Remvoe go1.7 from travis

Signed-off-by: Michael Crosby <crosbymichael@gmail.com>

.travis.yml

@@ -4,7 +4,6 @@ sudo: required
 language: go
 
 go:
-  - 1.7.x
   - 1.8.x
   - tip
 

design/plugins.md

@@ -1,6 +1,6 @@
 # containerd Plugin Model
 
-With go 1.8 we now have dynamically loaded plugins via go packages.  This seems to be a very easy and clean way to extent containerd.  It does have the drawback of only working on Linux right now but there is where we see the most need for swapping out defaults. 
+With go 1.8 we now have dynamically loaded plugins via go packages.  This seems to be a very easy and clean way to extend containerd.  It does have the drawback of only working on Linux right now but this is where we see the most need for swapping out defaults. 
 
 ## core
 
@@ -19,44 +19,11 @@ The core should be comprised of the following:
 
 The runtime code in the core provides API to create, list, and manage containers on the system.  It provides a runtime type that is responsible for creating, deleting, and loading containers.
 
-```go
-type Runtime interface {
-	Create(ctx context.Context, id string, opts CreateOpts) (Container, error)
-	Containers() ([]Container, error)
-	Delete(ctx context.Context, c Container) error
-}
-```
-
-There is a common `Container` interface with common methods for interacting with the container as well as platform specific container interfaces.
-
-
-```go
-type Container interface {
-	Info() ContainerInfo
-	Start(context.Context) error
-	State(context.Context) (State, error)
-	Events(context.Context) (<-chan ContainerEvent, error)
-	Kill(context.Context) error 
-}
-
-type LinuxContainer interface {
-	Pause(context.Context) error
-	Resume(context.Context) error
-	Exec(context.Context, ExecOpts) (uint32, error)
-	Signal(c context.Context, pid uint32, s os.Signal) error 
-}
-
-type WindowsContainer interface {
-	Exec(context.Context, ExecOpts) (uint32, error)
-	Signal(c context.Context, pid uint32, s os.Signal) error 
-}
-```
-
 ### Monitoring 
 
 The monitoring subsystem is a way to collect events and metrics from various subsystems.
 With the monitoring subsystem you can monitor various types, subsystems, and objects within the core.
-This can be use to collect metrics for containers and monitor OOM events when supported.
+This can be used to collect metrics for containers and monitor OOM events when supported.
 An example of this is a prometheus monitor that exports container metrics such as cpu, memory, io, and network information.
 
 ```go

linux/main.go

@@ -0,0 +1,5 @@
+package main
+
+func main() {
+	panic("build as plugin not executable")
+}

linux/runtime.go

@@ -31,12 +31,13 @@ func init() {
 }
 
 func New(ic *containerd.InitContext) (interface{}, error) {
-	if err := os.MkdirAll(ic.State, 0700); err != nil {
+	path := filepath.Join(ic.State, runtimeName)
+	if err := os.MkdirAll(path, 0700); err != nil {
 		return nil, err
 	}
 	c, cancel := context.WithCancel(ic.Context)
 	return &Runtime{
-		root:          ic.State,
+		root:          path,
 		events:        make(chan *containerd.Event, 2048),
 		eventsContext: c,
 		eventsCancel:  cancel,

plugin.go

@@ -46,13 +46,13 @@ var register = struct {
 	r: make(map[string]*Registration),
 }
 
-// Load loads all plugins at the provided that into containerd
+// Load loads all plugins at the provided path into containerd
 func Load(path string) (err error) {
 	defer func() {
 		if v := recover(); v != nil {
 			rerr, ok := v.(error)
 			if !ok {
-				panic(v)
+				rerr = fmt.Errorf("%s", v)
 			}
 			err = rerr
 		}

services/execution/service.go

@@ -133,6 +133,8 @@ func (s *Service) Delete(ctx context.Context, r *api.DeleteRequest) (*google_pro
 
 func (s *Service) List(ctx context.Context, r *api.ListRequest) (*api.ListResponse, error) {
 	resp := &api.ListResponse{}
+	s.mu.Lock()
+	defer s.mu.Unlock()
 	for _, c := range s.containers {
 		state, err := c.State(ctx)
 		if err != nil {