## Type of Change
- [ ] Bug fix
- [x] New feature
- [ ] Breaking change
- [x] Documentation update
## Motivation and Context
Adds the `container copy` (aliased as `cp`) command to copy files
between a running container and the local filesystem.
I saw [#1023](https://github.com/apple/container/pull/1023) and the
feedback from @dcantah — the previous attempt relied on tar being
installed inside the container.
This implementation takes the recommended approach:
file transfers go through the guest agent via the existing
`copyIn`/`copyOut` methods on the core `Containerization`, with no
dependency on container tooling.
## Testing
- [x] Tested locally
- [x] Added/updated tests
- [x] Added/updated docs
---------
Co-authored-by: jwhur <57657645+JaewonHur@users.noreply.github.com>
- Closes#1565.
- `runtime-config.toml` suggested the config
pertains to just the runtime (managing container
workloads) aspect of the system.
- Also renames `ADDR` columns to `IP` in
`container ls` and `container builder status`.
- Closes#1566.
- Currently on `system start` we only remove the read-only user config
if the user config file exists. This means that changes to the config
are processed properly on restart, but a removal is ignored, and the old
read-only settings linger.
- Closes#1533
- This PR aligns `volume list --format json` date
encoding with volume inspect by using ISO-8601
instead of the default numeric reference-date
format.
- Bumps the github-actions group with 1 update:
- [actions/labeler](https://github.com/actions/labeler).
- Updates `actions/labeler` from 6.0.1 to 6.1.0
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
- This change adds two types to support opaque runtime
data passing through the APIServer:
- RuntimeConfiguration `data`
- RuntimeLinuxData type
- LinuxRuntimeData defines runtime specific information.
With this change, it will support `variant` only. The
optional RuntimeConfiguration `data` field encodes
runtime specific data to pass through the APIServer.
It is decoded as needed by the runtime.
- The idea is to eventually move all runtime specific data
into the `data` field so that the APIServer is only aware
of generic container information.
- Closes#1318.
- Closes#1378.
- Reduces the complexity and coupling for IP allocation.
- Runtimes connect to networks for the life of the running container.
The runtime shuts down on connection loss.
- Networks automatically deallocate a runtime's IP address and hostname
record on connection loss.
- Removes AllocatedAttachment as this is no longer necessary. The
`bootstrap()` XPC now takes a `NetworkBootstrapInfo` array which
parallels the attachments in the bundle config and provides the network
plugin attributes needed to create VM network interface configurations.
## Type of Change
- [x] Bug fix
- [ ] New feature
- [ ] Breaking change
- [ ] Documentation update
## Motivation and Context
Simplify IP allocation and make deallocation more reliable.
## Testing
- [x] Tested locally
- [ ] Added/updated tests
- [ ] Added/updated docs
- This adds support for reading configurations from
a three layer hierarchy:
1. User provided TOML
2. Install root TOML
3. Code defaults
- We add some code to support plugin configurations
via the ConfigurationLoader. Each plugin can provide
a struct with an accompanying id that gets used to
parse the scoped section of the TOML.
- Closes#1521.
- Using URL for filesystem paths is bad practice. FilePath is safer and
more ergonomic.
- Same pattern as #1480 (HostDNSResolver) and #1518 (PacketFilter).
- Part of #1318.
- Part of #1378.
- Removes network plugin `deallocate()`, and allocate takes an
`XPCServerSession` that registers an `onDisconnect` handler that
performs deallocation.
- ContainerService now tracks `networkSessions` for allocations.
- Part of #1318.
- Part of #1378.
- `RouteHandler` now accepts an additional `XPCServerSession` which
include a disconnect handler.
- `XPCServer` configures a session for each client connection and passes
it to every route handler.
- `XPCServer.route(fn:)` can be used to wrap existing route handlers
that don't care about sessions.
- `XPCClient` adds functions to support client-side session handling.
- Update all existing routes.
- Closes#1520.
- Using URL for filesystem paths is bad practice. FilePath is safer and
more ergonomic.
- Same pattern as #1480 (HostDNSResolver) and #1518 (PacketFilter).
There is a bug in the artifact downloading where a single artifact
matching a pattern is extracted flat to the working directory without
creating a subdirectory. This meant the glob
`pr-coverage-*/pr-number.txt` never matched, silently skipping the
comment step on every run.
Fixed by using `merge-multiple: true` with an explicit `path`, giving a
deterministic download location regardless of artifact count.
## Type of Change
- [X] Bug fix
- [ ] New feature
- [ ] Breaking change
- [ ] Documentation update
## Motivation and Context
Fix the GitHub actions to post coverage numbers
## Testing
- [ ] Tested locally
- [ ] Added/updated tests
- [ ] Added/updated docs
This adds a swift Decoder that decodes from a swift-configuration
ConfigSnapshot, and adds the ability to modify how select types are
decoded within the
ConfigSnapshotDecoder. This is something that JSONEncoder already does
with URLs specifically, however this enables users to configure this
behavior for any arbitrary type regardless of existing Codable
conformance.
The decoder will help us expand on top of (pr:
https://github.com/apple/container/pull/1425), which switches user
configuration to TOML. With this PR we will be able to provide a multi
layered TOML if desired, allowing container to ship a default TOML
instead of encoding them in code constants.
- Closes#1517.
- Using URL for filesystem paths is bad practice. FilePath is safer and
more ergonomic.
- Sibling of #1480 (HostDNSResolver), same pattern.
This utilizes swift test `--skip-build` in combination with a single big
swift test compilation step. By doing this we reduce duplicated swift
compilation between unit and integration.
## Type of Change
- [ ] Bug fix
- [X] New feature
- [ ] Breaking change
- [ ] Documentation update
## Motivation and Context
Coverage GitHub action is taking too long, so hopefully reducing
duplicated work decreases time.
## Testing
- [X] Tested locally
- [ ] Added/updated tests
- [ ] Added/updated docs
- Discussion topic #1336.
- This change migrates away from using `UserDefaults`,
instead providing a TOML configuration mechanism for
user configurable settings. All existing system property
settings keys are supported in the new configuration
file. However, users will have to migrate any settings
they have configured in the `UserDefaults` into TOML
for these settings to take effect.
- Breaking changes:
* `container system property get` is removed in favor of
users directly utilizing `container system property list --format toml | jq<>`.
* `container system property set` is removed since the TOML
configuration is effectively immutable during the lifetime of the
`container` daemon. Uses can edit the TOML they have in their home
directory, however no changes will take effect until the daemon is
restarted via `container system stop && container system start`
* `container system property list --format table` is removed as
generating tabular format is non-trivial and the new TOML format is
intended to be human readable
This extends the github actions to by default run the coverage
generation targets. The targets will generate the coverage artifacts,
which will then be used by a seprate github action to comment on the PR
the coverage numbers.
This is a sanity check for testing information for new features and
refactors alike. It will allow reviewers to quickly identify if there
are major issues with new tests, or large changes to coverage percentage
that indicate a problem.
This improves the integration test suite speed by running all
integration tests in a single `swift test` invocation. We maintain the
one `@Suite` at a time behavior by introducing a new test trait that
uses a single mutex to limit inter-suite parallelism.
- Currently we need to use `pfctl` to create the packet filter rule for
container-to-host networking (`host.docker.internal`). This approach
unfortunately has a few limitations that require documentation.
- Part of #1404.
- Evolves API to use NetworkResource that conforms to ManagedResource.
- BREAKING CHANGE - compile time impact due to changes to Swift client
API signatures. No change to persistent data or API server XPC
protocols.
## Type of Change
- [x] Bug fix
- [ ] New feature
- [ ] Breaking change
- [ ] Documentation update
## Motivation and Context
`CLITest.run(arguments:stdin:currentDirectory:env:)` reads stdout to EOF
before stderr begins draining
(`Tests/CLITests/Utilities/CLITest.swift:182-183`). If a child process
writes more than a pipe-buffer's worth of data to stderr (~64 KB on
macOS), the child blocks in `write()` on stderr while we block in
`readDataToEndOfFile()` on stdout, and neither side makes progress until
`process.waitUntilExit()` returns. This is a latent deadlock that any
future test or CLI verbosity bump can trigger.
This PR drains both streams concurrently using `readabilityHandler`
closures backed by `Mutex<Data>` buffers. After
`process.waitUntilExit()` returns the handlers are cleared and
`readDataToEndOfFile()` flushes any bytes the kernel buffered between
the last handler invocation and exit. The error path also clears
handlers so a failed `process.run()` does not leak callbacks. `Mutex`
matches the locking primitive the file already uses for `commandSeq`.
Fixes#1456
## Testing
- [x] Tested locally
- [ ] Added/updated tests
- [ ] Added/updated docs
`swift build --target CLITests` passes against the change. `make
swift-fmt-check` is clean. Behavior is unchanged for any test command
that previously fit within a single pipe buffer, so the existing
CLITests still exercise the helper end-to-end.
I did not add a new test that emits >64 KB to stderr because every
existing CLITest invocation goes through `executablePath` (the
`container` binary), and reproducing the deadlock requires a child that
emits a controllable amount on stderr. Happy to follow up with a small
refactor that extracts the drain into a static helper plus a regression
test that drives it via `/bin/sh` if that would be useful.
---------
Signed-off-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
- Fixes#1281. Supersedes #1306.
- When `-a` runs as a regular user (such as via
`brew install container`), two launchctl failures
bubble up as a non-zero exit and break Homebrew's
post-install:
- `bootout` against `system` requires root.
- `launchctl print` lists Mach endpoint names that
aren't loaded jobs, so `bootout` returns
`No such process` on them.
- Skip `system` when not root, and treat individual
`bootout` failures as non-fatal. Same fix applied to
the non-`-a` `xargs` pipeline.
Signed-off-by: Patrick Linnane <patrick@linnane.io>
-Resolves #1466.
- `container image push <ref>` previously produced no
standard output on success. Print the fully qualified
reference (`image.reference`) after the push completes
so callers can pipe the output into subsequent
commands.
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
- Don't prefix match localhost, use exact
match only.
- Only check default local domain if one is
defined.
- Use IPv4Address for private IP CIDR tests.
- A recent change moved the list and delete
network XPC calls within an availability check,
making them only available on macOS 26.
However, there are some code flows that rely
on being able to call `list` on the NetworksService,
including anything that uses `networkClient.builtin`.
- `container` does not officially support macOS 15
and as a result we do not have GitHub action runners
for validation on macOS 15. That said, we try to avoid
breaking compatibility on macOS 15 as much as
possible.
Signed-off-by: Kathryn Baldauf <k_baldauf@apple.com>
- Closes#1450.
- If we require that bundle-packaged installs must populate a
Unix-structured `bin` and `libexec`, we don't need the code that
searches the app bundle install root. As it stands, we enumerate both
the app bundle and the Linux paths, meaning we're looking at the same
plugins (in different paths) twice.
AJ EmoryandGitHub
Saehej KangandGitHub
RajandGitHub
J LoganandGitHub
Shaurya SinghandGitHub
ChengHao YangandGitHub
Noah ThorntonandGitHub
Chris GeorgeandGitHub
iliyaandGitHub
Danny CanterandGitHub
Kathryn BaldaufandGitHub
Patrick LinnaneandGitHub