containerd/runtime/v1/linux/proc/exec.go
Michael Crosby 831a41b958 Fix process locking and state management
There were races with the way process states.  This displayed in ways,
especially around pausing the container for atomic operations.  Users
would get errors like, cannnot delete container in paused state and
such.

This can be eaisly reproduced with `docker` and the following command:

```bash
> (for i in `seq 1 25`; do id=$(docker create  alpine usleep 50000);docker start $id;docker commit $id;docker wait $id;docker rm $id; done)
```

This two issues that this fixes are:

* locks must be held by the owning process, not the state operations.
* If a container ends up being paused but before the operation
completes, the process exists, make sure we resume the container before
setting the the process as exited.

Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
2018-11-09 11:40:37 -05:00

256 lines
5.9 KiB
Go

// +build !windows
/*
Copyright The containerd Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package proc
import (
"context"
"fmt"
"io"
"os"
"path/filepath"
"sync"
"syscall"
"time"
"golang.org/x/sys/unix"
"github.com/containerd/console"
"github.com/containerd/containerd/runtime/proc"
"github.com/containerd/fifo"
runc "github.com/containerd/go-runc"
specs "github.com/opencontainers/runtime-spec/specs-go"
"github.com/pkg/errors"
)
type execProcess struct {
wg sync.WaitGroup
execState execState
mu sync.Mutex
id string
console console.Console
io runc.IO
status int
exited time.Time
pid int
closers []io.Closer
stdin io.Closer
stdio proc.Stdio
path string
spec specs.Process
parent *Init
waitBlock chan struct{}
}
func (e *execProcess) Wait() {
<-e.waitBlock
}
func (e *execProcess) ID() string {
return e.id
}
func (e *execProcess) Pid() int {
e.mu.Lock()
defer e.mu.Unlock()
return e.pid
}
func (e *execProcess) ExitStatus() int {
e.mu.Lock()
defer e.mu.Unlock()
return e.status
}
func (e *execProcess) ExitedAt() time.Time {
e.mu.Lock()
defer e.mu.Unlock()
return e.exited
}
func (e *execProcess) SetExited(status int) {
e.mu.Lock()
defer e.mu.Unlock()
e.execState.SetExited(status)
}
func (e *execProcess) setExited(status int) {
e.status = status
e.exited = time.Now()
e.parent.Platform.ShutdownConsole(context.Background(), e.console)
close(e.waitBlock)
}
func (e *execProcess) Delete(ctx context.Context) error {
e.mu.Lock()
defer e.mu.Unlock()
return e.execState.Delete(ctx)
}
func (e *execProcess) delete(ctx context.Context) error {
e.wg.Wait()
if e.io != nil {
for _, c := range e.closers {
c.Close()
}
e.io.Close()
}
pidfile := filepath.Join(e.path, fmt.Sprintf("%s.pid", e.id))
// silently ignore error
os.Remove(pidfile)
return nil
}
func (e *execProcess) Resize(ws console.WinSize) error {
e.mu.Lock()
defer e.mu.Unlock()
return e.execState.Resize(ws)
}
func (e *execProcess) resize(ws console.WinSize) error {
if e.console == nil {
return nil
}
return e.console.Resize(ws)
}
func (e *execProcess) Kill(ctx context.Context, sig uint32, _ bool) error {
e.mu.Lock()
defer e.mu.Unlock()
return e.execState.Kill(ctx, sig, false)
}
func (e *execProcess) kill(ctx context.Context, sig uint32, _ bool) error {
pid := e.pid
if pid != 0 {
if err := unix.Kill(pid, syscall.Signal(sig)); err != nil {
return errors.Wrapf(checkKillError(err), "exec kill error")
}
}
return nil
}
func (e *execProcess) Stdin() io.Closer {
return e.stdin
}
func (e *execProcess) Stdio() proc.Stdio {
return e.stdio
}
func (e *execProcess) Start(ctx context.Context) error {
e.mu.Lock()
defer e.mu.Unlock()
return e.execState.Start(ctx)
}
func (e *execProcess) start(ctx context.Context) (err error) {
var (
socket *runc.Socket
pidfile = filepath.Join(e.path, fmt.Sprintf("%s.pid", e.id))
)
if e.stdio.Terminal {
if socket, err = runc.NewTempConsoleSocket(); err != nil {
return errors.Wrap(err, "failed to create runc console socket")
}
defer socket.Close()
} else if e.stdio.IsNull() {
if e.io, err = runc.NewNullIO(); err != nil {
return errors.Wrap(err, "creating new NULL IO")
}
} else {
if e.io, err = runc.NewPipeIO(e.parent.IoUID, e.parent.IoGID, withConditionalIO(e.stdio)); err != nil {
return errors.Wrap(err, "failed to create runc io pipes")
}
}
opts := &runc.ExecOpts{
PidFile: pidfile,
IO: e.io,
Detach: true,
}
if socket != nil {
opts.ConsoleSocket = socket
}
if err := e.parent.runtime.Exec(ctx, e.parent.id, e.spec, opts); err != nil {
close(e.waitBlock)
return e.parent.runtimeError(err, "OCI runtime exec failed")
}
if e.stdio.Stdin != "" {
sc, err := fifo.OpenFifo(ctx, e.stdio.Stdin, syscall.O_WRONLY|syscall.O_NONBLOCK, 0)
if err != nil {
return errors.Wrapf(err, "failed to open stdin fifo %s", e.stdio.Stdin)
}
e.closers = append(e.closers, sc)
e.stdin = sc
}
var copyWaitGroup sync.WaitGroup
if socket != nil {
console, err := socket.ReceiveMaster()
if err != nil {
return errors.Wrap(err, "failed to retrieve console master")
}
if e.console, err = e.parent.Platform.CopyConsole(ctx, console, e.stdio.Stdin, e.stdio.Stdout, e.stdio.Stderr, &e.wg, &copyWaitGroup); err != nil {
return errors.Wrap(err, "failed to start console copy")
}
} else if !e.stdio.IsNull() {
if err := copyPipes(ctx, e.io, e.stdio.Stdin, e.stdio.Stdout, e.stdio.Stderr, &e.wg, &copyWaitGroup); err != nil {
return errors.Wrap(err, "failed to start io pipe copy")
}
}
copyWaitGroup.Wait()
pid, err := runc.ReadPidFile(opts.PidFile)
if err != nil {
return errors.Wrap(err, "failed to retrieve OCI runtime exec pid")
}
e.pid = pid
return nil
}
func (e *execProcess) Status(ctx context.Context) (string, error) {
s, err := e.parent.Status(ctx)
if err != nil {
return "", err
}
// if the container as a whole is in the pausing/paused state, so are all
// other processes inside the container, use container state here
switch s {
case "paused", "pausing":
return s, nil
}
e.mu.Lock()
defer e.mu.Unlock()
// if we don't have a pid then the exec process has just been created
if e.pid == 0 {
return "created", nil
}
// if we have a pid and it can be signaled, the process is running
if err := unix.Kill(e.pid, 0); err == nil {
return "running", nil
}
// else if we have a pid but it can nolonger be signaled, it has stopped
return "stopped", nil
}