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>
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>
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>
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>
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>
This reverts most of commit 24c05b7, as otherwise it causes
a few regressions (docker cli, TestDockerSwarmSuite/TestServiceLogsTTY).
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
The TestExecInTTY test case is sometimes failing like this:
> execin_test.go:332: unexpected carriage-return in output "PID USER TIME COMMAND\r\n 1 root 0:00 cat\r\n 7 root 0:00 ps\r\n"
or this:
> execin_test.go:332: unexpected carriage-return in output "PID USER TIME COMMAND\r\n 1 root 0:00 cat\n 7 root 0:00 ps\n"
(this is easy to repro with `go test -run TestExecInTTY -count 1000`).
This is caused by a race between
- an Init() (in this case it is is (*linuxSetnsInit.Init(), but
(*linuxStandardInit).Init() is no different in this regard),
which creates a pty pair, sends pty master to runc, and execs
the container process,
and
- a parent runc process, which receives the pty master fd and calls
ClearONLCR() on it.
One way of fixing it would be to add a synchronization mechanism
between these two, so Init() won't exec the process until the parent
sets the flag. This seems excessive, though, as we can just move
the ClearONLCR() call to Init(), putting it right after console.NewPty().
Note that bug only happens in the TestExecInTTY test case, but
from looking at the code it seems like it can happen in runc run
or runc exec, too.
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
This fixes the following failure:
> sudo runc run -b bundle ctr </dev/null
> WARN[0000] exit status 2
> ERRO[0000] container_linux.go:367: starting container process caused: process_linux.go:459: container init caused:
The "exit status 2" with no error message is caused by SIGHUP
which is sent to init by the kernel when we are losing the
controlling terminal. If we choose to ignore that, we'll get
panic in console.Current(), which is addressed by [1].
Otherwise, the issue here is simple: the code assumes stdin
is opened to a terminal, and fails to work otherwise. Some
standard Linux tools (e.g. stty, top) do the same (modulo panic),
while some others (reset, tput) use the trick of trying
all the three std streams (starting with stderr as it is least likely
to be redirected), and if all three fails, open /dev/tty.
This commit does a similar thing (see initHostConsole).
It also replaces the call to console.Current(), which may panic
(see [1]), by reusing the t.hostConsole.
Finally, a simple test case is added.
Fixes: https://github.com/opencontainers/runc/issues/2485
[1] https://github.com/containerd/console/pull/37
Signed-off-by: Kir Kolyshkin <kolyshkin@gmail.com>
ec0d23a92f ("tty: close epollConsole on errors") fixed a significant
issue, but the cleanup was not ideal (especially if the function is
changed in future to add additional error conditions to those currently
present). Using the defer-named-error trick avoids this issue and makes
the code more readable.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
make sure epollConsole is closed before returning an error. It solves
a hang when using these commands with a container that uses a
terminal:
runc run foo &
ssh root@localhost runc exec foo echo hello
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
This removes usages of docker/pkg/term to set raw terminal, handle interrupt
and restore the terminal, and instead use containerd/console and handle
interrupt ourselves.
Signed-off-by: Daniel Dao <dqminh89@gmail.com>
This moves all console code to use github.com/containerd/console library to
handle console I/O. Also move to use EpollConsole by default when user requests
a terminal so we can still cope when the other side temporarily goes away.
Signed-off-by: Daniel Dao <dqminh89@gmail.com>
And use it only in local tooling that is forwarding the pseudoterminal
master. That way runC no longer has an opinion on the onlcr setting
for folks who are creating a terminal and detaching. They'll use
--console-socket and can setup the pseudoterminal however they like
without runC having an opinion. With this commit, the only cases
where runC still has applies SaneTerminal is when *it* is the process
consuming the master descriptor.
Signed-off-by: W. Trevor King <wking@tremily.us>
This is a fix for rootless containers and general io handling. The
higher level systems must preparte the IO for the container in the
detach case and make sure it is setup correctly for the container's
process.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
This allows for higher-level orchestrators to be able to have access to
the master pty file descriptor without keeping the runC process running.
This is key to having (detach && createTTY) with a _real_ pty created
inside the container, which is then sent to a higher level orchestrator
over an AF_UNIX socket.
This patch is part of the console rewrite patchset.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
This implements {createTTY, detach} and all of the combinations and
negations of the two that were previously implemented. There are some
valid questions about out-of-OCI-scope topics like !createTTY and how
things should be handled (why do we dup the current stdio to the
process, and how is that not a security issue). However, these will be
dealt with in a separate patchset.
In order to allow for late console setup, split setupRootfs into the
"preparation" section where all of the mounts are created and the
"finalize" section where we pivot_root and set things as ro. In between
the two we can set up all of the console mountpoints and symlinks we
need.
We use two-stage synchronisation to ensures that when the syscalls are
reordered in a suboptimal way, an out-of-place read() on the parentPipe
will not gobble the ancilliary information.
This patch is part of the console rewrite patchset.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
In user namespaces, we need to make sure we don't chown() the console to
unmapped users. This means we need to get both the UID and GID of the
root user in the container when changing the owner.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Add a waitgroup to wait for the io.Copy of stdout/err to finish before
existing runc. The problem happens more in exec because it is really
fast and the pipe has data buffered but not yet read after the process
has already exited.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
This flag allows systems that are running runc to allocate tty's that
they own and provide to the container.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
When we are using user namespaces we need to make sure that when we do
not have a TTY we change the ownership of the pipe()'s used for the
process to the root user within the container so that when you call
open() on any of the /proc/self/fd/*'s you do not get an EPERM.
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
When we use ```cat | runc``` or ```runc /dev/stdin < config.json```,
it will fail and output ```FATA[0000] Container start failed: inappropriate ioctl for device```.
It is hard for the users to find out the reason from the message:
the config.json enables the terminal but the user redirect the stdin
to an non-terminal file.
After this patch, the output will be
```FATA[0000] Container start failed: Failed to set the terminal from the stdin: inappropriate ioctl for device```
So the user can disable the terminal in the config.json.
See the #202
Cc: W. Trevor King <wking@tremily.us> (@wking)
Signed-off-by: Lai Jiangshan <jiangshanlai@gmail.com>
Should compile now without errors but changes needed to be added for each system so it actually works.
main_unsupported.go is a new file with all the unsupported commands
Fixes#9
Signed-off-by: Marianna <mtesselh@gmail.com>
