Starting with go1.19, the Go runtime on Windows now supports the `netgo` build-
flag to use a native Go DNS resolver. Prior to that version, the build-flag
only had an effect on non-Windows platforms. When using the `netgo` build-flag,
the Windows's host resolver is not used, and as a result, custom entries in
`etc/hosts` are ignored, which is a change in behavior from binaries compiled
with older versions of the Go runtime.
From the go1.19 release notes: https://go.dev/doc/go1.19#net
> Resolver.PreferGo is now implemented on Windows and Plan 9. It previously
> only worked on Unix platforms. Combined with Dialer.Resolver and Resolver.Dial,
> it's now possible to write portable programs and be in control of all DNS name
> lookups when dialing.
>
> The net package now has initial support for the netgo build tag on Windows.
> When used, the package uses the Go DNS client (as used by Resolver.PreferGo)
> instead of asking Windows for DNS results. The upstream DNS server it discovers
> from Windows may not yet be correct with complex system network configurations,
> however.
Our Windows binaries are compiled with the "static" (`make/binary-daemon`)
script, which has the `netgo` option set by default. This patch unsets the
`netgo` option when cross-compiling for Windows.
Co-authored-by: Bjorn Neergaard <bjorn.neergaard@docker.com>
Signed-off-by: Bjorn Neergaard <bjorn.neergaard@docker.com>
(cherry picked from commit 53d1b12bc0)
Signed-off-by: Bjorn Neergaard <bjorn.neergaard@docker.com>
Docker with containerd integration emits "Exists" progress action when a
layer of the currently pulled image already exists. This is different
from the non-c8d Docker which emits "Already exists".
This makes both implementations consistent by emitting backwards
compatible "Already exists" action.
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
(cherry picked from commit a7bc65fbd8)
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
TestProxyNXDOMAIN has proven to be susceptible to failing as a
consequence of unlocked threads being set to the wrong network
namespace. As the failure mode looks a lot like a bug in the test
itself, it seems prudent to add a check for mismatched namespaces to the
test so we will know for next time that the root cause lies elsewhere.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 871cf72363)
Signed-off-by: Cory Snider <csnider@mirantis.com>
osl.setIPv6 mistakenly captured the calling goroutine's thread's network
namespace instead of the network namespace of the thread getting its
namespace temporarily changed. As this function appears to only be
called from contexts in the process's initial network namespace, this
mistake would be of little consequence at runtime. The libnetwork unit
tests, on the other hand, unshare network namespaces so as not to
interfere with each other or the host's network namespace. But due to
this bug, the isolation backfires and the network namespace of
goroutines used by a test which are expected to be in the initial
network namespace can randomly become the isolated network namespace of
some other test. Symptoms include a loopback network server running in
one goroutine being inexplicably and randomly being unreachable by a
client in another goroutine.
Capture the original network namespace of the thread from the thread to
be tampered with, after locking the goroutine to the thread.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 6d79864135)
Signed-off-by: Cory Snider <csnider@mirantis.com>
Swapping out the global logger on the fly is causing tests to flake out
by logging to a test's log output after the test function has returned.
Refactor Resolver to use a dependency-injected logger and the resolver
unit tests to inject a private logger instance into the Resolver under
test.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit d4f3858a40)
Signed-off-by: Cory Snider <csnider@mirantis.com>
tstwriter mocks the server-side connection between the resolver and the
container, not the resolver and the external DNS server, so returning
the external DNS server's address as w.LocalAddr() is technically
incorrect and misleading. Only the protocols need to match as the
resolver uses the client's choice of protocol to determine which
protocol to use when forwarding the query to the external DNS server.
While this change has no material impact on the tests, it makes the
tests slightly more comprehensible for the next person.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 0cc6e445d7)
Signed-off-by: Cory Snider <csnider@mirantis.com>
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
(cherry picked from commit 34964c2454)
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
Our resolver is just a forwarder for external DNS so it should act like
it. Unless it's a server failure or refusal, take the response at face
value and forward it along to the client. RFC 8020 is only applicable to
caching recursive name servers and our resolver is neither caching nor
recursive.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 41356227f2)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The error returned by DecodeConfig was changed in
b6d58d749c and caused this to regress.
Allow empty request bodies for this endpoint once again.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 967c7bc5d3)
Signed-off-by: Cory Snider <csnider@mirantis.com>
Make RepoTags and RepoDigests empty arrays instead of nil.
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
(cherry picked from commit 1be26e9f0c)
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
This fixes a bug where, if a user pulls an image with a tag != `latest` and
a specific platform, we return an NotFound error for the wrong (`latest`) tag.
see: https://github.com/moby/moby/issues/45558
This bug was introduced in 779a5b3029
in the changes to `daemon/images/image_pull.go`, when we started returning the error from the call to
`GetImage` after the pull. We do this call, if pulling with a specified platform, to check if the platform
of the pulled image matches the requested platform (for cases with single-arch images).
However, when we call `GetImage` we're not passing the image tag, only name, so `GetImage` assumes `latest`
which breaks when the user has requested a different tag, since there might not be such an image in the store.
Signed-off-by: Laura Brehm <laurabrehm@hey.com>
(cherry picked from commit f450ea64e6)
Signed-off-by: Laura Brehm <laurabrehm@hey.com>
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
(cherry picked from commit 3309e45ca1)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
When the `ServerAddress` in the `AuthConfig` provided by the client is
empty, default to the default registry (registry-1.docker.io).
This makes the behaviour the same as with the containerd image store
integration disabled.
Signed-off-by: Paweł Gronowski <pawel.gronowski@docker.com>
(cherry picked from commit 2ad499f93e)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Now that most uses of reexec have been replaced with non-reexec
solutions, most of the reexec.Init() calls peppered throughout the test
suites are unnecessary. Furthermore, most of the reexec.Init() calls in
test code neglects to check the return value to determine whether to
exit, which would result in the reexec'ed subprocesses proceeding to run
the tests, which would reexec another subprocess which would proceed to
run the tests, recursively. (That would explain why every reexec
callback used to unconditionally call os.Exit() instead of returning...)
Remove unneeded reexec.Init() calls from test and example code which no
longer needs it, and fix the reexec.Init() calls which are not inert to
exit after a reexec callback is invoked.
Signed-off-by: Cory Snider <csnider@mirantis.com>
(cherry picked from commit 4e0319c878)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This workflow will just build the bin-image bake target.
Signed-off-by: CrazyMax <crazy-max@users.noreply.github.com>
(cherry picked from commit 135d8f04f9)
Allows to build a non-runnable image that contains bundles.
Signed-off-by: CrazyMax <crazy-max@users.noreply.github.com>
(cherry picked from commit ae1ca67178)
Ports over all the previous image delete logic, such as:
- Introduce `prune` and `force` flags
- Introduce the concept of hard and soft image delete conflics, which represent:
- image referenced in multiple tags (soft conflict)
- image being used by a stopped container (soft conflict)
- image being used by a running container (hard conflict)
- Implement delete logic such as:
- if deleting by reference, and there are other references to the same image, just
delete the passed reference
- if deleting by reference, and there is only 1 reference and the image is being used
by a running container, throw an error if !force, or delete the reference and create
a dangling reference otherwise
- if deleting by imageID, and force is true, remove all tags (otherwise soft conflict)
- if imageID, check if stopped container is using the image (soft conflict), and
delete anyway if force
- if imageID was passed in, check if running container is using the image (hard conflict)
- if `prune` is true, and the image being deleted has dangling parents, remove them
This commit also implements logic to get image parents in c8d by comparing shared layers.
Signed-off-by: Laura Brehm <laurabrehm@hey.com>
(cherry picked from commit cad97135b3)
Signed-off-by: Laura Brehm <laurabrehm@hey.com>
