Apr 26 03:46:56 foo.bar systemd[1]: Couldn't stat device /dev/char/10:200: No such file or directory
messages from systemd on every container start, when runc uses systemd
cgroup driver, and the system runs an old (< v240) version of systemd
(the message was presumably eliminated by [1]).
[1] https://github.com/systemd/systemd/pull/10996/commits/d5aecba6e0b7c73657c4cf544ce57289115098e7
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no need to parallelize lint and lint-extra jobs,
and they only differ with the arguments to golangci-lint.
Given that the longest time spent in these jobs is installing
libseccomp-dev, and that the second linter run can probably
benefit a lot from caching, it makes sense to merge them.
Move lint-extra from a separate job to a step in lint job.
The implementation is motivated by [1] and relies on the fact
that the last commit being fetched is the merge commit. So,
we need to set fetch-depth to 2 to be able to see the diff of
the merge commit -- and this is what golangci-lint is using.
[1] https://github.com/golangci/golangci-lint-action/issues/449#issuecomment-1096995821
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since the recent bump of actions/setup-go to v3 (commit
9d2268b9db), specifying "stable:" is no longer needed
when we want to try a beta or rc version of Go.
Remove it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The CAP_CHECKPOINT_RESTORE linux capability provides the ability to
update /proc/sys/kernel/ns_last_pid. However, because this file is under
/proc, and by default both K8s and CRI-O specify that /proc/sys should
be mounted as Read-Only, by default even with the capability specified,
a process will not be able to write to ns_last_pid.
To get around this, a pod author can specify a volume mount and a
hostpath to bind-mount /proc/sys/kernel/ns_last_pid. However, runc does
not allow specifying mounts under /proc.
This commit adds /proc/sys/kernel/ns_last_pid to the validProcMounts
string array to enable a pod author to mount ns_last_pid as read-write.
The default remains unchanged; unless explicitly requested as a volume
mount, ns_last_pid will remain read-only regardless of whether or not
CAP_CHECKPOINT_RESTORE is specified.
Signed-off-by: Irwin D'Souza <dsouzai.gh@gmail.com>
Now the only remaining file that needs shellcheck warnings to be fixed
is bash-completion. Note that in Makefile's TODO.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is done to make sure the script is working correctly in different
environments (distro and kernel versions). In addition, we can see in
test logs which kernel features are enabled.
Note that I didn't want to have a separate job for GHA CI, so I just
added this to the end of shellcheck one.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
One particularly bad one is ${codes[@]} which is fine in bash 4.4+,
but gives "codes[@]: unbound variable" with older bash versions,
such as with bash 4.2 used on CentOS 6. It's good that this is the only
array in the script that can potentially be empty.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
... and add this file to shellcheck target in Makefile.
These:
In script/check-config.sh line 27:
kernelMinor="${kernelVersion#$kernelMajor.}"
^----------^ SC2295 (info): Expansions inside ${..} need to be quoted separately, otherwise they match as patterns.
Did you mean:
kernelMinor="${kernelVersion#"$kernelMajor".}"
In script/check-config.sh line 103:
source /etc/os-release 2>/dev/null || /bin/true
^-------------^ SC1091 (info): Not following: /etc/os-release was not specified as input (see shellcheck -x).
In script/check-config.sh line 267:
NET_CLS_CGROUP $netprio
^------^ SC2206 (warning): Quote to prevent word splitting/globbing, or split robustly with mapfile or read -a.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Like this one:
In ./script/check-config.sh line 215:
if [ "$kernelMajor" -lt 5 ] || [ "$kernelMajor" -eq 5 -a "$kernelMinor" -le 1 ]; then
^-- SC2166 (warning): Prefer [ p ] && [ q ] as [ p -a q ] is not well defined.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Allow wrap_bad and wrap_good to have an optional arguments.
2. Remove unneeded echos; this fixes the shellcheck warnings like
In ./script/check-config.sh line 178:
echo "$(wrap_bad 'cgroup hierarchy' 'nonexistent??')"
^-- SC2005 (style): Useless echo? Instead of 'echo $(cmd)', just use 'cmd'.
3. Fix missing color argument in calls to wrap_color (when printing the
hint about how to install apparmor).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case statfs("/sys/fs/cgroup/unified") fails with any error other
than ENOENT, current code panics. As IsCgroup2HybridMode is called from
libcontainer/cgroups/fs's init function, this means that any user of
libcontainer may panic during initialization, which is ugly.
Avoid panicking; instead, do not enable hybrid hierarchy support and
report the error (under debug level, not to confuse anyone).
Basically, replace the panic with "turn off hybrid mode support"
(which makes total sense since we were unable to statfs its root).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The variable $ROOTLESS, as set by helpers.bash and used in many places,
provides the same value as $EUID which is always set by bash. Since we
are using bash, we can rely on $EUID being omnipresent.
Modify all uses accordingly, and since the value is known to be a
number, omit the quoting.
Similarly, replace all uses of $(id -u) to $EUID.
Do some trivial cleanups along the way, such as
- simplify some if A; then B; to A && B;
- do not use [[ instead of [ where not necessary.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test requires both rootless and root, which does not make sense.
Remove the rootless part.
Fixes: d41a273da
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This makes it work similar to all the other variables we use as binary
flags.
The new 'shellcheck disable' is due to a bug in shellcheck (basically,
it does not track the scope of variables or execution order, assuming
everything is executed as soon as it is seen).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Strictly speaking, == is for [[ only, not for [ / test,
and, unlike =, the right side is a pattern.
To avoid confusion, use =. In cases where we compare with empty string,
use -z instead.
Keep using [[ in some cases since it does not require quoting the left
and right side of comparison (I trust shellcheck on that one).
This should have no effect (other than the code being a tad more
strict).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The only implementation of these is linuxContainer. It does not make
sense to have an interface with a single implementation, and we do not
foresee other types of containers being added to runc.
Remove BaseContainer and Container interfaces, moving their methods
documentation to linuxContainer.
Rename linuxContainer to Container.
Adopt users from using interface to using struct.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since LinuxFactory has become the means to specify containers state
top directory (aka --root), and is only used by two methods (Create
and Load), it is easier to pass root to them directly.
Modify all the users and the docs accordingly.
While at it, fix Create and Load docs (those that were originally moved
from the Factory interface docs).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
StartInitialization does not have to be a method of Factory (while
it is clear why it was done that way initially, now we only have
Linux containers so it does not make sense).
Fix callers and docs accordingly.
No change in functionality.
Also, since this was the only user of libcontainer.New with the empty
string as an argument, the corresponding check can now be removed
from it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The only implementation is LinuxFactory, let's use this directly.
Move the piece of documentation about Create from removed factory.go to
the factory_linux.go.
The LinuxFactory is to be removed later.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Require go 1.17 from now on, since go 1.16 is no longer supported.
Drop go1.16 compatibility.
NOTE we also have to install go 1.18 from Vagrantfile, because
Fedora 35 comes with Go 1.16.x which can't be used.
Note the changes to go.mod and vendor are due to
https://go.dev/doc/go1.17#tools
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Function strings.Title is deprecated as of Go 1.18, because it does not
handle some corner cases good enough. In this case, though, it is
perfectly fine to use it since we have a single ASCII word as an
argument, and strings.Title won't be removed until at least Go 2.0.
Suppress the deprecation warning.
The alternative is to not capitalize the namespace string; this will break
restoring of a container checkpointed by earlier version of runc.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
What used to be godoc.org is now pkg.go.dev, and while the old URLs
still work, they might be broken in the future.
Updated badges are generated via https://pkg.go.dev/badge/
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Fix function docs. In particular, remove the part
which is not true ("verifies that the user isn't trying to set up any
mounts they don't have the rights to do"), and fix the part that
says "that doesn't resolve to root" (which is no longer true since
commit d8b669400a).
2. Replace fmt.Sscanf (which is slow and does lots of allocations)
with strings.TrimPrefix and strconv.Atoi.
3. Add a benchmark for rootlessEUIDMount. Comparing the old and the new
implementations:
name old time/op new time/op delta
RootlessEUIDMount-4 1.01µs ± 2% 0.16µs ± 1% -84.15% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
RootlessEUIDMount-4 224B ± 0% 80B ± 0% -64.29% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
RootlessEUIDMount-4 7.00 ± 0% 1.00 ± 0% -85.71% (p=0.008 n=5+5)
Note this code is already tested (in rootless_test.go).
Fixes: d8b669400a
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
While doing the previous fix, I went over all the tests in this file and
made sure they were named correctly. This patch just adds a small
sentence to clarify the intent, and does some minor improvements to some
other test names.
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
There was a typo and instead of "empty" we should have used "non-empty".
Let's add a small sentence explaining the intent (like other tests in
this file) and let's highlight what we expect to happen in this test (to
ignore the listenerPath).
Fixes: #3415
Signed-off-by: Rodrigo Campos <rodrigoca@microsoft.com>
The following failure was observed in CI (on centos-stream-8 in
integration-cgroup suite):
not ok 42 runc delete
(from function `fail' in file tests/integration/helpers.bash, line 338,
in test file tests/integration/delete.bats, line 30)
`[ "$output" = "" ] || fail "cgroup not cleaned up correctly: $output"' failed
....
cgroup not cleaned up correctly: /sys/fs/cgroup/pids/system.slice/tmp-bats\x2drun\x2d68012-runc.IPOypI-state-testbusyboxdelete-runc.zriC8C.mount
/sys/fs/cgroup/cpu,cpuacct/system.slice/tmp-bats\x2drun\x2d68012-runc.IPOypI-state-testbusyboxdelete-runc.zriC8C.mount
...
Apparently, this is a cgroup systemd creates for a mount unit which
appears then runc does internal /proc/self/exe bind-mount. The test
case should not take it into account.
The second problem with this test is it does not check that cgroup
actually exists when the container is running (so checking that it
was removed after makes less sense). For example, in rootless mode
the cgroup might not have been created.
Fix the find arguments to look for a specific cgroup name, and add
a check that these arguments are correct (i.e. the cgroup is found
when the container is running).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
golangci-lint-action v3 no longer installs golang itself, and the
version that comes with Ubuntu is not new/good enough.
Install go 1.17.x explicitly.
Introduce GO_VERSION environment variable to avoid duplication,
and use it instead of 1.x in other places, so that implicit go update
won't bring some unexpected failures.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently, "systemctl --user --no-pager show-environment" is useless
without DBUS_SESSION_BUS_ADDRESS or XDG_RUNTIME_DIR set:
$ echo $DBUS_SESSION_BUS_ADDRESS, $XDG_RUNTIME_DIR
unix:path=/run/user/1000/bus, /run/user/1000
$ systemctl --user --no-pager show-environment | grep DBUS_SESS
DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/1000/bus
$ unset DBUS_SESSION_BUS_ADDRESS
$ systemctl --user --no-pager show-environment | grep DBUS_SESS
DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/1000/bus
$ unset XDG_RUNTIME_DIR
$ systemctl --user --no-pager show-environment | grep DBUS_SESS
Failed to connect to bus: $DBUS_SESSION_BUS_ADDRESS and $XDG_RUNTIME_DIR not defined (consider using --machine=<user>@.host --user to connect to bus of other user)
So, it does not make sense to try it to get the address.
Also, it does not make sense to suggest "systemctl --user start dbus"
either, for the same reason, so remove that suggestion from the error
message text.
Since DBUS_SESSION_BUS_ADDRESS environment variable, on which the code
relies, is et by dbus-run-session (or dbus-launch, or something similar
that is supposed to be run during the login process), add a suggestion
to re-login.
Finally, fix the following linter warning:
> error-strings: error strings should not be capitalized or end with punctuation or a newline (revive)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a way to prevent the code doing something really bad when a
variable it uses is not set. Good to have since it helps to catch some
logical errors etc.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Audit all checks for non-empty variables (i.e. ' -z ', ' -n ',
' != ""' and '= ""'), and fix those cases where a variable might be
unset. Those variables (that might not be set) are
- RUNC_USE_SYSTEMD
- BATS_RUN_TMPDIR
- AUX_UID
- AUX_DIR
- SD_PARENT_NAME
- REL_PARENT_PATH
- ROOT
- HAVE_CRIU
- ROOTLESS_FEATURES
- and a few test-specific or file-specific variables
This should allow us to enable set -u.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Add "unset ALT_ROOT" since it should not be used after teardown is
called.
2. Remove "rm -rf $ALT_ROOT". It is not needed, because ALT_ROOT is a
subdirectory of ROOT, which is removed in teardown_bundle.
3. Checking for ALT_ROOT being non-empty is a leftover from the era when
teardown() was called as the first step from setup(). Since commit
41670e21f0 this is no longer the case, so the condition
is no longer needed (plus, the `set -u` which is about to be added
should catch any possible use of unset ALT_ROOT).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Some tests (those in help.bats and version.bats) do not use setup_bundle
(as they do not need to start any containers), and thus they do not set
$ROOT. As a consequence, these tests now call "runc --root /state" which
is not nice.
Make adding --root conditional (only if $ROOT is set).
Amazingly, this change breaks help.bats tests under rootless, because
"sudo rootless" does not change the value of XDG_RUNTIME_DIR which still
points to root-owned directory, and as a result we have this:
> runc foo -h (status=1):
> the path in $XDG_RUNTIME_DIR must be writable by the user
> time="2022-02-08T07:04:57Z" level=error msg="mkdir /run/user/0/runc: permission denied"
This could be fixed by adding proper $ROOT, but it's easier just to skip
those tests under non-root.
NOTE that version.bats is not broken because -v is handled by urfave/cli
very early, so app.Before function is not run.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Older bash versions treats variable as unset if nothing has been
assigned to it. Here is an example from CentOS 7 system:
[kir@localhost ~]$ bash -u -c 'x() { local args=(); echo "${args[@]}"; }; x'
bash: args[@]: unbound variable
[kir@localhost ~]$ echo $BASH_VERSION
4.2.46(2)-release
Rewrite to work around this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently, not all files listed in /sys/kernel/cgroup/delegate must
exist in every cgroup, so we should ignore ENOENT.
Dot not ignore ENOENT on the directory itself though.
Change cgroupFilesToChown to not return ".", and refactor it to not do
any dynamic slice appending in case we're using the default built-in
list of files.
Fixes: 35d20c4e0
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It is questionable whether runc list should return an empty list of
containers when non-existent --root is specified or not.
The current behavior is the directory is always created and then the
empty list of container is shown.
To my mind, specifying a non-existent root is an error and should be
reported as such. This is what this patch does.
For backward compatibility, if --root is not set (i.e. a default is
used), ENOENT is not reported.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. In case --root option is not provided, do nothing.
2. Instead of checking if root value is empty string, check it after
filepath.Abs, and reject "/". Improve docstring while at it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Variable xdgRuntimeDir is only checked to be non-empty. Change it to
a boolean.
2. Refactor so that os.Getenv is not called twice.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is a mix of styles when handling CLI commands. In most cases we
return an error, which is handled from app.Run in main.go (it calls
fatal if there is an error).
In a few cases, though, we call fatal(err) from random places.
Let's be consistent and always return an error. The only exception is
runc exec, which needs to exit with a particular exit code.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The value of root is already an absolute path since commit
ede8a86ec1, so it does not make sense to call filepath.Abs()
again.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Don't require CAT or MBA because we don't detect those correctly (we
don't support L2 or L3DATA/L3CODE for example, and in the future
possibly even more). With plain "ClosId mode" we don't really care: we
assign the container to a pre-configured CLOS without trying to do
anything smarter.
Moreover, this was a duplicate/redundant check anyway, as for CAT and
MBA there is another specific sanity check that is done if L3 or MB
is specified in the config.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
1. This valid warning is reported by shellcheck v0.8.0:
In tests/integration/helpers.bash line 38:
KERNEL_MINOR="${KERNEL_VERSION#$KERNEL_MAJOR.}"
^-----------^ SC2295 (info): Expansions inside ${..} need to be quoted separately, otherwise they match as patterns.
Did you mean:
KERNEL_MINOR="${KERNEL_VERSION#"$KERNEL_MAJOR".}"
Fix this.
2. These (invalid) warnings are also reported by the new version:
In tests/integration/events.bats line 13:
@test "events --stats" {
^-- SC2030 (info): Modification of status is local (to subshell caused by @bats test).
In tests/integration/events.bats line 41:
[ "$status" -eq 0 ]
^-----^ SC2031 (info): status was modified in a subshell. That change might be lost.
Basically, this is happening because shellcheck do not really track
the call tree and/or local variables. This is a known (and reported)
deficiency, and the alternative to disabling these warnings is moving
the code around, which is worse due to more changes in git history.
So we have to silence/disable these.
3. Update shellcheck to 0.8.0.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Foreground runc exec and runc run forwards all the signals (that it can)
to the process being run.
Since Go 1.14, go runtime uses SIGURG for async preemptive scheduling.
This means that runc regularly receives SIGURG and, in case of
foreground runc run/exec, it gets forwarded to the container process.
For example:
[kir@kir-rhat runc]$ sudo ./runc --debug exec xx67 sleep 1m
...
DEBU[0000] child process in init()
DEBU[0000] setns_init: about to exec
DEBU[0000]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
DEBU[0000]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
DEBU[0000]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
...
Or, with slightly better debug messages from commit 58c1ff39a5:
DEBU[0000]signals.go:102 main.(*signalHandler).forward() forwarding SIGURG to 819784
DEBU[0000]signals.go:102 main.(*signalHandler).forward() forwarding SIGURG to 819784
Obviously, this signal is an internal implementation detail of Go
runtime, and should not be forwarded to the container process.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Remove intelrtd.Manager interface, since we only have a single
implementation, and do not expect another one.
Rename intelRdtManager to Manager, and modify its users accordingly.
Remove NewIntelRdtManager from factory.
Remove IntelRdtfs. Instead, make intelrdt.NewManager return nil if the
feature is not available.
Remove TestFactoryNewIntelRdt as it is now identical to TestFactoryNew.
Add internal function newManager to be used for tests (to make sure
some testing is done even when the feature is not available in
kernel/hardware).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
TestGetContainerStats test a function that is smaller than the test
itself, and only calls a couple of other functions (which are
represented by mocks). It does not make sense to have it.
mockIntelRdtManager is only needed for TestGetContainerStats
and TestGetContainerState, which basically tests that Path
is called. Also, it does not make much sense to have it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since we are looking up the path to newuidmap/newgidmap in one context,
and executing those in another (libct/nsenter), it might make sense to
use a stricter rules for looking up path to those binaries.
Practically it means that if someone wants to use custom newuidmap and
newgidmap binaries from $PATH, it would be impossible to use these from
the current directory by means of PATH=.:$PATH; instead one would have
to do something like PATH=$(pwd):$PATH.
See https://go.dev/blog/path-security for background.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These were introduced in commit d8b669400 back in 2017, with a TODO
of "make binary names configurable". Apparently, everyone is happy with
the hardcoded names. In fact, they *are* configurable (by prepending the
PATH with a directory containing own version of newuidmap/newgidmap).
Now, these binaries are only needed in a few specific cases (when
rootless is set etc.), so let's look them up only when needed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We only have one implementation of config validator, which is always
used. It makes no sense to have Validator interface.
Having validate.Validator field in Factory does not make sense for all
the same reasons.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Those are *always* /proc/self/exe init, and it does not make sense
to ever change these. More to say, if InitArgs option func (removed
by this commit) is used to change these parameters, it will break
things, since "init" is hardcoded elsewhere.
Remove this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 029b73c1b replaced a regular expression with code checking the
characters. Despite what the comment said about ASCII, the check was
performed rune by rune, not byte by byte.
Note the check was still correct, basically comparing int32's, but the
byte by byte way is a tad faster and more straightforward. The change
also fixes the issue of a misleading comment.
Benchmark before/after shows a modest improvement:
name old time/op new time/op delta
CheckPropertyName-4 164ns ± 2% 123ns ± 2% -24.73% (p=0.029 n=4+4)
name old alloc/op new alloc/op delta
CheckPropertyName-4 96.0B ± 0% 64.0B ± 0% -33.33% (p=0.029 n=4+4)
name old allocs/op new allocs/op delta
CheckPropertyName-4 6.00 ± 0% 4.00 ± 0% -33.33% (p=0.029 n=4+4)
Fixes: 029b73c1b
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 643f8a2b40 renamed isValidName to checkPropertyName, but fell
short of renaming its test and benchmark. Fix that.
Fixes: 643f8a2b40
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Here's how it looks now:
$ runc --debug exec ctid sleep 1h
...
DEBU[0000]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
DEBU[0000]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
DEBU[0022]signals.go:102 main.(*signalHandler).forward() sending signal to process terminated
DEBU[0022]signals.go:102 main.(*signalHandler).forward() sending signal to process urgent I/O condition
This is obviously not very readable.
Use unix.SignalName, plus a numeric representation of the signal, since
SignalName does not know all signals.
Add PID while we're at it.
With this commit:
DEBU[0000]signals.go:103 main.(*signalHandler).forward() forwarding signal 23 (SIGURG) to 891345
DEBU[0020]signals.go:103 main.(*signalHandler).forward() forwarding signal 45 () to 891345
DEBU[0020]signals.go:103 main.(*signalHandler).forward() forwarding signal 23 (SIGURG) to 891345
DEBU[0020]signals.go:103 main.(*signalHandler).forward() forwarding signal 23 (SIGURG) to 891345
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
After adding some debug info to cloned_binary.c I found out that
is_self_cloned() is not working right when runc binary is on tmpfs,
resulting in one extra re-exec of runc.
With some added debug:
$ mkdir bin
$ sudo mount -t tmpfs tmp bin
$ sudo cp runc bin
$ sudo ./bin/runc --debug exec xxx true
DEBU[0000] nsexec[763590]: => is_self_cloned
DEBU[0000] nsexec[763590]: got seals 1 (want 15)
DEBU[0000] nsexec[763590]: <= is_self_cloned, is_cloned = 0
DEBU[0000] nsexec[763590]: try_bindfd: 5
DEBU[0000] nsexec[763590]: re-exec itself...
DEBU[0000] nsexec[763590]: => is_self_cloned
DEBU[0000] nsexec[763590]: got seals 1 (want 15)
DEBU[0000] nsexec[763590]: <= is_self_cloned, is_cloned = 0
DEBU[0000] nsexec[763590]: try_bindfd: -1
DEBU[0000] nsexec[763590]: fallback to make_execfd: 5
DEBU[0000] nsexec[763590]: re-exec itself...
DEBU[0000] nsexec[763590]: => is_self_cloned
DEBU[0000] nsexec[763590]: got seals 15 (want 15)
DEBU[0000] nsexec[763590]: <= is_self_cloned, is_cloned = 1
From the above, it is seen that
- `is_self_cloned` returns 0,
- `try_bindfd` is called and succeeds,
- runc re-execs itself,
- the second call to `is_self_cloned` returns 0 again (because GET_SEALS returns 1),
- runc falls back to `make_execfd`, and re-execs again,
- finally, the third `is_self_cloned` returns 1.
I guess that the code relied on the following (quoting fcntl(2)):
> Currently, file seals can be applied only to a file descriptor
> returned by memfd_create(2) (if the MFD_ALLOW_SEALING was employed).
> On other filesystems, all fcntl() operations that operate on seals
> will return EINVAL.
It looks like in case of a file on tmpfs it returns 1 (F_SEAL_SEAL).
With the fix:
DEBU[0000] nsexec[768367]: => is_self_cloned
DEBU[0000] nsexec[768367]: got seals 1 (want 15)
DEBU[0000] nsexec[768367]: no CLONED_BINARY_ENV
DEBU[0000] nsexec[768367]: <= is_self_cloned, is_cloned = 0
DEBU[0000] nsexec[768367]: try_bindfd: 5
DEBU[0000] nsexec[768367]: re-exec itself...
DEBU[0000] nsexec[768367]: => is_self_cloned
DEBU[0000] nsexec[768367]: got seals 1 (want 15)
DEBU[0000] nsexec[768367]: fstatfs says ro = 1
DEBU[0000] nsexec[768367]: fstat says nlink = 1
DEBU[0000] nsexec[768367]: <= is_self_cloned, is_cloned = 1
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was introduced in an initial commit, back in the day when criu was
a highly experimental thing. Today it's not; most users who need it have
it packaged by their distro vendor.
The usual way to run a binary is to look it up in directories listed in
$PATH. This is flexible enough and allows for multiple scenarios (custom
binaries, extra binaries, etc.). This is the way criu should be run.
Make --criu a hidden option (thus removing it from help). Remove the
option from man pages, integration tests, etc. Remove all traces of
CriuPath from data structures.
Add a warning that --criu is ignored and will be removed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is needed since the future commits will touch this code, and then
the lint-extra CI job complains.
> libcontainer/factory.go#L245
> var-naming: var fdsJson should be fdsJSON (revive)
and
> libcontainer/init_linux.go#L181
> error-strings: error strings should not be capitalized or end with punctuation or a newline (revive)
and
> notify_socket.go#L94
> receiver-naming: receiver name n should be consistent with previous receiver name s for notifySocket (revive)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since the next commit is going to touch this structure, our CI
(lint-extra) is about to complain about improperly named field:
> Warning: var-naming: struct field ContainerId should be ContainerID (revive)
Make it happy.
Brought to use by gopls rename.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This method was removed earlier by commit 097c6d7425,
but the documentation was not updated. Fix it.
Fixes: 097c6d7425
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Those were added by commit 59c5c3ac0 back in Apr 2015, but AFAICS were
never used and are obsoleted by more generic container hooks (initially
added by commit 05567f2c94 in Sep 2015).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The utils.Annotations was used here before only because it made it
possible to distinguish between "key not found" and "empty value" cases.
With the previous commit, utils.SearchLabels can do that, and so it
makes sense to use it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using strings.Split generates temporary strings for GC to collect.
Rewrite the function to not do that.
Also, add a second return value, so that the caller can distinguish
between an empty value found and no key found cases.
Fix the test accordingly.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit 551629417 we can (and should) use Info() to get access to
file stat. Do this.
While going over directory entries, a parallel runc delete can remove
an entry, and with the current code it results in a fatal error (which
was not observed in practice, but looks quite possible). To fix,
add a special case to continue on ErrNotExist.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case of a read-only /dev, it's better to move on and let whatever is
run in a container to handle any possible errors.
This solves runc exec for a user with read-only /dev.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since we already called fstat, we know the current file uid. In case it
is the same as the one we want it to be, there's no point in trying
chown.
Remove the specific /dev/null check, as the above also covers it
(comparing /dev/null uid with itself is true).
This also fixes runc exec with read-only /dev for root user.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Use os/file Chown method instead of bare unix.Fchown as it already have
access to underlying fd, and produces nice-looking errors. This allows
us to remove our error wrapping and some linter annotations.
We still use unix.Fstat since os.Stat access to os-specific fields
like uid/gid is not very straightforward. The only change here is to use
file name (rather than fd) in the error text.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For the Nth time I wanted to replace parsing mountinfo with
statfs and the check for superblock magic, but it is not possible
since some code relies of mount options check which can only
be obtained via mountinfo.
Add a note about it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 4d1d6185ab added this
nsenter_unsupported.go file in order for nsenter to be a valid (but
empty, non-functional) Go package on unsupported platforms.
As a result, runc can be build successfully without CGO, which results
in a non-working and hard-to-debug binary (see issue 3330).
As the functionality of being able to compile a package which is
definitely not working is questionable, and I can't think of any use
cases, let's remove the file.
With this, runc can no longer be build without CGO:
[kir@kir-rhat runc]$ CGO_ENABLED=0 make runc
go build -trimpath "-buildmode=pie" -tags "seccomp" -ldflags "-X main.gitCommit=v1.0.0-452-g00f56786-dirty -X main.version=1.1.0-rc.1+dev " -o runc .
go build github.com/opencontainers/runc/libcontainer/nsenter: build constraints exclude all Go files in /home/kir/go/src/github.com/opencontainers/runc/libcontainer/nsenter
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
(on Go 1.18 this is actually an error)
> libcontainer/factory_linux.go:341:10: fmt.Errorf format %w has arg e of wrong type interface{}
Unfortunately, fixing it results in an errorlint warning:
> libcontainer/factory_linux.go#L344 non-wrapping format verb for fmt.Errorf. Use `%w` to format errors (errorlint)
so we have to silence that one.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit 12e99a0f8d we do require Go >= 1.16, so this file
is no longer needed.
Also, this actually ensures that go >= 1.16 is used (otherwise
libcontainer/cgroups/getallpids.go won't compile).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This will make releases much simpler. I've back-filled the changelog
with everything since runc 1.0.0 (there's not much point going further
back than that).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Commit 7cfb107f2c started using unix.Geteuid(), unix.Getegid() as
uid/gid argument for chown. It seems that it should have removed chown
entirely, since, according to mkdir(2),
> The newly created directory will be owned by the effective user ID of
> the process. If the directory containing the file has the
> set-group-ID bit set, or if the filesystem is mounted with BSD
> group semantics (mount -o bsdgroups or, synonymously mount -o grpid),
> the new directory will inherit the group ownership from its parent;
> otherwise it will be > owned by the effective group ID of the process.
So, the only effect of the chown after mkdir is ignoring the sgid bit
on the parent directory (which is probably not the right thing to do).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This check was always broken, and it slipped through the cracks because
we never run it without additional architectures now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It was released about a month ago. I don't see anything major
in the changelog but it makes sense to keep tracking upstream deps.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using "$@" instead of $1 in update_config() allows us to use it from
hooks.bats, where jq is used with more options than usual.
We need to disable SC2016 as otherwise shellcheck sees $something inside
single quotes and think we are losing the shell expansion (we are not).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This can be used to specify a different runc binary, for example:
sudo -E RUNC=$PWD/runc.mine tests/integration/cwd.bats
A different (but compatible enough) runtime also works:
sudo -E RUNC=/usr/local/bin/crun tests/integration/cwd.bats
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The new mount option "rro" makes the mount point recursively read-only,
by calling `mount_setattr(2)` with `MOUNT_ATTR_RDONLY` and `AT_RECURSIVE`.
https://man7.org/linux/man-pages/man2/mount_setattr.2.html
Requires kernel >= 5.12.
The "rro" option string conforms to the proposal in util-linux/util-linux Issue 1501.
Fix issue 2823
Similary, this commit also adds the following mount options:
- rrw
- r[no]{suid,dev,exec,relatime,atime,strictatime,diratime,symfollow}
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
My GPG keys are not available inside the container, so it makes little
sense to try to sign the binaries inside the container's release.sh. The
solution is to split things into separate build and sign stages, with
signing ocurring after the in-Docker build.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In a (quite common) case RDT is not supported by the kernel/hardware,
it does not make sense to parse /proc/cpuinfo and /proc/self/mountinfo,
and yet the current code does it (on every runc exec, for example).
Fortunately, there is a quick way to check whether RDT is available --
if so, kernel creates /sys/fs/resctrl directory. Check its existence,
and skip all the other initialization if it's not present.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This test was written back in the day when findIntelRdtMountpointDir
was using its own mountinfo parser. Commit f1c1fdf911 changed that,
and thus this test is actually testing moby/sys/mountinfo parser, which
does not make much sense.
Remove the test, and drop the io.Reader argument since we no longer need
to parse a custom file.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case resctrl filesystem can not be found in /proc/self/mountinfo
(which is pretty common on non-server or non-x86 hardware), subsequent
calls to Root() will result in parsing it again and again.
Use sync.Once to avoid it. Make unit tests call it so that Root() won't
actually parse mountinfo in tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When writing netlink messages, it is possible to have a byte array
larger than UINT16_MAX which would result in the length field
overflowing and allowing user-controlled data to be parsed as control
characters (such as creating custom mount points, changing which set of
namespaces to allow, and so on).
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Since commit 7296dc1712, type intelRdtData is only used by tests,
and since commit 79d292b9f, its only member is config.
Change the test to use config directly, and remove the type.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This adds a new GHA CI job which runs a few extra linters. This is only
done for pull requests, and should only warn about new code.
The justification is simple: we want more linters, but since this is not
a new project, adding a new linter meaning we have to fix all the
existing warnings. In some cases having all the warnings fixed is
difficult and takes time, plus it is usually a low priority task.
Therefore, we are stuck with inability to add new linters because we
can't fix all their warnings. Meanwhile, new pull requests add more
code which is not linted.
This is an attempt to break this vicious cycle. Let's enable godot
and revive for now and see how it is going.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It never returns any error, so let's drop it (in case it needs to be
re-added, it is easy to do so).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Those were never used (ctx was added by the initial commit, and
error was added by commit 25fd4a6757).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This function never returned anything other than nil, and its return
value is discarded since it is only called from defer.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since commit 8850636eb3 (February 2015) this function is no longer
used (replaced by (*ConfigValidator).rootfs), so let's remove it,
together with its unit tests (which were added by commit 917c1f6d6 in
April 2016).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Delegating cgroups to the container enables more complex workloads,
including systemd-based workloads. The OCI runtime-spec was
recently updated to explicitly admit such delegation, through
specification of cgroup ownership semantics:
https://github.com/opencontainers/runtime-spec/pull/1123
Pursuant to the updated OCI runtime-spec, change the ownership of
the container's cgroup directory and particular files therein, when
using cgroups v2 and when the cgroupfs is to be mounted read/write.
As a result of this change, systemd workloads can run in isolated
user namespaces on OpenShift when the sandbox's cgroupfs is mounted
read/write.
It might be possible to implement this feature in other cgroup
managers, but that work is deferred.
Signed-off-by: Fraser Tweedale <ftweedal@redhat.com>
This is not used since commit 5e7b48f7c0 (23 Mar 2017).
In case there are external users, they should switch to
opencontainers/selinux.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Move test case comments to doc strings, and use t.Run.
Suggested-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no need to run hooks when `Config.Hooks` is just an empty map,
(dlv) p p.config.Config.Hooks
github.com/opencontainers/runc/libcontainer/configs.Hooks []
Signed-off-by: Dave Chen <dave.chen@arm.com>
1. Instead of distinguishing between errors and warnings, let's treat all
errors as warnings, thus simplifying the code. This changes the
function behaviour for input like hugepages-BadNumberKb --
previously, the error from Atoi("BadNumber") was considered fatal,
now it's just another warnings.
2. Move the warning logging to HugePageSizes, thus simplifying the test
case, which no longer needs to read what logrus writes. Note that we
do not want to log all the warnings (as chances are very low we'll
get any, and if we do this means the code need to be updated), only
the first one.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
I have noticed that libct/cg/fs allocates 8K during init on every runc
execution:
> init github.com/opencontainers/runc/libcontainer/cgroups/fs @1.5 ms, 0.028 ms clock, 8512 bytes, 13 allocs
Apparently this is caused by global HugePageSizes variable init, which
is only used from GetStats (i.e. it is never used by runc itself).
Remove it, and use HugePageSizes() directly instead. Make it init-once,
so that GetStats (which, I guess, is periodically called by kubernetes)
does not re-read huge page sizes over and over.
This also removes 12 allocs and 8K from libct/cg/fs init section:
> $ time GODEBUG=inittrace=1 ./runc --help 2>&1 | grep cgroups/fs
> init github.com/opencontainers/runc/libcontainer/cgroups/fs @1.5 ms, 0.003 ms clock, 16 bytes, 1 allocs
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Since GetHugePageSize do not have any external users (checked by
sourcegraph), and no internal user ever uses its second return value
(the error), let's drop it.
2. Rename GetHugePageSize -> HugePageSizes (drop the Get prefix as per
Go guidelines, add suffix since we return many sizes).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using null bytes as control characters for sending strings via netlink
opens us up to a user explicitly putting a null byte in a mount string
(which JSON will happily let you do) and then causing us to open a mount
path different to the one expected.
In practice this is more of an issue in an environment such as
Kubernetes where you may have path-based access control policies (which
are more susceptible to these kinds of flaws).
Found by Google Project Zero.
Fixes: 9c444070ec ("Open bind mount sources from the host userns")
Reported-by: Felix Wilhelm <fwilhelm@google.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
As reported in [1], in a case where read-only fuse (sshfs) mount
is used as a volume without specifying ro flag, the kernel fails
to remount it (when adding various flags such as nosuid and nodev),
returning EPERM.
Here's the relevant strace line:
> [pid 333966] mount("/tmp/bats-run-PRVfWc/runc.RbNv8g/bundle/mnt", "/proc/self/fd/7", 0xc0001e9164, MS_NOSUID|MS_NODEV|MS_REMOUNT|MS_BIND|MS_REC, NULL) = -1 EPERM (Operation not permitted)
I was not able to reproduce it with other read-only mounts as the source
(tried tmpfs, read-only bind mount, and an ext2 mount), so somehow this
might be specific to fuse.
The fix is to check whether the source has RDONLY flag, and retry the
remount with this flag added.
A test case (which was kind of hard to write) is added, and it fails
without the fix. Note that rootless user need to be able to ssh to
rootless@localhost in order to sshfs to work -- amend setup scripts
to make it work, and skip the test if the setup is not working.
[1] https://github.com/containers/podman/issues/12205
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The staticcheck linter points out that the err != nil comparison
after system.Exec is always true:
> libcontainer/standard_init_linux.go#L253
> SA4023: this comparison is always true (staticcheck)
> libcontainer/system/linux.go#L43
> SA4023(related information): github.com/opencontainers/runc/libcontainer/system.Exec never returns a nil interface value (staticcheck)
Indeed, Exec either returns an error or does not return at all.
Remove the (useless) check.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Decapitalize errors.
2. Rename isValidName to checkPropertyName.
3. Make it return a specific error.
Suggested-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit fb4c27c4b7 (went into v1.0.0-rc93) fixed a bug with
read-only tmpfs, but introduced a bug with read-only /dev.
This happens because /dev is a tmpfs mount and is therefore remounted
read-only a bit earlier than before.
To fix,
1. Revert the part of the above commit which remounts all tmpfs mounts
as read-only in mountToRootfs.
2. Reuse finalizeRootfs (which is already used to remount /dev
read-only) to also remount all ro tmpfs mounts that were previously
mounted rw in mountPropagate.
3. Remove the break in finalizeRootfs, as now we have more than one
mount to care about.
4. Reorder the if statements in finalizeRootfs to perform the fast check
(for ro flag) first, and compare the strings second. Since /dev is
most probably also a tmpfs mount, do the m.Device check first.
Add a test case to validate the fix and prevent future regressions;
make sure it fails before the fix:
✗ runc run [ro /dev mount]
(in test file tests/integration/mounts.bats, line 45)
`[ "$status" -eq 0 ]' failed
runc spec (status=0):
runc run test_busybox (status=1):
time="2021-11-12T12:19:48-08:00" level=error msg="runc run failed: unable to start container process: error during container init: error mounting \"devpts\" to rootfs at \"/dev/pts\": mkdir /tmp/bats-run-VJXQk7/runc.0Fj70w/bundle/rootfs/dev/pts: read-only file system"
Fixes: fb4c27c4b7
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Also, add a simple test and a benchmark (just out of sheer curiosity).
Benchmark results:
name old time/op new time/op delta
IsValidName-4 540ns ± 3% 45ns ± 1% -91.76% (p=0.008 n=5+5)
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 1cd71dfd7 added isSecSuffix, but the same thing can be done
easily without a regex. This is faster and saves some init time and
memory.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
parseMountOption already returns way too many values, making the code
kind of hard to read.
Since all of the return values are used as is to populate the fields of
configs.Mount, let's change it to return (semi-)populated *configs.Mount
instead.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This makes the repeated calls to parseMountOptions faster,
and decreases the amount of garbage to collect.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These two maps are the same, except that mountPropagationMapping
has an extra element with key of "" and value of 0. Since the
code already checks for f != 0, this extra element is not a problem.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Eliminate some of these allocations when starting runc:
> init github.com/opencontainers/runc/libcontainer/specconv @10 ms, 0.11 ms clock, 5408 bytes, 70 allocs
Most of this (4K) is the two regexes, which are left intact for now.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When I tried to start a rootless container under a different/wrong user,
I got:
$ ../runc/runc --systemd-cgroup --root /tmp/runc.$$ run 445
ERRO[0000] runc run failed: operation not permitted
This is obviously not good enough. With this commit, the error is:
ERRO[0000] runc run failed: fchown fd 9: operation not permitted
Alas, there are still some code that returns unwrapped errnos from
various unix calls.
This is a followup to commit d8ba4128b2 which wrapped many, but not
all, bare unix errors. Do wrap some more, using either os.PathError or
os.SyscallError.
While at it,
- use os.SyscallError instead of os.NewSyscallError;
- use errors.Is(err, os.ErrXxx) instead of os.IsXxx(err).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
A bug in systemd-249.6-2.fc35.x86_64 prevents rootless containers from
start when systemd manager is used.
Apparently, "config exclude" is not working in F35 dnf shell either, so
use a workaround of specifying --exclude from the command line.
This should fix runc CI for the time being.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Sometimes a container cgroup already exists but is frozen.
When this happens, runc init hangs, and it's not clear what is going on.
Refuse to run in a frozen cgroup; add a test case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently runc allows multiple containers to share the same cgroup (for
example, by having the same cgroupPath in config.json). While such
shared configuration might be OK, there are some issues:
- When each container has its own resource limits, the order of
containers start determines whose limits will be effectively applied.
- When one of containers is paused, all others are paused, too.
- When a container is paused, any attempt to do runc create/run/exec
end up with runc init stuck inside a frozen cgroup.
- When a systemd cgroup manager is used, this becomes even worse -- such
as, stop (or even failed start) of any container results in
"stopTransientUnit" command being sent to systemd, and so (depending on
unit properties) other containers can receive SIGTERM, be killed after a
timeout etc.
Any of the above may lead to various hard-to-debug situations in production
(runc init stuck, cgroup removal error, wrong resource limits, init not
reaping zombies etc.).
One obvious solution is to refuse a non-empty cgroup when starting a new
container. This would be a breaking change though, so let's make it in
steps, with the first step is issue a warning and a deprecated notice
about a non-empty cgroup.
Later (in runc 1.2) we will replace this warning with an error.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently, if a container is paused (i.e. its cgroup is frozen), runc exec
just hangs, and it is not obvious why.
Refuse to exec in a paused container. Add a test case.
In case runc exec in a paused container is a legit use case,
add --ignore-paused option to override the check. Document it,
add a test case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Also rename `Vagrantfile.fedora%d` to `Vagrantfile.fedora` so that
we do not need to reset the commit log on upgrading the Fedora release.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Some test directories are created using mktemp -d, and so they have
permissions set to 0700 and are thus inaccessible to a user inside
userns. This was workarounded for $ROOT in userns.bats before.
Now, when we have updated Cirrus CI config to use Fedora 35 (rather than
34), userns tests fail:
> runc run failed: unable to start container process: error during
> container init: error preparing rootfs: mount
> /tmp/bats-run-4pCERd/runc.f66gCC/bundle/rootfs:/tmp/bats-run-4pCERd/runc.f66gCC/bundle/rootfs,
> flags: 0x5000: permission denied
Fedora 34 image used kernel v5.11, while Fedora 35 has v5.15.
Apparently, the newer kernel also checks that the parent directories
are accessible by the user before doing mount.
Move the old workaround from userns.bats to helpers.bats, drop the r bit
(not needed), and add $BATS_RUN_TMPDIR (also created by mktemp -d) to
fix userns.bats test failures under Fedora 35.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The misc cgroup controller, introduced in Linux 5.13, is still unknown
to systemd, and thus it cannot delegate it. Add an appropriate fixup
to the test case, similar to an earlier commit 601cf5825f.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Runtime spec says:
> sysctl (object, OPTIONAL) allows kernel parameters to be modified at
> runtime for the container. For more information, see the sysctl(8)
> man page.
and sysctl(8) says:
> variable
> The name of a key to read from. An example is
> kernel.ostype. The '/' separator is also accepted in place of a '.'.
Apparently, runc config validator do not support sysctls with / as a
separator. Fortunately this is a one-line fix.
Add some more test data where / is used as a separator.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add a unit test to check that bind mounts that have a part of its
path non accessible by others still work when using user namespaces.
To do this, we also modify newRoot() to return rootfs directories that
can be traverse by others, so the rootfs created works for all test
(either running in a userns or not).
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
Go 1.15 is not supported since Go 1.17 release (16 Aug 2021), and some
packages that we use already require Go 1.16+ (notably,
github.com/cilium/ebpf v0.7.0).
Let's require Go 1.16+.
Remove Go version requirement from README when describing dependencies,
since it is no longer needed:
$ GO=go1.15.15 make vendor
go1.15.15 mod tidy
go mod tidy: go.mod file indicates go 1.16, but maximum supported version is 1.15
make: *** [Makefile:141: vendor] Error 1
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Jobs verify/compile-buildtags and verify/deps relied on whatever Go
version is available from the Ubuntu-20.04 image, which seems to be
1.15.x).
Job test/cross-i386 was installing whatever Go version is considered to
be the default one by actions/setup-go@v2, which seems to be go 1.15.15
at the moment.
Fix all three jobs to install Go 1.x (which should translate to latest
stable Go version, i.e. 1.17.2 as of now).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The previous commit fixed an issue opening bind mount sources. This
commit just adds integration tests to make sure we don't regress on this
in the future.
Signed-off-by: Alban Crequy <alban@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
The source of the bind mount might not be accessible in a different user
namespace because a component of the source path might not be traversed
under the users and groups mapped inside the user namespace. This caused
errors such as the following:
# time="2020-06-22T13:48:26Z" level=error msg="container_linux.go:367:
starting container process caused: process_linux.go:459:
container init caused: rootfs_linux.go:58:
mounting \"/tmp/busyboxtest/source-inaccessible/dir\"
to rootfs at \"/tmp/inaccessible\" caused:
stat /tmp/busyboxtest/source-inaccessible/dir: permission denied"
To solve this problem, this patch performs the following:
1. in nsexec.c, it opens the source path in the host userns (so we have
the right permissions to open it) but in the container mntns (so the
kernel cross mntns mount check let us mount it later:
https://github.com/torvalds/linux/blob/v5.8/fs/namespace.c#L2312).
2. in nsexec.c, it passes the file descriptors of the source to the
child process with SCM_RIGHTS.
3. In runc-init in Golang, it finishes the mounts while inside the
userns even without access to the some components of the source
paths.
Passing the fds with SCM_RIGHTS is necessary because once the child
process is in the container mntns, it is already in the container userns
so it cannot temporarily join the host mntns.
This patch uses the existing mechanism with _LIBCONTAINER_* environment
variables to pass the file descriptors from runc to runc init.
This patch uses the existing mechanism with the Netlink-style bootstrap
to pass information about the list of source mounts to nsexec.c.
Rootless containers don't use this bind mount sources fdpassing
mechanism because we can't setns() to the target mntns in a rootless
container (we don't have the privileges when we are in the host userns).
This patch takes care of using O_CLOEXEC on mount fds, and close them
early.
Fixes: #2484.
Signed-off-by: Alban Crequy <alban@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
Export getIntelRdtRoot function as Root.
This is needed by google/cadvisor, which is (ab)using GetIntelRdtPath,
removed by commit 7296dc1712.
While at it, do some minimal refactoring to always use Root()
internally, not relying on variable value. Other than that it's just
some renaming.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If the container binary to be run is removed in between runc create
and runc start, the latter spits the following error:
> can't exec user process: no such file or directory
This is a bit confusing since we don't see what file is missing.
Wrap the unix.Exec error into os.PathError, like in many other cases,
to provide some context. Remove the error wrapping from
(*linuxStandardInit).Init as it is now redundant.
With this patch, the error is now:
> exec /bin/false: no such file or directory
Reported-by: Daniel J Walsh <dwalsh@redhat.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case hugetlb is not supported, GetStats() should not error out,
and yet it does.
Assume that if GetHugePageSize return an error, hugetlb is
not supported (this is what cgroup v1 manager do).
Fixes: 89a87adb
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Users would like to have the possibility to skip checks for their
tests the same way they are skipped within the tests in runc.
Not exposing this variable makes it very hard to test components
that use this library. To avoid copying-and-pasting the code
into outside projects this variable sould be exposed to the users.
Signed-off-by: Itamar Holder <iholder@redhat.com>
Commit f30244ee1b broke the scenario of using Dockefile for
anything but making a release. This happened because it installed
native libseccomp build to a temporary directory, and so linking against
libseccomp required setting a few environment variables.
Let's fix this, and simplify libseccomp installation. Instead of using
temporary directories, let's install native libseccomp to a specified
directory, all the cross-builds to its subdirectories, and set
PKG_CONFIG_PATH and LD_LIBRARY_PATH in Dockerfile so that the built
library will found by pkg-config and the dynamic linker (without setting
LD_LIBRARY_PATH, ld picks up distro-provided libseccomp.so).
While at it, fix some bugs introduced by the abovementioned commit.
This fixes building runc in make targets like shell, dbuild,
integration, unittest -- i.e. those that depend on runcimage.
Fixes: f30244ee1b
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently, we can create subcgroup in a rootless container with systemd cgroupv2 on centos8.
But after the container exited, the container cgroup and its subcgroup will not be removed.
Fix this by removing all directories recursively.
Fixes: https://github.com/opencontainers/runc/issues/3225
Signed-off-by: Kang Chen <kongchen28@gmail.com>
Commit 7a0302f0d7 already removed "runc init" from runc help output,
as this is an internal option not supposed to be used by the end user.
Let's remove runc init completion, too.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In some setups, multiple cgroups are used inside a container,
and sometime there is a need to execute a process in a particular
sub-cgroup (in case of cgroup v1, for a particular controller).
This is what this commit implements.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Check that runc run and runc exec put the process on the same cgroups v2
when using hybrid mode.
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently the parent process of the container is moved to the right
cgroup v2 tree when systemd is using a hybrid model (last line with 0::):
$ runc --systemd-cgroup run myid
/ # cat /proc/self/cgroup
12:cpuset:/system.slice/runc-myid.scope
11:blkio:/system.slice/runc-myid.scope
10:devices:/system.slice/runc-myid.scope
9:hugetlb:/system.slice/runc-myid.scope
8:memory:/system.slice/runc-myid.scope
7:rdma:/
6:perf_event:/system.slice/runc-myid.scope
5:net_cls,net_prio:/system.slice/runc-myid.scope
4:freezer:/system.slice/runc-myid.scope
3:pids:/system.slice/runc-myid.scope
2:cpu,cpuacct:/system.slice/runc-myid.scope
1:name=systemd:/system.slice/runc-myid.scope
0::/system.slice/runc-myid.scope
However, if a second process is executed in the same container, it is
not moved to the right cgroup v2 tree:
$ runc exec myid /bin/sh -c 'cat /proc/self/cgroup'
12:cpuset:/system.slice/runc-myid.scope
11:blkio:/system.slice/runc-myid.scope
10:devices:/system.slice/runc-myid.scope
9:hugetlb:/system.slice/runc-myid.scope
8:memory:/system.slice/runc-myid.scope
7:rdma:/
6:perf_event:/system.slice/runc-myid.scope
5:net_cls,net_prio:/system.slice/runc-myid.scope
4:freezer:/system.slice/runc-myid.scope
3:pids:/system.slice/runc-myid.scope
2:cpu,cpuacct:/system.slice/runc-myid.scope
1:name=systemd:/system.slice/runc-myid.scope
0::/user.slice/user-1000.slice/session-8.scope
This commit makes that processes executed with exec are placed into the
right cgroup v2 tree. The implementation checks if systemd is using a
hybrid mode (by checking if cgroups v2 is mounted in
/sys/fs/cgroup/unified), if yes, the path of the cgroup v2 slice for
this container is saved into the cgroup path list.
The fs group driver has a similar issue, in this case none of the runc
run or runc exec commands put the process in the right cgroups v2. This
commit also fixes that.
Having the processes of the container in its own cgroup v2 is useful
for any BPF programs that rely on bpf_get_current_cgroup_id(), like
https://github.com/kinvolk/inspektor-gadget/ for instance.
[@kolyshkin: rebased]
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The function used here, cgroups.EnterPid, silently skips non-existing
paths, and it does not look like a good idea to do so for an existing
container with already configured cgroups.
Switch to cgroups.WriteCgroupProc which does not do that, so in case
a cgroup does not exist, we'll get an error.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
No need to add a file name to the error messages, as errors from
OpenFile and (*os.File).Write both contain the file name already.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kir Kolyshkin (5):
ci/gha: test criu-dev with latest go
ci/gha: remove debug info
CI/GHA: switch to OBS criu repo
Dockerfile: fix apt-key warning
Dockerfile: use Debian_11 repo for criu
LGTMs: mrunalp cyphar
As commits 120f74060 and a58718013 were added independently,
criu-dev go version was left at 1.16.x. Fix this.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes
> Warning: apt-key is deprecated. Manage keyring files in trusted.gpg.d instead (see apt-key(8)).
Apparently, "the internets" disagree with the above, instead suggesting
using /usr/share/keyrings and a signed-by= declaration in sources.list.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The Debian_11 was not available in this repo at the time when commit 24d318b8b
was made, so we had to use Debian_10 URL for Debian 11 (apparently without any
consequences).
Now Debian_11 is available, so let's switch to it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It is assumed that m.config is not nil, so these checks are redundant
(in case it is nil, NewManager panics and this code is unreachable).
Note that cgroups/manager.New checks that config is not nil.
Remove them.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Cgroup controllers should never panic, and yet sometimes they do.
Add a unit test to check that controllers never panic when called with
nil arguments and/or resources, and fix a few found cases.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
cgName and cgParent are only used when cgPath is empty, so move
their cleaning to the body of the appropriate "if" statement.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This way we
- won't re-initialize the paths if they were provided;
- will always have paths ready for every method.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is already documented but I guess more explanations (in particular,
why the path is being removed from paths) won't hurt.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Separate path initialization logic from Apply to initPaths,
and call initPaths from NewManager, so:
- we can error out early (in NewManager rather than Apply);
- always have m.paths available (e.g. in Destroy or Exists).
- do not unnecessarily call subsysPath from Apply in case
the paths were already provided.
2. Add a check for non-nil cgroups.Resources to NewManager,
since initPaths, as well as some controller's Apply methods,
need it.
3. Move cgroups.Resources.Unified check from Apply to NewManager,
so we can error out early (same check exists in Set).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Make Rootless and Systemd flags part of config.Cgroups.
2. Make all cgroup managers (not just fs2) return error (so it can do
more initialization -- added by the following commits).
3. Replace complicated cgroup manager instantiation in factory_linux
by a single (and simple) libcontainer/cgroups/manager.New() function.
4. getUnifiedPath is simplified to check that only a single path is
supplied (rather than checking that other paths, if supplied,
are the same).
[v2: can't -> cannot]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This implements cross-build for "make release", moving the build into a
container. This way we can support arm, arm64, ppc, and whatnot.
* script/seccomp.sh: separate out of script/release.sh, amend to support
cross-compile and save needed environment variables to a file.
* Dockerfile: add installing libseccomp from source, as this is needed
for release builds.
* script/release.sh: amend to support more architectures in addition to
the native build. Additional arches can be added by specifying
"-a <arch>" argument (can be specified multiple times), or
"make RELEASE_ARGS="-a arm64" release" if called via make.
All supported architectures can be enabled via "make releaseall".
* Makefile: move "release" target to "localrelease", add "release" and
"releaseall" targets to build via the Dockerfile. This is done because
most distros (including Fedora and openSUSE) lack cross-glibc, which is
needed to cross-compile libseccomp.
* Makefile: remove 'cross' and 'localcross' targets, as this is now done
by the release script.
* .github/workflows/validate.yum: amend the release CI job to cross-build
for supported architectures, remove cross job.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There is no need to have a static version of recvtty and/or sd-helper
binary.
This speeds up script/release.sh a bit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. The seccompagent target it built in the same way as others in contrib,
so there is no need to have a separate rule.
2. Mark seccompagent as phony, because it is (it rarely happens, but I
actually just had an issue because this was absent).
3. Add seccompagent binary to clean target.
Fixes: e21a9ee81
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Wire through CRIU's support to change the mount context on restore.
This is especially useful if restoring a container in a different pod.
Single container restore uses the same SELinux process label and
same mount context as during checkpointing. If a container is being
restored into an existing pod the process label and the mount context
needs to be changed to the context of the pod.
Changing process label on restore is already supported by runc. This
patch adds the possibility to change the mount context.
Signed-off-by: Adrian Reber <areber@redhat.com>
runc delete -f is not working for a paused container, since in cgroup v1
SIGKILL does nothing if a process is frozen (unlike cgroup v2, in which
you can kill a frozen process with a fatal signal).
Theoretically, we only need this for v1, but doing it for v2 as well is
OK.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Dismantle and remove struct cgroupData. It contained three unrelated
entities (cgroup paths, pid, and resources), and made the code
harder to read. Most importantly, though, it is not needed.
Now, subsystems' Apply methods take path, resources, and pid.
To a reviewer -- the core of the changes is in fs.go and paths.go,
the rest of it is adapting to the new signatures and related test
changes.
2. Dismantle and remove struct cgroupTestUtil. This is a followup
to the previous item -- since cgroupData is gone, there is nothing
to hold in cgroupTestUtil. The change itself is very small (see
util_test.go), but this patch is big because of it -- mostly
because we had to replace helper.cgroup.Resources with
&config.Resources{}.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case c.Path is set, c.Name and c.Parent are not used, and so
calls to utils.CleanPath are entirely unnecessary. Move them to
inside of the "if" statement body.
Get rid of the intermediate cgPath variable, it is not needed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now fs.go is not very readable as its public API functions are
intermixed with internal stuff about getting cgroup paths.
Move that out to paths.go, without changing any code.
Same for the tests -- move paths-related tests to paths_test.go.
This commit is separate to make the review easier.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As ExpandSlice("system.slice") returns "/system.slice", there is no need
to call it for such paths (and the slash will be added by path.Join
anyway).
The same optimization was already done for v2 as part of commit
bf15cc99b1.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
We were checking if a unit is a slice two times. Consolidate those
checks, and improve comments while we're at it.
The code is the same in v1 and v2 but it's too complicated to factor it
out, thus we just do the same changes in v1.go and v2.go.
No functional change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As the error may contain anything, it may not be clear to a user that
the whole (create or run) operation failed. Amend the errors.
Also, change the code flow in create to match that of run, so we don't
have to add the fake "return nil" at the end.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
All three callers* of startContainer call revisePidFile and createSpec
before calling it, so it makes sense to move those calls to inside of
the startContainer, and drop the spec argument.
* -- in fact restore does not call revisePidFile, but it should.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Error messages should not usually contain newlines.
Testing shows that the error runc delete prints is the same before and
after this commit:
[kir@kir-rhat runc-tst]$ sudo ../runc/runc delete xx3
ERRO[0000] cannot delete container xx3 that is not stopped: running
[kir@kir-rhat runc-tst]$
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This commit adds the config.json as generated by the script. Note that
the diff is minimal if you see this commit with "git show -w". The
differences are mostly whitespaces and some ordering.
We add a simple test that runs this and expects sucess.
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Currently, if the log level is not set to e.g. "debug", runc init sends
some debug logs to the parent, which parses and discards it.
It is better to not send those in the first place.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The code already parses an environment variable into an integer twice,
and we're about to add a third one.
Factor it out to getenv_int().
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This makes it possible to use bail() even if logging is not set up
(yet), so we don't have to think whether it's OK to use it or not.
In addition, this might help some unit tests that do not set log
forwarding.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of passing _LIBCONTAINER_LOGLEVEL as a string
(like "debug" or "info"), use a numeric value.
Also, simplify the init log level passing code -- since we actually use
the same level as the runc binary, just get it from logrus.
This is a preparation for the next commit.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Two new seccomp actions have been added to the libseccomp-golang
dependency, which can be now supported by runc, too.
ActKillThread kills the thread that violated the rule. It is the same as
ActKill. All other threads from the same thread group will continue to
execute.
ActKillProcess kills the process that violated the rule. All threads in
the thread group are also terminated. This action is only usable when
libseccomp API level 3 or higher is supported.
Signed-off-by: Sascha Grunert <sgrunert@redhat.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
These are just boilerplate and are only really useful for the two
actions which require us to set a default errno/aux value (ActErrno and
ActTrace).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Kir Kolyshkin (6):
runc --debug: shorter caller info
libct/logs: do not show caller in nsexec logs
libct/logs: parse log level implicitly
libct/logs: test: make more robust
libct/logs: remove ConfigureLogging
init.go, main.go: don't use logs.ConfigureLogging
LGTMs: thaJeztah cyphar
Kir Kolyshkin (3):
libct/system: add I and P process states
libct/system.Stat: fix/improve/speedup
libct/system/proc_test: fix, improve, add benchmark
LGTMs: thaJeztah cyphar
Add functional test to check seccomp notify end-to-end. This test uses the
sample seccomp agent from the contrib/cmd folder.
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
Implement sample seccomp agent. It's also used in integration tests in
the following commit.
Instructions how to use it in contrib/cmd/seccompagent/README.md
Signed-off-by: Alban Crequy <alban@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
Extend the SetupSeccomp tests by adding the following cases:
- Test nil config
- Test empty config
- Test bad action and architecture
- Test all possible actions
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
This commit implements support for the SCMP_ACT_NOTIFY action. It
requires libseccomp-2.5.0 to work but runc still works with older
libseccomp if the seccomp policy does not use the SCMP_ACT_NOTIFY
action.
A new synchronization step between runc[INIT] and runc run is introduced
to pass the seccomp fd. runc run fetches the seccomp fd with pidfd_get
from the runc[INIT] process and sends it to the seccomp agent using
SCM_RIGHTS.
As suggested by @kolyshkin, we also make writeSync() a wrapper of
writeSyncWithFd() and wrap the error there. To avoid pointless errors,
we made some existing code paths just return the error instead of
re-wrapping it. If we don't do it, error will look like:
writing syncT <act>: writing syncT: <err>
By adjusting the code path, now they just look like this
writing syncT <act>: <err>
Signed-off-by: Alban Crequy <alban@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Co-authored-by: Rodrigo Campos <rodrigo@kinvolk.io>
SendFds is a helper function for sending a set of file descriptors and a message
over a unix domain socket.
Signed-off-by: Mauricio Vásquez <mauricio@kinvolk.io>
if function returns error before WriteJSON defer, error will not be
printed out, so move this defer as early as possible and use logrus to
print out error if returns before it.
Signed-off-by: xiadanni <xiadanni1@huawei.com>
openSUSE comes with site-config package, which makes configure select
${prefix}/lib64 as libdir on x86_64, unless explicitly specified.
Since release.sh relies on a particular libdir path (for pkgconfig), it
breaks things:
> + make -C /home/kir/git/runc PKG_CONFIG_PATH=/tmp/tmp.QgIJ1sR5c9/lib/pkgconfig COMMIT_NO= EXTRA_FLAGS=-a 'EXTRA_LDFLAGS=-w -s -buildid=' static
> make[1]: Entering directory '/home/kir/git/runc'
> CGO_ENABLED=1 go build -trimpath -a -tags "seccomp netgo osusergo" -ldflags "-extldflags -static -X main.gitCommit=v1.0.0-204-g963e0146 -X main.version=1.0.0+dev -w -s -buildid=" -o runc .
> Package libseccomp was not found in the pkg-config search path.
> Perhaps you should add the directory containing `libseccomp.pc'
To fix, we have to explicitly specify libdir.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 9f3d7534ea enabled logrus to show information about log
caller, if --debug is set.
The problem is, the file name and in many cases the function name have a
long prefix of github.com/opencontainers/runc (this is with -trimpath,
and without it it's worse).
Add a function to trim the prefix.
Note all this happens only when --debug is given.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 9f3d7534ea enabled logrus to show information about log
caller, if --debug is set. It is helpful in many scenarios, but does
not work very well when we are debugging runc init, for example:
# runc --debug run -d xx4557
DEBU[0000]libcontainer/logs/logs.go:45 github.com/opencontainers/runc/libcontainer/logs.processEntry() nsexec[279687]: logging set up
DEBU[0000]libcontainer/logs/logs.go:45 github.com/opencontainers/runc/libcontainer/logs.processEntry() nsexec[279687]: logging set up
DEBU[0000]libcontainer/logs/logs.go:45 github.com/opencontainers/runc/libcontainer/logs.processEntry() nsexec[279687]: => nsexec container setup
DEBU[0000]libcontainer/logs/logs.go:45 github.com/opencontainers/runc/libcontainer/logs.processEntry() nsexec[279687]: update /proc/self/oom_score_adj to '30'
As we're merely forwarding the logs here, printing out filename:line and
function is useless and clutters the logs a log.
To fix, create and use a copy of the standard logger with caller info
turned off.
With this in place, nsexec logs are sane again:
# runc --debug --log-format=text run -d xe34
DEBU[0000] nsexec[293595]: logging set up
DEBU[0000] nsexec[293595]: logging set up
DEBU[0000] nsexec[293595]: => nsexec container setup
DEBU[0000] nsexec[293595]: update /proc/self/oom_score_adj to '30'
This patch also changes Logf to Log in processEntry, as this is what it
should be.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
There's no need to call logrus.ParseLevel as logrus.Level already
implements encoding.TextUnmarshaler interface.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
When playing with the log forwarder, I broke it, but all the units tests
were still passing. This happened because test cases were merely looking
for a word (like "kitten") in the log output, which also happened to be
there in case of an error (as a part of an error message produced by log
forwarder).
Make the test a bit more robust by
- looking for a complete log message, not just part of it;
- also checking that log file does NOT contain errors.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
- add missing colons before error message;
- unify error messages after cmd.Start and cmd.Wait, so that they show
context and the error itself.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of reading a single message, do read all the logs from the init,
and use DisallowUnknownFields for stricter checking.
While at it, use reapChildren to reap zombies (and add an extra check).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The test was not working since at least commit 64bb59f592
renamed pid to stage2_pid (or maybe even earlier), so the pid
was never received (i.e. pid.Pid was 0).
The problem was not caught because os.FindProcess never return an error
on Unix.
Factor out and fix pid decode function:
- use DisallowUnknownInput to get error if JSON will be changed;
- check pids to make sure they are valid
- and use unix.Wait4 to reap zombies.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Make sure we close all file descriptors at the end of the test.
2. Make sure we close child fds after the start.
3. Use newPipe for logs as well, for simplicity and uniformity.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Treat warning as errors only in the CI. We can enforce it in the source
code (like setting CFLAGS in libcontainer/nsenter/nsenter.go), but that
can force other downstream to patch the code if thei C compiler produces
warnings. For that reason, we do it only on the CI.
Todays CGO warnings are quite hidden in the CI (only shown for the
compilation step, that is collapsed) and CI is green anyways. With this
patch, CI fails if a warning is introduced.
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
Go 1.17 introduce this new (and better) way to specify build tags.
For more info, see https://golang.org/design/draft-gobuild.
As a way to seamlessly switch from old to new build tags, gofmt (and
gopls) from go 1.17 adds the new tags along with the old ones.
Later, when go < 1.17 is no longer supported, the old build tags
can be removed.
Now, as I started to use latest gopls (v0.7.1), it adds these tags
while I edit. Rather than to randomly add new build tags, I guess
it is better to do it once for all files.
Mind that previous commits removed some tags that were useless,
so this one only touches packages that can at least be built
on non-linux.
Brought to you by
go1.17 fmt ./...
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These are not needed as these packages (libcontainer/cgroups,
libcontainer/cgroups/fs, and libcontainer/cgroups/systemd) can
not be built under non-linux anyway (for various reasons).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Only some libcontainer packages can be built on non-linux platforms
(not that it make sense, but at least go build succeeds). Let's call
these "good" packages.
For all other packages (i.e. ones that fail to build with GOOS other
than linux), it does not make sense to have linux build tag (as they
are broken already, and thus are not and can not be used on anything
other than Linux).
Remove linux build tag for all non-"good" packages.
This was mostly done by the following script, with just a few manual
fixes on top.
function list_good_pkgs() {
for pkg in $(find . -type d -print); do
GOOS=freebsd go build $pkg 2>/dev/null \
&& GOOS=solaris go build $pkg 2>/dev/null \
&& echo $pkg
done | sed -e 's|^./||' | tr '\n' '|' | sed -e 's/|$//'
}
function remove_tag() {
sed -i -e '\|^// +build linux$|d' $1
go fmt $1
}
SKIP="^("$(list_good_pkgs)")"
for f in $(git ls-files . | grep .go$); do
if echo $f | grep -qE "$SKIP"; then
echo skip $f
continue
fi
echo proc $f
remove_tag $f
done
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was added by commit 5aa82c950 back in the day when we thought
runc is going to be cross-platform. It's very clear now it's Linux-only
package.
While at it, further clarify it in README that we're Linux only.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For files that end with _linux.go or _linux_test.go, there is no need to
specify linux build tag, as it is assumed from the file name.
In addition, rename libcontainer/notify_linux_v2.go -> libcontainer/notify_v2_linux.go
for the file name to make sense.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add a test case for an issue fixed by the previous commit.
Unfortunately, this is somewhat complicated as there's no easy way to
create a transient unit, so a binary, sd-helper, had to be added. On top
of that, an ability to create a parent/pod cgroup is added to
helpers.bash, which might be useful for future integration tests.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As reported in issue 3084, sometimes setting CPU quota period fails
when a new period is lower and a parent cgroup has CPU quota limit set.
This happens as in cgroup v1 the quota and the period can not be set
together (this is fixed in v2), and since the period is being set first,
new_limit = old_quota/new_period may be higher than the parent cgroup
limit.
The fix is to retry setting the period after the quota, to cover all
possible scenarios.
Add a test case to cover a regression caused by an earlier version of
this patch (ignoring a failure of setting invalid period when quota is
not set).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This reverts commit 814f3ae1d9. This
changed the on-disk state which breaks runc when it has to operate on
containers started with an older runc version. Working around this is
far more complicated than just reverting it.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
ConfigError was added by commit e918d02139, while removing runc own
error system, to preserve a way for a libcontainer user to distinguish
between a configuration error and something else.
The way ConfigError is implemented requires a different type of check
(compared to all other errors defined by error.go). An attempt was made
to rectify this, but the resulting code became even more complicated.
As no one is using this functionality (of differentiating a "bad config"
type of error from other errors), let's just drop the ConfigError type.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
All the errors returned from Validate should tell about a configuration
error. Some were lacking a context, so add it.
While at it, fix abusing fmt.Errorf and logrus.Warnf where the argument
do not contain %-style formatting.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Looking into data generated by setting
GODEBUG="inittrace=1"
I have noticed this line:
init github.com/opencontainers/runc/libcontainer/cgroups/devices @1.2 ms, 0.020 ms clock, 10512 bytes, 133 allocs
This is the leader for both bytes and allocs among the packages from
this repo, and all of it is caused by a single regex:
> var devicesListRegexp = regexp.MustCompile(`^([abc])\s+(\d+|\*):(\d+|\*)\s+([rwm]+)$`)
It seems that the same parsing can be done without relying on
a regular expression, no decrease in readability, and 2x faster
(according to the benchmark added), and also makes runc start
slightly faster and leaner.
Before:
BenchmarkParseLine-4 176240 6768 ns/op 6576 B/op 64 allocs/op
After:
BenchmarkParseLine-4 322441 3535 ns/op 5520 B/op 53 allocs/op
[v2: single split with SplitFunc; fix a typo in error message]
[v3: rebase after 3159 merge; re-ran benchmarks (results are similar)]
Co-authored-by: Sebastiaan van Stijn <github@gone.nl>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
... and add the file to be checked by shellcheck.
The warnings fixed are:
In tests/integration/helpers.bash line 10:
INTEGRATION_ROOT=$(dirname "$(readlink -f "$BASH_SOURCE")")
^----------^ SC2128: Expanding an array without an index only gives the first element.
In tests/integration/helpers.bash line 22:
TESTDATA="${INTEGRATION_ROOT}/testdata"
^------^ SC2034: TESTDATA appears unused. Verify use (or export if used externally).
In tests/integration/helpers.bash line 42:
echo "runc $@ (status=$status):" >&2
^-- SC2145: Argument mixes string and array. Use * or separate argument.
^-----^ SC2154: status is referenced but not assigned.
In tests/integration/helpers.bash line 43:
echo "$output" >&2
^-----^ SC2154: output is referenced but not assigned.
In tests/integration/helpers.bash line 77:
| .linux.gidMappings += [{"hostID": '"$(($ROOTLESS_GIDMAP_START + 10))"', "containerID": 1, "size": 20}]
^--------------------^ SC2004: $/${} is unnecessary on arithmetic variables.
In tests/integration/helpers.bash line 78:
| .linux.gidMappings += [{"hostID": '"$(($ROOTLESS_GIDMAP_START + 100))"', "containerID": 1000, "size": '"$(($ROOTLESS_GIDMAP_LENGTH - 1000))"'}]'
^--------------------^ SC2004: $/${} is unnecessary on arithmetic variables.
^---------------------^ SC2004: $/${} is unnecessary on arithmetic variables.
In tests/integration/helpers.bash line 125:
base_path=$(gawk '$(NF-2) == "cgroup" && $NF ~ /\<'${g}'\>/ { print $5; exit }' /proc/self/mountinfo)
^--^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
base_path=$(gawk '$(NF-2) == "cgroup" && $NF ~ /\<'"${g}"'\>/ { print $5; exit }' /proc/self/mountinfo)
In tests/integration/helpers.bash line 127:
eval CGROUP_${g^^}_BASE_PATH="${base_path}"
^----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
eval CGROUP_"${g^^}"_BASE_PATH="${base_path}"
In tests/integration/helpers.bash line 229:
if [ "x$CGROUP_UNIFIED" = "xyes" ]; then
^----------------^ SC2268: Avoid x-prefix in comparisons as it no longer serves a purpose.
Did you mean:
if [ "$CGROUP_UNIFIED" = "yes" ]; then
In tests/integration/helpers.bash line 234:
eval cgroup=\$${var}${REL_CGROUPS_PATH}
^----^ SC2086: Double quote to prevent globbing and word splitting.
^-----------------^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
eval cgroup=\$"${var}""${REL_CGROUPS_PATH}"
In tests/integration/helpers.bash line 236:
cat $cgroup/$source
^-----^ SC2086: Double quote to prevent globbing and word splitting.
^-----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
cat "$cgroup"/"$source"
In tests/integration/helpers.bash line 242:
current="$(get_cgroup_value $1)"
^-- SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
current="$(get_cgroup_value "$1")"
In tests/integration/helpers.bash line 245:
echo "current" $current "!?" "$expected"
^------^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
echo "current" "$current" "!?" "$expected"
In tests/integration/helpers.bash line 257:
[ $(id -u) != "0" ] && user="--user"
^------^ SC2046: Quote this to prevent word splitting.
In tests/integration/helpers.bash line 259:
current=$(systemctl show $user --property $source $SD_UNIT_NAME | awk -F= '{print $2}')
^-----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
current=$(systemctl show $user --property "$source" $SD_UNIT_NAME | awk -F= '{print $2}')
In tests/integration/helpers.bash line 261:
[ "$current" = "$expected" ] || [ -n "$expected2" -a "$current" = "$expected2" ]
^-- SC2166: Prefer [ p ] && [ q ] as [ p -a q ] is not well defined.
In tests/integration/helpers.bash line 309:
check_cgroup_value "cpu.weight" $weight
^-----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
check_cgroup_value "cpu.weight" "$weight"
In tests/integration/helpers.bash line 310:
check_systemd_value "CPUWeight" $weight
^-----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
check_systemd_value "CPUWeight" "$weight"
In tests/integration/helpers.bash line 383:
if [ $CGROUP_UNIFIED = "no" -a ! -e "${CGROUP_MEMORY_BASE_PATH}/memory.memsw.limit_in_bytes" ]; then
^-- SC2166: Prefer [ p ] && [ q ] as [ p -a q ] is not well defined.
In tests/integration/helpers.bash line 412:
local cpu_count=$(grep -c '^processor' /proc/cpuinfo)
^-------^ SC2155: Declare and assign separately to avoid masking return values.
In tests/integration/helpers.bash line 450:
sleep $delay
^----^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
sleep "$delay"
In tests/integration/helpers.bash line 453:
echo "Command \"$@\" failed $attempts times. Output: $output"
^-- SC2145: Argument mixes string and array. Use * or separate argument.
In tests/integration/helpers.bash line 471:
runc state $1
^-- SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
runc state "$1"
In tests/integration/helpers.bash line 472:
if [ $2 == "checkpointed" ]; then
^-- SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
if [ "$2" == "checkpointed" ]; then
In tests/integration/helpers.bash line 484:
mkdir $dir
^--^ SC2086: Double quote to prevent globbing and word splitting.
Did you mean:
mkdir "$dir"
In tests/integration/helpers.bash line 497:
kill -9 $(cat "$dir/pid")
^---------------^ SC2046: Quote this to prevent word splitting.
In tests/integration/helpers.bash line 508:
export ROOT=$(mktemp -d "$BATS_RUN_TMPDIR/runc.XXXXXX")
^--^ SC2155: Declare and assign separately to avoid masking return values.
In tests/integration/helpers.bash line 512:
cd "$ROOT/bundle"
^---------------^ SC2164: Use 'cd ... || exit' or 'cd ... || return' in case cd fails.
Did you mean:
cd "$ROOT/bundle" || exit
In tests/integration/helpers.bash line 535:
cd "$INTEGRATION_ROOT"
^--------------------^ SC2164: Use 'cd ... || exit' or 'cd ... || return' in case cd fails.
Did you mean:
cd "$INTEGRATION_ROOT" || exit
For more information:
https://www.shellcheck.net/wiki/SC2145 -- Argument mixes string and array. ...
https://www.shellcheck.net/wiki/SC2034 -- TESTDATA appears unused. Verify u...
https://www.shellcheck.net/wiki/SC2046 -- Quote this to prevent word splitt...
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It is not used since PR 2757, as all tests are run with cd to bundle
directory.
runc_spec argument count checking is removed since otherwise shellcheck
complains:
> SC2120: runc_spec references arguments, but none are ever passed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Previous commits removed all its users -- the only one left is package's
own unit tests.
Modify those unit tests to configure logrus directly, and remove
ConfigureLogging for good. The world is better without it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This function is somewhat strange and I always wanted to remove it,
as it tries to satisfy both init.go and main.go, which have somewhat
different needs.
It is more straightforward and readable to configure logrus directly.
While at it, simplify errors on panic (errors from logrus.ParseLevel
and strconv.Atoi already contain value which they fail to parse, and
panic already contains enough context to figure out what's wrong).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Use the term "clos group" instead of "container_id group" as the group
that a container belongs to is not necessarily tied to its container id.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
Those states are available since Linux 4.14 (kernel commits
8ef9925b02c23e3838d5 and 06eb61844d841d003). Before this
patch, they were shown as unknown.
This is mostly cosmetical.
Note that I is described in /proc/pid/status as just "idle", although
elsewhere it says it's an idle kernel thread. Let's have it as "idle"
for brevity.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Remove PID field as it is useless.
2. Rewrite parseStat() to make it faster and more correct:
- do not use fmt.Scanf as it is very slow;
- avoid splitting data into 20+ fields, of which we only need 2;
- make sure to not panic on short lines and other bad input;
- add some bad input tests (some fail with old code);
- use LastIndexByte instead of LastIndex.
Benchmarks:
before (from the previous commit message):
> BenchmarkParseStat-4 116415 10804 ns/op
> BenchmarkParseRealStat-4 240 4781769 ns/op
after:
> BenchmarkParseStat-4 1164948 1068 ns/op
> BenchmarkParseRealStat-4 331 3458315 ns/op
We are seeing 10x speedup in a synthetic benchmark, and about 1.4x
speedup in a real world benchmark.
While at it, do not ignore any possible errors, and properly wrap those.
[v2: use pkg/errors more, remove t.Logf from test]
[v3: rebased; drop pkg/errors; gofumpt'ed]
[v4: rebased; improved description]
[v5: rebased; mention bad input tests, added second benchmark results]
[v6: remove PID field, do not use strings.Split, further speedup]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Add a test case that tests parentheses in command.
2. Replace individual comparisons with reflect.DeepEqual.
This also fixes wrong %-style types in Fatalf statements.
3. Replace Fatalf with Errorf so we don't bail out on the first
failure, and do not check result on error.
4. Add two benchmarks. On my laptop, they show:
BenchmarkParseStat-4 116415 10804 ns/op
BenchmarkParseRealStat-4 240 4781769 ns/op
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
runc resolves symlink before doing bind mount. So
we should save original path while formatting CriuReq for
dump and restore.
"checkpoint: resolve symlink for external bind mount" is merged as
da22625f6986f0ef196eaa1f8bb6adce098f0fb7(PR 2902) previously. And reverted
in commit 70fdc0573dced3464e9c31d674559f77c1de3973(PR 3043) duo to behavior changes
caused by commit 0ca91f44f1664da834bc61115a849b56d22f595f(Fixes: CVE-2021-30465)
Signed-off-by: Liu Hua <weldonliu@tencent.com>
This is helpful to kubernetes in cases it knows for sure that the freeze
is not required (since it created the systemd unit with no device
restrictions).
As the code is trivial, no tests are required.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add a test for freezeBeforeSet, checking various scenarios including
those that were failing before the fix in the previous commit.
[v2: add more cases, add a check before creating a unit.]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes the behavior intended to avoid freezing containers/control
groups without it being necessary. This is important for end users of
libcontainer who rely on the behavior of no freeze.
The previous implementation would always get error trying to get
DevicePolicy from the Unit via dbus, since the Unit interface doesn't
contain DevicePolicy.
Signed-off-by: Odin Ugedal <odin@uged.al>
The two exceptions I had to add to codespellrc are:
- CLOS (used by intelrtd);
- creat (syscall name used in tests/integration/testdata/seccomp_*.json).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 2bab4a5 resulted in a warning from gcc:
nsexec.c: In function ‘write_log’:
nsexec.c:171:2: warning: ignoring return value of ‘write’, declared with attribute warn_unused_result [-Wunused-result]
171 | write(logfd, json, ret);
| ^~~~~~~~~~~~~~~~~~~~~~~
As there's nothing we can or want to do in case write fails,
let's just tell the compiler we're not going to use it.
Fixes: 2bab4a5
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Today we support the seccomp build tag only that is used by default.
However, we are not testing that compiling without any build tag works.
I found the CI didn't catch this when working on #2682, that the CI was
green but compilation without build tags was broken.
We test compilation without build tags only, compilation with the only
build tag supported is done extensively in other actions.
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
The contents of the pointer returned on asprintf() error are undefined
i.e., it can be anything there. We set it to NULL on error so that
free() afterwards won't get a garbage pointer.
This patch applies the above to message and stage as well to be
consistent with what we do for json.
Signed-off-by: Kailun Qin <kailun.qin@intel.com>
According to C standards, `size_t` is always an unsigned integer type.
Thus, checking unsigned expressions to be less than zero is not needed.
Signed-off-by: Kailun Qin <kailun.qin@intel.com>
Possibly there was a specific reason to use a rune for this, but I noticed
that there's various parts in the code that has to convert values from a
string to this type. Using a string as type for this can simplify some of
that code.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
GCP images description at [1] claims that:
- For CentOS 8 and CentOS Stream 8, the PowerTools repository is
enabled.
- For CentOS 7, EPEL is enabled.
Apparently,
- we do not need epel for centos-stream-8;
- powertools is not enabled on centos-stream-8 despite [1].
Anyway, the less yum commands the better, as we have seen those fail
sometimes due to occasional networking problems etc.
[1] https://cloud.google.com/compute/docs/images/os-details#centos
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
opencontainers/runc issue 3026 describes a scenario in which OpenFile
failed to open a legitimate existing cgroupfs file. Added debug
(similar to what this commit does) shown that cgroupFd is no longer
opened to "/sys/fs/cgroup", but to "/" (it's not clear what caused it,
and the source code is not available, but they might be using the same
process on the both sides of the container/chroot/pivot_root/mntns
boundary, or remounting /sys/fs/cgroup).
Consider such use incorrect, but give a helpful hint as two what is
going on by wrapping the error in a more useful message.
NB: this can potentially be fixed by reopening the cgroupFd once we
detected that it's screwed, and retrying openat2. Alas I do not have
a test case for this, so left this as a TODO suggestion.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As reported in issue 3119, there is a race in nsexec logging
that can lead to garbled json received by log forwarder, which
complains about it with a "failed to decode" error.
This happens because dprintf (used since the very beginning of nsexec
logging introduced in commit ba3cabf932) relies on multiple write(2)
calls, and with additional logging added by 64bb59f592 a race is
possible between runc init parent and its children.
The fix is to prepare a string and write it using a single call to
write(2).
[v2: NULLify json on error from asprintf]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kir Kolyshkin (3):
libct/cg: GetAllPids: optimize for go 1.16+
libct/cg: improve GetAllPids and readProcsFile
libct/cg: move GetAllPids out of utils.go
LGTMs: AkihiroSuda cyphar
Closes#3133
Fix reading cgroup files from the top cgroup directory, i.e.
/sys/fs/cgroup.
The code was working for for any subdirectory of /sys/fs/cgroup, but
for dir="/sys/fs/cgroup" a fallback (open and fstatfs) was used, because
of the way the function worked with the dir argument.
Fix those cases, and add unit tests to make sure they work. While at it,
make the rules for dir and name components more relaxed, and add test
cases for this, too.
While at it, improve OpenFile documentation, and remove a duplicated
doc comment for openFile.
Without these fixes, the unit test fails the following cases:
file_test.go:67: case {path:/sys/fs/cgroup name:cgroup.controllers}: fallback
file_test.go:67: case {path:/sys/fs/cgroup/ name:cgroup.controllers}: openat2 /sys/fs/cgroup//cgroup.controllers: invalid cross-device link
file_test.go:67: case {path:/sys/fs/cgroup/ name:/cgroup.controllers}: openat2 /sys/fs/cgroup///cgroup.controllers: invalid cross-device link
file_test.go:67: case {path:/ name:/sys/fs/cgroup/cgroup.controllers}: fallback
file_test.go:67: case {path:/ name:sys/fs/cgroup/cgroup.controllers}: fallback
file_test.go:67: case {path:/sys/fs/cgroup/cgroup.controllers name:}: openat2 /sys/fs/cgroup/cgroup.controllers/: not a directory
Here "fallback" means openat2-based implementation fails, and the fallback code
is used (and works).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Run unit tests irrespective of the underlying system configuration, i.e.
even if RDT has not been enabled or is not supported. The tests do not
depend on real kernel interfaces.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
Check that the ClosID directory pre-exists if no L3 or MB schema has
been specified. Conform with the following line from runtime-spec
(config-linux):
If closID is set, and neither of l3CacheSchema and memBwSchema are
set, runtime MUST check if corresponding pre-configured directory
closID is present in mounted resctrl. If such pre-configured directory
closID exists, runtime MUST assign container to this closID and
generate an error if directory does not exist.
Add a TODO note for verifying existing schemata against L3/MB
parameters.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
Handle ClosID parameter of IntelRdt. Makes it possible to use
pre-configured classes/ClosIDs and avoid running out of available IDs
which easily happens with per-container classes.
Remove validator checks for empty L3CacheSchema and MemBwSchema fields
in order to be able to leave them empty, and only specify ClosID for
a pre-configured class.
Signed-off-by: Markus Lehtonen <markus.lehtonen@intel.com>
This was initially added by commits 41d9d26513 and 4a8f0b4db4,
apparently to implement docker run --cgroup container:ID, which was
never merged. Therefore, this code is not and was never used.
It needs to be removed mainly because having it makes it much harder to
understand how cgroup manager works (because with this in place we have
not one or two but three sets of cgroup paths to think about).
Note if the paths are known and there is a need to add a PID to existing
cgroup, cgroup manager is not needed at all -- something like
cgroups.WriteCgroupProc or cgroups.EnterPid is sufficient (and the
latter is what runc exec uses in (*setnsProcess).start).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
filepath.WalkDir function, introduced in Go 1.16, doesn't do stat(2)
on every entry, and is therefore somewhat faster (see below).
Since we have to support Go 1.15, keep the old version for backward
compatibility.
Add a quick benchmark, which shows approximately 3x improvement:
$ go1.15.15 test -bench AllPid -run xxx .
BenchmarkGetAllPids-4 48 23528839 ns/op
$ go version
go version go1.16.6 linux/amd64
$ go test -bench AllPid -run xxx .
BenchmarkGetAllPids-4 147 7700170 ns/op
(Unrelated but worth noting -- go 1.17rc2 is pushing it even further)
$ go1.17rc2 test -bench AllPid -run xxx .
BenchmarkGetAllPids-4 164 6820994 ns/op
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since every cgroup directory is guaranteed to have cgroup.procs file,
we don't have to do filename comparison in GetAllPids() and just read
cgroup.procs in every directory.
While at it, switch readProcsFile to use our own OpenFile.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Add testing against criu-dev branch instead of a released version
(happens to be criu v3.15 at the moment), to check how it works.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Currently there's no way to distinguish between the two cases:
- runc exec failed;
- the command executed returned 1.
This was possible before commit 8477638aab, as runc exec exited with
the code of 255 if exec itself has failed. The code of 255 is the same
convention as used by e.g. ssh.
Re-introduce the feature, document it, and add some tests so it won't be
broken again.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
What it takes is add an empty buildid, which, together with previously
added strip invocation, results in reproducible build!
NB: earlier versions of this patch also added the following:
1. non-random libseccomp install $prefix;
2. "objcopy --enable-deterministic-archives $prefix/lib/libseccomp.a"
to strip ar dates and UIDs/GIDs;
3. "-B=0x00" to EXTRA_LDFLAGS to have non-variable NT_GNU_BUILD_ID.
Apparently, all this is not needed with strip.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This patch
* drops the default `-w` flag for `make static`, which helps with
debugging the static runc binary;
* adds `EXTRA_LDFLAGS="-w -s"` to `script/release.sh` to disable DWARF
generation and symbol table for the release runc binary;
* adds strip in `script/release.sh` for a further size-optimized release
runc binary.
Signed-off-by: Kailun Qin <kailun.qin@intel.com>
As pointed out in TODO item added by commit 64bb59f59, it is not
necessary to have a special sync mechanism for cgroupns, as the parent
adds runc init to cgroup way earlier (before sending nl bootstrap data.
This sync was added by commit df3fa115f9, which was also added a
second cgroup manager.Apply() call, later removed in commit
d1ba8e39f8. It seems the original author had the idea to wait for
that second Apply().
Fixes: df3fa115f9
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
runc init is special. For one thing, it needs to do a few things before
main(), so we have func init() that checks if we're init and does that.
What happens next is main() is called, which does some options parsing,
figures out it needs to call initCommand.Action and so it does.
Now, main() is entirely unnecessary -- we can do everything right from
init().
Hopefully the change makes things slightly less complicated.
From a user's perspective, the only change is runc help no longer
lists 'runc init` (which I think it also good).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Replace ioutil.TempDir (mostly) with t.TempDir, which require no
explicit cleanup.
While at it, fix incorrect usage of os.ModePerm in libcontainer/intelrdt
test. This is supposed to be a mask, not mode bits.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Use t.TempDir instead of ioutil.TempDir. This means no need for an
explicit cleanup, which removes some code, including newTestBundle
and newTestRoot.
2. Move newRootfs invocation down to newTemplateConfig, removing a need
for explicit rootfs creation. Also, remove rootfs from tParam as it
is no longer needed (there was a since test case in which two
containers shared the same rootfs, but it does not look like it's
required for the test).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes using runc with podman on my system (Fedora 34).
> $ podman --runtime `pwd`/runc run --rm --memory 4M fedora echo it works
> Error: unable to start container process: error adding seccomp filter rule for syscall bdflush: permission denied: OCI permission denied
The problem is, libseccomp returns EPERM when a redundant rule (i.e. the
rule with the same action as the default one) is added, and podman (on
my machine) sets the following rules in config.json:
<....>
"seccomp": {
"defaultAction": "SCMP_ACT_ERRNO",
"architectures": [
"SCMP_ARCH_X86_64",
"SCMP_ARCH_X86",
"SCMP_ARCH_X32"
],
"syscalls": [
{
"names": [
"bdflush",
"io_pgetevents",
<....>
],
"action": "SCMP_ACT_ERRNO",
"errnoRet": 1
},
<....>
(Note that defaultErrnoRet is not set, but it defaults to 1).
With this commit, it works:
> $ podman --runtime `pwd`/runc run --memory 4M fedora echo it works
> it works
Add an integration test (that fails without the fix).
Similar crun commit:
* https://github.com/containers/crun/commit/08229f3fb904c5ea19a7d9
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Rather than silently ignoring unknown syscalls, print a warning.
While at it, fix imports ordering (stdlib, others, ours).
[v2: demote Warn to Debug]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
As Cirrus CI does not provide a real terminal this uses the same
'ssh -tt' workaround as the Vagrant setup. This sets up the
CentOS 7 and 8 to allow SSH as root to localhost so that we can run
all the tests via 'ssh -tt'.
Not going through vagrant reduces CI times for CentOS 7 and 8 from 6
minutes to 4 minutes.
Signed-off-by: Adrian Reber <areber@redhat.com>
Without this, the test case fails with
> Writing 1000000 to /sys/fs/cgroup/cpu,cpuacct/runc-cgroups-integration-test/cpu.rt_period_us
> /tmp/bats-run-106836/bats.116418.src: line 548: /sys/fs/cgroup/cpu,cpuacct/runc-cgroups-integration-test/cpu.rt_period_us: Permission denied
Since we do not currently have a setup to test this, this went
unnoticed (can be seen in RHEL8 though).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Adrian Reber <areber@redhat.com>
Since commit f09a3e1b8d, the value passed on to read starts with
a slash, resulting in the first element of the array to be empty.
As a result, the test tries to write to the top-level cgroup, which
fails when rootless:
> # Writing 1000000 to /sys/fs/cgroup/cpu,cpuacct//cpu.rt_period_us
> # /tmp/bats-run-106184/bats.115768.src: line 548: /sys/fs/cgroup/cpu,cpuacct//cpu.rt_period_us: Permission denied
To fix, remove the leading slash.
An alternative fix would be to do "for ((i = 1;" instead of "i = 0", but
that seems less readable.
Fixes: f09a3e1b8d
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
"`runc` X.Y.Z should implement the X.Y version of the specification." is no longer correct.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
TestPodSkipDevicesUpdate checks that updating a pod having SkipDevices: true
does not result in spurious "permission denied" errors in a container
running under the pod. The test is somewhat similar in nature to the
@test "update devices [minimal transition rules]" in tests/integration,
but uses a pod.
This tests the validity of freezeBeforeSet in v1.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was initially added by commit 3e5c199708 because Set (with
r.Freezer = Frozen) was not able to freeze a container.
Now (see a few previous commits) Set can do the freeze, so the explicit
Freeze is no longer needed.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Introduce freezeBeforeSet, which contains the logic of figuring out
whether we need to freeze/thaw around setting systemd unit properties.
In particular, if SkipDevices is set, and the current unit properties
allow all devices, there is no need to freeze and thaw, as systemd
won't write any device rules in this case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In addition to freezing and thawing a container via Pause/Resume,
there is a way to also do so via Set.
This way was broken though and is being fixed by a few preceding
commits. The test is added to make sure this is fixed and won't regress.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The t.Name() usage in libcontainer/integration prevented subtests
to be used, since in such case it returns a string containing "/",
and thus it can't be used to name a container.
Fix this by replacing slashes with underscores where appropriate.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
m.Freeze method changes m.cgroups.Resources.Freezer field, which should
not be done while we're temporarily freezing the cgroup in Set. If this
field is changed, and r == m.cgroups.Resources (as it often happens),
this results in inability to freeze the container using Set().
To fix, add and use a method which does not change r.Freezer field.
A test case for the bug will be added separately.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is necessary in order for runc to be able to configure device
cgroups with --systemd-cgroup on distributions that have very strict
SELinux policies such as openSUSE MicroOS[1].
The core issue here is that systemd is adding its own BPF policy that
has an SELinux label such that runc cannot interact with it. In order to
work around this, we can just ignore the policy -- in theory this
behaviour is not correct but given that the most obvious case
(--systemd-cgroup) will still handle updates correctly, this logic is
reasonable.
[1]: https://bugzilla.suse.com/show_bug.cgi?id=1182428
Fixes: d0f2c25f52 ("cgroup2: devices: replace all existing filters when attaching")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
We need to update the eBPF library so that we can get the raw syscall
errors from bpf(2) syscalls using errors.Is.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
This test the issues fixed by the two preceding commits.
Co-Authored-By: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Odin Ugedal <odin@uged.al>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
If a control group is frozen, all its descendants will report FROZEN
in freezer.state cgroup file.
OTOH cgroup v2 cgroup.freeze is not reporting the cgroup as frozen
unless it is frozen directly (i.e. not via an ancestor).
Fix the discrepancy between v1 and v2 drivers behavior by
looking into freezer.self_freezing cgroup file, which, according
to kernel documentation, will show 1 iff the cgroup was frozen directly.
Co-authored-by: Kir Kolyshkin <kolyshkin@gmail.com>
Signed-off-by: Odin Ugedal <odin@uged.al>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Since device updates in cgroup v2 are atomic for systemd, there is no
need to freeze the processes before running the updates.
Signed-off-by: Odin Ugedal <odin@uged.al>
Run device update tests on cgroup v2, and add a test verifying that we
don't allow access to devices when we don't intend to.
Signed-off-by: Odin Ugedal <odin@uged.al>
Add CAP_SYSLOG to ensure that /dev/kmsg can be accesses on systems where
the sysctl kernel.dmesg_restrict = 1.
Signed-off-by: Odin Ugedal <odin@uged.al>
The main change is the switch to Go 1.13-style "%w" error wrapping,
dropping one of the github.com/pkg/errors dependencies we have left.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Lines in /etc/group longer than 64 characters breaks the current
implementation of group parser. This is caused by bufio.Scanner
buffer limit.
Fix by re-using the fix for a similar problem in golang os/user,
namely https://go-review.googlesource.com/c/go/+/283601.
Add some tests.
Co-authored-by: Andrey Bokhanko <andreybokhanko@gmail.com>
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Same as in other places (other parsers here, as well as golang os/user
parser and glibc parser all tolerate extra space at BOL and EOL).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Don't ignore close(2) return code, rather bail if there is any
unexpected failures. By checking the close return code we make sure we
don't introduce the same bug (closing an already closed fd) I've fixed
in the previous patch.
As a side note, we are not handling in this patch when close(2) returns
EINTR and the go runtime, since go 1.14, sends SIGURG to preempt
goroutines. This should not happen here though, as nsenter is guaranteed
to be executed before the go runtime starts.
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
This was closed in the child[1], before calling clone_parent (so runc
INIT will have this fd closed too), there is no point closing it again.
This was not causing issues because we ignore the return code of
close(2) and no one was opening a new fd between both calls to close.
However, with the new patches that I'm working on (PR #2576), this
problem is no longer inocuos: we do open a new fd in that PR, sometimes
that fd is allocated between the two close(2) calls and, as the lowest
fd is allocated to the new fd, sometimes the second close ends up
incorrectly closing this new fd.
Before it was not a problem in practice, but it was incorrect
nevertheless.
This seems to be long standing bug, present since at least 2018
(a54316bae), when SYNC_GRANDCHILD was introduced.
[1]: https://github.com/opencontainers/runc/blob/5547b5774f71f75a088e7432fa961778750a0fbd/libcontainer/nsenter/nsexec.c#L888
Co-authored-by: Alban Crequy <alban@kinvolk.io>
Signed-off-by: Rodrigo Campos <rodrigo@kinvolk.io>
When running a script from an azure file share interrupted syscall
occurs quite frequently, to remedy this add retries around execve
syscall, when EINTR is returned.
Signed-off-by: Maksim An <maksiman@microsoft.com>
The test is failing like this:
not ok 70 runc run --no-pivot must not expose bare /proc
# (in test file tests/integration/no_pivot.bats, line 20)
# `[[ "$output" == *"mount: permission denied"* ]]' failed
# runc spec (status=0):
#
# runc run --no-pivot test_no_pivot (status=1):
# unshare: write error: Operation not permitted
Apparently, a recent kernel commit db2e718a47984b9d prevents
root from doing unshare -r unless it has CAP_SETFPCAP.
Add the capability for this specific test.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These are not used anywhere outside of the package
(I have also checked the only external user of the package
(github.com/google/cadvisor).
No changes other than changing the case. The following
identifiers are now private:
* IntelRdtTasks
* NewLastCmdError
* NewStats
Brought to you by gorename.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
... the stack, so every caller will automatically benefit from it.
The only change that it causes is the user in
libcontainer/process_linux.go will get a better error message.
[v2: typo fix]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For errors that only have a string and an underlying error, using
fmt.Errorf with %w to wrap an error is sufficient.
In this particular case, the code is simplified, and now we have
unwrappable errors as a bonus (same could be achieved by adding
(*LastCmdError).Unwrap() method, but that's adding more code).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Initially, this was copied over from libcontainer/cgroups, where it made
sense as for cgroup v1 we have multiple controllers and mount points.
Here, we only have a single mount, so there's no need for the whole
type.
Replace all that with a simple error (which is currently internal since
the only user is our own test case).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case getIntelRdtData() returns an error, d is set to nil.
In case the error returned is of NotFoundError type (which happens
if resctlr mount is not found in /proc/self/mountinfo), the function
proceeds to call d.join(), resulting in a nil deref and a panic.
In practice, this never happens in runc because of the checks in
intelrdt() function in libcontainer/configs/validate, which raises
an error in case any of the parameters are set in config but
the IntelRTD itself is not available (that includes checking
that the mount point is there).
Nevertheless, the code is wrong, and can result in nil dereference
if some external users uses Apply on a system without resctrl mount.
Fix this by removing the exclusion.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In cases we have something like
if y != "" {
x = y
}
where both x and y are strings, and x was not set before,
it makes no sense to have a condition, as such code is
equivalent to mere
x = y
Simplify such cases by removing "if".
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This removes libcontainer's own error wrapping system, consisting of a
few types and functions, aimed at typization, wrapping and unwrapping
of errors, as well as saving error stack traces.
Since Go 1.13 now provides its own error wrapping mechanism and a few
related functions, it makes sense to switch to it.
While doing that, improve some error messages so that they start
with "error", "unable to", or "can't".
A few things that are worth mentioning:
1. We lose stack traces (which were never shown anyway).
2. Users of libcontainer that relied on particular errors (like
ContainerNotExists) need to switch to using errors.Is with
the new errors defined in error.go.
3. encoding/json is unable to unmarshal the built-in error type,
so we have to introduce initError and wrap the errors into it
(basically passing the error as a string). This is the same
as it was before, just a tad simpler (actually the initError
is a type that got removed in commit afa844311; also suddenly
ierr variable name makes sense now).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Errors from unix.* are always bare and thus can be used directly.
Add //nolint:errorlint annotation to ignore errors such as these:
libcontainer/system/xattrs_linux.go:18:7: comparing with == will fail on wrapped errors. Use errors.Is to check for a specific error (errorlint)
case errno == unix.ERANGE:
^
libcontainer/container_linux.go:1259:9: comparing with != will fail on wrapped errors. Use errors.Is to check for a specific error (errorlint)
if e != unix.EINVAL {
^
libcontainer/rootfs_linux.go:919:7: comparing with != will fail on wrapped errors. Use errors.Is to check for a specific error (errorlint)
if err != unix.EINVAL && err != unix.EPERM {
^
libcontainer/rootfs_linux.go:1002:4: switch on an error will fail on wrapped errors. Use errors.Is to check for specific errors (errorlint)
switch err {
^
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Do this for all errors except one from unix.*.
This fixes a bunch of errorlint warnings, like these
libcontainer/generic_error.go:25:15: type assertion on error will fail on wrapped errors. Use errors.As to check for specific errors (errorlint)
if le, ok := err.(Error); ok {
^
libcontainer/factory_linux_test.go:145:14: type assertion on error will fail on wrapped errors. Use errors.As to check for specific errors (errorlint)
lerr, ok := err.(Error)
^
libcontainer/state_linux_test.go:28:11: type assertion on error will fail on wrapped errors. Use errors.As to check for specific errors (errorlint)
_, ok := err.(*stateTransitionError)
^
libcontainer/seccomp/patchbpf/enosys_linux.go:88:4: switch on an error will fail on wrapped errors. Use errors.Is to check for specific errors (errorlint)
switch err {
^
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Instead of using errors.Wrap, use fmt.Errorf with %w for error wrapping.
Also, use errors.Is instead of direct error comparison.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Use fmt.Errorf with %w instead.
Convert the users to the new wrapping.
This fixes an errorlint warning.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This one is tough as errorlint insists on using errors.Is, and the
latter is known to not work for Go 1.13 which we still support.
So, add a nolint annotation to suppress the warning, and a TODO to
address it later.
For intelrdt, we can do the same, but it is easier to reuse the very
same function from fscommon (note we can't use fscommon for other stuff
as it expects cgroupfs).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This should result in no change when the error is printed, but make the
errors returned unwrappable, meaning errors.As and errors.Is will work.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Errors from os.Open, os.Symlink etc do not need to be wrapped, as they
are already wrapped into os.PathError.
Error from unix are bare errnos and need to be wrapped. Same
os.PathError is a good candidate.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Errors returned by unix are bare. In some cases it's impossible to find
out what went wrong because there's is not enough context.
Add a mountError type (mostly copy-pasted from github.com/moby/sys/mount),
and mount/unmount helpers. Use these where appropriate, and convert error
checks to use errors.Is.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This builds on top of recently introduced fscommon.ParseError.
Errors returned from parsers (mostly ones used by GetStats()) are all
different, and many are incomplete. For example, in many cases errors
from strconv.ParseUint are returned as is, meaning there is no context
telling which file we were reading. Similarly, errors from
fscommon.ParseKeyValue should be wrapped to add more context.
Same is true for scanner.Err().
OTOH, errors from fscommon.GetCgroup* do have enough context and there
is no need to wrap them.
Fix all the above.
While at it, add missing scanner.Err() checks.
[v2: use parseError, not ParseError]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This builds on top of recently introduced fscommon.ParseError.
Errors returned from parsers (mostly ones used by GetStats()) are all
different, and many are incomplete. For example, in many cases errors
from strconv.ParseUint are returned as is, meaning there is no context
telling which file we were reading. Similarly, errors from
fscommon.ParseKeyValue should be wrapped to add more context.
Same is true for scanner.Err().
One special case that repeats a few times is "malformed line: xxx".
Add and use a helper for that to simplify things.
OTOH, errors from fscommon.GetCgroup* do have enough context and there
is no need to wrap them.
Fix all the above.
While at it, add a missing scanner.Err() check.
[v2: use parseError not ParseError]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Introduce ParseError type as a way to unify error messages related to
file parsing. Use it from GetCgroup* functions.
2. Do not discard the error from strconv.Parse{Int,Uint} -- it contains
the value being parsed, and the details about the error.
2. As the error above already contains the value, drop it from format.
[v2: use path.Join in Error]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. Do not wrap errors returned from fscommon.GetCgroupParamUint -- those
errors already have enough context.
2. Instead of parsing "max" ourselves, use GetCgroupParamUint which does
it, and then convert MaxUint64 to 0 (we do it historically since
commit 087b953dc5, and while using MaxUint64 as is seems fine,
there may be some existing users who rely on the old behavior).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Errors from strconv.Atoi are already descriptive enough, and contain the
value being converted, so our error messages do not need to contain it.
While at it, use %w to wrap errors.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The error from fscommon.GetCgroup* already contains the file name and so
on, so there's no need to wrap it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The error returned from strconv.ParseUint is already pretty descriptive,
something like:
strconv.ParseUint: parsing "000d": invalid syntax
So, there is no need to add more context to it.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Using fmt.Errorf for errors that do not have %-style formatting
directives is an overkill. Switch to errors.New.
Found by
git grep fmt.Errorf | grep -v ^vendor | grep -v '%'
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
An errror from ioutil.WriteFile already contains file name, so there is
no need to duplicate that information.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Kir Kolyshkin (3):
libct/cg/fs/blkio: do not set weight == 0
libct/cg/fs2: set per-device io weight if available
tests/int/cgroups: add test for bfq per-device weight
LGTMs: AkihiroSuda mrunalp cyphar
Closes#3022
Current runc man pages are ugly (no proper man page formatting)
and very short (mostly just a copy-paste from the "runc <command>
--help" output. They are also somewhat obsoleted as not all CLI updates
were propagated to man/*.
This commits makes the first step to solving this.
In short:
- added some more information about some options;
- lots of formatting fixes;
- use references to other man pages and web pages;
- fix SYNOPSYS (formatting, mostly);
- removed the repeated description of <container_id> from every page;
- added SEE ALSO;
- something else I forgot.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Now runc puts dump/restore logs in c.root defaultly, which will be deleted
when container exits. So if checkpinting/restoring failed, we can not get
these logs and analyze why.
This patch lets criu use its default if --work-path is not set:
- Use WorkDirectory found in criu's configfile.
- Use ImageDirectory.
Signed-off-by: Liu Hua <weldonliu@tencent.com>
Per-device weight is supported since kernel v5.4 (kernel commit
795fe54c2a8), so let's set those if supplied.
[v2: implement a more relaxed check in bfqDeviceWeightSupported]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Commit 52390d6804 made this parameters obsoleted, but they are
still shown in e.g. runc update --help output.
Hide them (and maybe in 5 years we can remove them).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This was removed from runc exec by commit f61c6e413f about 5 years ago,
so it's time to remove it entirely.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
For per-device weight, you can set weight and/or leaf weight.
The problem is, with the recent fix to use BFQ on cgroup v1,
if per-device weights are set, the code tries to set device
weight to blkio.bfq.weight, and the leaf weight to
blkio.leaf_weight_device. The latter file does not exist on
kernels v5.0, meaning one can not set any per-device weights
at all.
The fix is to only set weights if they are non-zero (i.e. set).
The test case will come in a following commit.
Fixes: 6339d8a0dd
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is a better place as cgroups itself is using these.
Should help with moving more stuff common in between fs and fs2 to
fscommon.
Looks big, but this is just moving the code around:
fscommon/{fscommon,open}.go -> cgroups/file.go
fscommon/fscommon_test.go -> cgroups/file_test.go
and fixes for TestMode moved to a different package.
There's no functional change.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The `errors.Is(err, unix.EINVAL)` check in `haveBpfProgReplace()` was
broken because the `cilium/ebpf` library did not "wrap" errors.
https://github.com/cilium/ebpf/blob/v0.6.0/link/program.go#L72
So the eBPF support of runc was broken for kernel prior to 5.6.
This commit bumps up cilium/ebpf to contain cilium/ebpf PR 320.
Fix opencontainers/runc issue 3008
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
- Update the blkio cgroup to support the BFQ I/O Scheduler, that has
replaced CFQ in the Linux kernel.
- BFQ is controlled through blkio.bfq.weight[_device] instead of
CFQ's blkio.weight[_device] in cgroups v1.
- BFQ does not support blkio.leaf_weight[_device], so that behavior
remains untouched.
- Do not change behavior on legacy CFQ systems.
- Enable using blkio weights on BFQ systems.
Signed-off-by: Antti Kervinen <antti.kervinen@intel.com>
1. The meaning of SkipDevices is what it is -- do not set any
device-related options.
2. Reverts the part of commit 108ee85b82 which skipped the freeze
when the SkipDevices is set. Apparently, the freeze is needed on
update even if no Device* properties are being set.
3. Add "runc update" to "runc run [device cgroup deny]" test.
Fixes: 752e7a8249
Fixes: 108ee85b82
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Sebastiaan van Stijn (2):
libcontainer: relax validation for absolute paths
configs/validator: move cgroup validation to the list of checks
LGTMs: kolyshkin cyphar
Closes#3004
Commits 1f1e91b1a0 and 2192670a24
added validation for mountpoints to be an absolute path, to match the OCI
specs.
Unfortunately, the old behavior (accepting the path to be a relative path)
has been around for a long time, and although "not according to the spec",
various higher level runtimes rely on this behavior.
While higher level runtime have been updated to address this requirement,
there will be a transition period before all runtimes are updated to carry
these fixes.
This patch relaxes the validation, to generate a WARNING instead of failing,
allowing runtimes to update (but allowing them to update runc to the current
version, which includes security fixes).
We can remove this exception in a future patch release.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
It seems that we are triggering the mutli-attach fallback in the fedora
CI, but we don't have enough debugging information to really know what's
going on, so add some. Unfortunately the amount of information we have
available with eBPF programs in general is fairly limited (we can't get
their bytecode for instance).
We also demote the "more than one filter" warning to an info message
because it happens very often under the systemd cgroup driver (likely
when systemd configures the cgroup it isn't deleting our old program, so
when our apply code runs after the systemd one there are two running
programs).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
It turns out that the cilium eBPF library doesn't degrade gracefully if
BPF_F_REPLACE is not supported, so we need to work around it by treating
that case as we treat the more-than-one program case.
It also turns out that we weren't passing BPF_F_REPLACE explicitly, but
this is required by the cilium library (causing EINVALs).
Fixes: d0f2c25f52 ("cgroup2: devices: replace all existing filters when attaching")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Noticed that the check of trying to use both rootful and rootless
in NewDbusConnManager never worked, as we never set dbusInited to true.
Do that. While at it, protect this with the mutex (against the
case of two goroutines simultaneously calling NewDbusConnManager).
This is a rare call, so taking read-only then read-write mutex does not
make sense.
Fixes: c7f847ed3a
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Restoring an SELinux enabled container with Podman will result in
a container with the exactly same SELinux process labels as during
checkpointing. CRIU takes care of all the process labels.
Restoring multiple copies of a checkpointed container will result in all
containers having the same SELinux process labels, which might be
undesired.
When looking at Pods all container in a Pod share the process label
of the infrastructure container. To restore a container into and
existing Pod it is necessary to tell CRIU to restore the container
with the infrastructure container process label.
CRIU supports setting different process labels using --lsm-profile for a
long time and this just passes the process label information from runc
to CRIU.
Unfortunately CRIU has a bug as no one was using the --lsm-profile
option so this changes requires the upcoming CRIU version 3.16.
Signed-off-by: Adrian Reber <areber@redhat.com>
... and remove the one from tests/integration.
The idea is similar to the one for the test case being removed -- try
updating device rules many times to make sure we are not leaking eBPF
programs after every update/Set(). This is better though as we can
really change the device rules every time (which "runc update" can't)
and check that the rule is applied.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These were deprecated and moved; the stubs were included in the
last two (rc94, rc95) releases, so external consumers would have
the chance to update their code.
Removing this so that this doesn't get into v1.0.0 GA
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
These were deprecated and moved; the stubs were included in the
last two (rc94, rc95) releases, so external consumers would have
the chance to update their code.
Removing this so that this doesn't get into v1.0.0 GA
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This fixes isDbusError function, introduced by commit bacfc2c. Due to a
type error it was not working at all.
This also fixes the whole "retry on dbus disconnect" logic.
This also fixes a regression in startUnit (and cgroupManager.Apply()),
which should never return "unit already exists" error but it did.
Fixes: bacfc2c
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The runc update CLI is not able to modify devices, so let's set SkipDevices
(so that a cgroup controller won't try to update devices cgroup).
This helps use cases when some other device management (NVIDIA GPUs)
applies its configuration on top of what runc does.
Make sure we do not save SkipDevices into state.json.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
These functions are called from multiple places,
and if t.Helper() is not used, the context is not clear.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Explain where the "/dev/tty: no such device or address" error is coming
from, and provide ways to solve the issue.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
libcontainer/cgroups/devices/devices_emulator.go:261:9: `if` block ends with a `return` statement, so drop this `else` and outdent its block (golint)
} else {
^
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This prevents having to maintain GoDoc for the stub implementations,
and makes sure that the "stub" implementations have the same signature
as the "non-stub" versions.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
gofumpt (mvdan.cc/gofumpt) is a fork of gofmt with stricter rules.
Brought to you by
git ls-files \*.go | grep -v ^vendor/ | xargs gofumpt -s -w
Looking at the diff, all these changes make sense.
Also, replace gofmt with gofumpt in golangci.yml.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
strconv.ParseUint(..., 0) is not really safe, because on 32-bit
architectures it will trigger runtime errors when trying to parse large
numbers (which in the case of the cgroupv2 io controller, is almost
certainly going to happen).
Fixes: 1932917b71 ("libcontainer: add initial support for cgroups v2")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Kubelet and cAdvisor depend on the metrics having the same values as in
cgroupv1, but we didn't correctly map the number of read and write IOs
to the correct cgroupv1 stats table (blkio.io_serviced).
In addition, don't leak any extra stats in our output -- if users need
that information we can always add a new field for it.
Reported-by: Yashpal Choudhary <yashpal.c1995@gmail.com>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The idea is to mimic what kubelet is doing, with minimum amount of code.
First, create a slice with SkipDevices=true. It should have access to
all devices.
Next, create a scope within the above slice, allowing access to /dev/full
only.
Check that within that scope we can only access /dev/full and not other
devices (such as /dev/null).
Repeat the test with SkipDevices=false, make sure we can not access any
devices (as they are disallowed by a parent cgroup). This is done only
to assess the test correctness.
NOTE that cgroup v1 and v2 behave differently for SkipDevices=false
case, and thus the check is different. Cgroup v1 returns EPERM on
writing to devices.allow, so cgroup manager's Set() fails, and we check
for a particular error from m.Set(). Cgroup v2 allows to create a child
cgroup, but denies access to any device (despite access being enabled)
-- so we check the error from the shell script running in that cgroup.
Again, this is only about SkipDevices=false case.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The "runc run [device cgroup allow rm block device]" test calls lsblk
three times to get device name, minor and major number. This creates a
potential problem when the devices are changed between the calls.
Simplify the code by using bash read together with IFS (as there's no
way to have lsblk output MAJOR:MINOR pair without a semicolon).
Note that head -n 1 is not needed as read already reads a single line.
[v2: don't use PATH as CentOS7's lsblk does not support it.]
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Apparently not everyone compiles runc via the provided Makefile. For
example, one can just run "go build", in which case Version variable
is left empty, which leads to:
$ ./runc -v
runc version spec: 1.0.2-dev
go: go1.16.3
Surely, the main problem here is runc was built in a wrong way, but the
second problem is such output is very confusing -- it may seem that we
have runc 1.0.2.
To solve, make sure to _always_ add version (even if empty), and set the
default to "unknown".
NOTE this does not change anything in case runc is compiled via the
Makefile.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
It seems that golangci-lint didn't warn us about new lint errors that
were added after we enabled it, so just run the full thing and give us
all the errors on every PR run -- as long as we keep master lint-clean
it doesn't matter whether we set this or not.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Most of these were false positives or cases where we want to ignore the
lint, but the change to the BPF generation is actually useful.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Commit 108ee85b82 adds SkipDevices flag, which is used by kubernetes
to create cgroups for pods.
Unfortunately the above commit falls short, and systemd DevicePolicy and
DeviceAllow properties are still set, which requires kubernetes to set
"allow everything" rule.
This commit fixes this: if SkipDevices flag is set, we return
Device* properties to allow all devices.
Fixes: 108ee85b82
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This is to ensure that we aren't leaking eBPF programs after "runc
update". Unfortunately we cannot directly test the behaviour of cgroup
program updates in an integration test because "runc update" doesn't
support that behaviour at the moment.
So instead we rely on the fact that each "runc update" implicitly
triggers the devices rules to be updated. Without the previous patches
applied, this new test will fail with errors (on cgroupv2 systems).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
In the normal cases (only one existing filter or no existing filters),
just make use of BPF_F_REPLACE if there is one existing filter. However
if there is more than one filter applied, we should probably remove all
other filters since the alternative is that we will never remove our old
filters.
The only two other viable ways of solving this problem would be to use
BPF pins to either pin the eBPF program using a predictable name (so we
can always only replace *our* programs) or to switch away from custom
programs and instead use eBPF maps (which are pinned) and thus we just
update the map conntents to update the ruleset. Unfortunately these both
would add a hard requirement of bpffs and would require at least a minor
rewrite of the eBPF filtering code -- which is better left for another
time.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
There were several issues with the previous cgroupv2 devices filter
generator implementation, stemming from the previous implementation
using a few too many tricks to implement the correct cgroup behaviour
(rules were handled in reverse order, with wildcards having particularly
special interpretations). As a result, some slightly odd configurations
with rules in specific orders could result in incorrect filters being
generated.
By switching to the emulator which is already used by cgroupv1, we can
guarantee that the behaviour of filters in both cgroup versions will be
identical, as well as making use of the hardenings in the emulator (not
allowing users to add deny rules the kernel will ignore).
(Note that because the ordering of the devices emulator rules is
deterministic and based on the rule value, the existing test rules had
to be reordered slightly.)
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
The devices cgroup emulator is also useful for removing unneeded rules
as well as computing what the final default-allow state of the filter
will be (allow-list or deny-list).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
When running inside a Docker container, systemd is not available. The
new TestFdLeaksSystemd forgot to include the relevant t.Skip section.
Fixes: a7feb42395 ("libct/int: add TestFdLeaksSystemd")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Use "git describe --dirty --long" instead of "git rev-parse". As a
result, the commit ID will contain the closest tag, the number of commits
since the tag, and the (abbreviated) git commit sha (see example below).
NOTE that this tag is still unique and can be used instead of bare sha
for all git commands.
Example output of "runc -v | grep commit".
Before:
commit: 4d87573871
After:
commit: v1.0.0-rc95-9-g6f55d074
This means that
- the closest tag is v1.0.0-rc95
- there were 9 commits after the tag
- the abbreviated sha is 6f55d074
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
According to cgroup v2 documentation [1]:
> Freezing of the cgroup may take some time; when this action is
> completed, the “frozen” value in the cgroup.events control file will
> be updated to “1” and the corresponding notification will be issued.
Implement polling of cgroup.events, waiting for "frozen 1" to appear.
In case something goes wrong, limit the maximum number of retries and
return "undefined" after some time (currently 10s).
[1] https://www.kernel.org/doc/html/latest/admin-guide/cgroup-v2.html
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Before this patch, setFreezer does
- open/read/close (to check if the freezer is supported)
- open/write/close (to set the value)
- open/read/close (to check the value)
Three opens is a bit excessive. Refactor to only open the file once:
- open (to check if the freezer is supported)
- write (to set the value)
- seek/read (to check the value)
- close
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
In case configs.Undefined or any wrong value is passed, there is no need
to check whether the freezer is supported.
Move arguments check to the beginning to avoid an unnecessary call to
supportFreezer().
While at it, simplify the "whether to return an error if freezer is not
supported" check.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
Historically, we never returned an error from failed startUnit
or stopUnit. The startUnit case was fixed by commit 3844789.
It is time to fix stopUnit, too. The reasons are:
1. Ignoring an error from stopUnit means an unexpected trouble down the
road, for example a failure to create a container with the same name:
> time="2021-05-07T19:51:27Z" level=error msg="container_linux.go:380: starting container process caused: process_linux.go:385: applying cgroup configuration for process caused: Unit runc-test_busybox.scope already exists."
2. A somewhat short timeout of 1 second means the cgroup might
actually be removed a few seconds later but we might have a
race between removing the cgroup and creating another one
with the same name, resulting in the same error as amove.
So, return an error if removal failed, and increase the timeout.
Now, modify the systemd cgroup v1 manager to not mask the error from
stopUnit (stopErr) with the subsequent one from cgroups.RemovePath,
as stopErr is most probably the reason why RemovePath failed.
Note that for v1 we do want to remove the paths even in case of a
failure from stopUnit, as some were not created by systemd.
There's no need to do that for v2, thanks to unified hierarchy,
so no changes there.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
1. These tests can't be run in parallel since they do check
a global variable (mbaScEnabled).
2. findIntelRdtMountpointDir() relies on mbaScEnabled to be initially
set to the default value (false) and this the test fails if run
more than once:
> go test -count 2
> ...
> intelrdt_test.go:243: expected mbaScEnabled=false, got true
> --- FAIL: TestFindIntelRdtMountpointDir/Valid_mountinfo_with_MBA_Software_Controller_disabled (0.00s)
Fixes: 2c70d2384
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
runc resolves symlink before doing bind mount. So
we should save original path while formatting CriuReq for
checkpoint.
Signed-off-by: Liu Hua <weldonliu@tencent.com>
2021-04-25 09:50:00 +08:00
1585 changed files with 172609 additions and 113700 deletions
- name:notoolchain in go.mod# See https://github.com/opencontainers/runc/pull/4717, https://github.com/dependabot/dependabot-core/issues/11933.
run:|
if grep -q '^toolchain ' go.mod; then echo "Error: go.mod must not have toolchain directive, please fix"; exit 1; fi
- name:noexclude nor replace in go.mod
run:|
if grep -Eq '^\s*(exclude|replace) ' go.mod; then echo "Error: go.mod must not have exclude/replace directive, it breaks go install. Please fix"; exit 1; fi
commit:
runs-on:ubuntu-20.04
# Only check commits on pull requests.
if:github.event_name == 'pull_request'
permissions:
contents:read
pull-requests:read
runs-on:ubuntu-24.04
steps:
- name:get pr commits
if:github.event_name == 'pull_request'# Only check commits on pull requests.
@test -z "$$(git status --porcelain -- go.mod go.sum vendor/)"\
||(echo -e "git status:\n $$(git status -- go.mod go.sum vendor/)\nerror: vendor/, go.mod and/or go.sum not up to date. Run \"make vendor\" to update";exit 1)\
&&echo"all vendor files are up to date."
cross:runcimage
$(CONTAINER_ENGINE) run $(CONTAINER_ENGINE_RUN_FLAGS)\
runc fully supports cgroup v2 (unified mode) since v1.0.0-rc93.
To use cgroup v2, you might need to change the configuration of the host init system.
Fedora (>= 31) uses cgroup v2 by default and no extra configuration is required.
The following distributions are known to use cgroup v2 by default:
<!-- the list should be kept in sync with https://github.com/rootless-containers/rootlesscontaine.rs/blob/master/content/getting-started/common/cgroup2.md -->
- Fedora (since 31)
- Arch Linux (since April 2021)
- openSUSE Tumbleweed (since c. 2021)
- Debian GNU/Linux (since 11)
- Ubuntu (since 21.10)
- RHEL and RHEL-like distributions (since 9)
On other systemd-based distros, cgroup v2 can be enabled by adding `systemd.unified_cgroup_hierarchy=1` to the kernel cmdline.
## Am I using cgroup v2?
@@ -26,18 +34,18 @@ The recommended systemd version is 244 or later. Older systemd does not support
Make sure you also have the `dbus-user-session` (Debian/Ubuntu) or `dbus-daemon` (CentOS/Fedora) package installed, and that `dbus` is running. On Debian-flavored distros, this can be accomplished like so:
```console
$sudo apt install -y dbus-user-session
$systemctl --user start dbus
```bash
sudo apt install -y dbus-user-session
systemctl --user start dbus
```
## Rootless
On cgroup v2 hosts, rootless runc can talk to systemd to get cgroup permissions to be delegated.
<!-- TBD: features that were already deprecated and removed -->
- The latest release in May 2029 may not necessarily support cgroup v1, but there will be at least one maintained branch with the support for cgroup v1.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
lfbzhm