spiegel_podman/pkg/domain/infra/abi/terminal/sigproxy_linux.go
Matthew Heon ed9edf840a When container stops, drop sig-proxy errors to infos
The sig-proxy code is set up to error on failing to forward
signals to a container. This is reasonable in cases where the
container is running, but something strange went wrong - but when
the Kill fails because the container is stopped, we shouldn't
bother with aggressive Error logging since this is an expected
part of the container lifecycle - it stops, and then `podman run`
also stops, but there is a timing window in between where signals
will fail to be proxied, and we should not print angry errors
during that.

Signed-off-by: Matthew Heon <mheon@redhat.com>
2020-10-30 09:23:47 -04:00

56 lines
1.7 KiB
Go

package terminal
import (
"os"
"syscall"
"github.com/containers/podman/v2/libpod"
"github.com/containers/podman/v2/libpod/define"
"github.com/containers/podman/v2/libpod/shutdown"
"github.com/containers/podman/v2/pkg/signal"
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
)
// ProxySignals ...
func ProxySignals(ctr *libpod.Container) {
// Stop catching the shutdown signals (SIGINT, SIGTERM) - they're going
// to the container now.
shutdown.Stop()
sigBuffer := make(chan os.Signal, 128)
signal.CatchAll(sigBuffer)
logrus.Debugf("Enabling signal proxying")
go func() {
for s := range sigBuffer {
// Ignore SIGCHLD and SIGPIPE - these are mostly likely
// intended for the podman command itself.
// SIGURG was added because of golang 1.14 and its preemptive changes
// causing more signals to "show up".
// https://github.com/containers/podman/issues/5483
if s == syscall.SIGCHLD || s == syscall.SIGPIPE || s == syscall.SIGURG {
continue
}
if err := ctr.Kill(uint(s.(syscall.Signal))); err != nil {
if errors.Cause(err) == define.ErrCtrStateInvalid {
logrus.Infof("Ceasing signal forwarding to container %s as it has stopped", ctr.ID())
} else {
logrus.Errorf("Error forwarding signal %d to container %s: %v", s, ctr.ID(), err)
}
// If the container dies, and we find out here,
// we need to forward that one signal to
// ourselves so that it is not lost, and then
// we terminate the proxy and let the defaults
// play out.
signal.StopCatch(sigBuffer)
if err := syscall.Kill(syscall.Getpid(), s.(syscall.Signal)); err != nil {
logrus.Errorf("failed to kill pid %d", syscall.Getpid())
}
return
}
}
}()
}