In this way, we can restore the Terminal as soon as possible once the hijacked
connection end. This not only fix weird output if cli enable -D, but also
remove duplicate code.
Signed-off-by: Lei Jitang <leijitang@huawei.com>
Upstream-commit: 66d3dcc6f7648ab5af3c7f876c2b9c63ab2899ac
Component: engine
More descriptive error when running a container with a too long hostname (#21445)
Upstream-commit: 1e8d1b5faaa113944ed0bc6b565b90c490a5b654
Component: engine
These are the changes required due to the new binaries that containerd introduced.
The rpm, and deb packages now include 5 binaries.
docker, containerd, containerd-shim, ctr, and runc
The tar files also include all 5 binaries.
Signed-off-by: Ken Cochrane <KenCochrane@gmail.com>
Upstream-commit: bb66d7144f7d0a617f5985486ae36bbbaa5461ba
Component: engine
.bashrc was removed in 29fbc9cc1d67759974ec4cb746840ae160d842bc
but is still needed. It serves as a hook to customize the build environment.
Signed-off-by: Harald Albers <github@albersweb.de>
Upstream-commit: 8850b223ed4c04cc671a977cadc3942e5e5f7f64
Component: engine
This fix tries to fix issues encountered when running a container with a hostname
that is longer than HOST_NAME_MAX(64).
Previously, `could not synchronise with container process` was generated as the
length of the regex check was missing.
This fix covers the length check so that a hostname that is longer than
HOST_NAME_MAX(64) will be given a correct error message.
Several unit tests cases and additional integration test cases are added as well.
This fix closes#21445.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
Upstream-commit: fa44b4e81ee87f1c2b39ad335a22795415c8cdf2
Component: engine
The server configuration already keeps the current version
if the daemon. This patch changes the middleware logic
to use it rather than using the global value.
This removes the dockerversion package dependency from the api.
Signed-off-by: David Calavera <david.calavera@gmail.com>
Upstream-commit: accf28a7dbdd46203c661d30b80326df4d447cea
Component: engine
Contains fixes for:
- pid.max fix that is causing hang on network stats test.
- fix for early stdin close containerd-shim
- better logging for `could not synchronise with container process`
Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
Upstream-commit: 22d997b3745e278d5a2b0b85d56c0d2b166e3a74
Component: engine
Instead of implementing refcounts at each graphdriver, implement this in
the layer package which is what the engine actually interacts with now.
This means interacting directly with the graphdriver is no longer
explicitly safe with regard to Get/Put calls being refcounted.
In addition, with the containerd, layers may still be mounted after
a daemon restart since we will no longer explicitly kill containers when
we shutdown or startup engine.
Because of this ref counts would need to be repopulated.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Upstream-commit: 65d79e3e5e537039b244afd7eda29e721a93d84f
Component: engine
When a plugin is first found, it is loaded into the available plugins
even though it's not activated yet.
If activation fails it is taken out of the list.
While it is in the list, other callers may see it and try to check it's
manifest. If it is not fully activated yet, the manifest will be nil and
cause a panic.
This is especially problematic for drivers that are down and have not
been activated yet.
We could just not load the plugin into the available list until it's
fully active, however that will just cause multiple of the same plugin
to attemp to be loaded.
We could check if the manifest is nil and return early (instead of
panicing on a nil manifest), but this will cause a 2nd caller to receive
a response while the first caller is still waiting, which can be
awkward.
This change uses a condition variable to handle activation (instead of
sync.Once). If the plugin is not activated, callers will all wait until
it is activated and receive a broadcast from the condition variable
signaling that it's ok to proceed, in which case we'll check if their
was an error in activation and proceed accordingly.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Upstream-commit: baac2f48674d42863f9fdca5ef3056b4c6ed707e
Component: engine
This fix tries to fix Docker core dumps when removing network with special
characters. The issue is from the fact that when docker client tries to
pass the command to API, the networkID is not escaped in case of special
characters. This also means other commands (not just `docker network rm`)
may face the same issue (e.g., `docker network connect`).
This fix adds the URL path escape to properly handle it. In addition, an
integration test for network create and delete is added to cover the cases
in #21401.
This fix fixes#21401.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
(cherry picked from commit f8dc5562d0a74792c46b9382181ddf97e5d7cdac)
Signed-off-by: David Calavera <david.calavera@gmail.com>
Upstream-commit: f1542276081aa91396495059638d3ea30d4e5a10
Component: engine
Removes the seccomp buildtag when building runc.
Because seccomp isn't currently being built, this would cause
the build to fail.
Signed-off-by: Christopher Jones <tophj@linux.vnet.ibm.com>
Upstream-commit: 2caf09d37b5865882dfa8c60d5362fd6d2ba54fd
Component: engine
Use aws instead of s3cmd.
Make sure the release script works with the docker-prepended binary
names.
Signed-off-by: Tibor Vass <tibor@docker.com>
Upstream-commit: dd51e85c050076be09fc7414bda1f261b61e60ad
Component: engine
