Implement build cache based on history array

Based on work by KJ Tsanaktsidis

Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
Signed-off-by: KJ Tsanaktsidis <kjtsanaktsidis@gmail.com>

api/server/router/build/build_routes.go

@@ -70,7 +70,7 @@ func newImageBuildOptions(ctx context.Context, r *http.Request) (*types.ImageBui
 	var buildUlimits = []*units.Ulimit{}
 	ulimitsJSON := r.FormValue("ulimits")
 	if ulimitsJSON != "" {
-		if err := json.NewDecoder(strings.NewReader(ulimitsJSON)).Decode(&buildUlimits); err != nil {
+		if err := json.Unmarshal([]byte(ulimitsJSON), &buildUlimits); err != nil {
 			return nil, err
 		}
 		options.Ulimits = buildUlimits
@@ -79,7 +79,7 @@ func newImageBuildOptions(ctx context.Context, r *http.Request) (*types.ImageBui
 	var buildArgs = map[string]string{}
 	buildArgsJSON := r.FormValue("buildargs")
 	if buildArgsJSON != "" {
-		if err := json.NewDecoder(strings.NewReader(buildArgsJSON)).Decode(&buildArgs); err != nil {
+		if err := json.Unmarshal([]byte(buildArgsJSON), &buildArgs); err != nil {
 			return nil, err
 		}
 		options.BuildArgs = buildArgs
@@ -87,12 +87,21 @@ func newImageBuildOptions(ctx context.Context, r *http.Request) (*types.ImageBui
 	var labels = map[string]string{}
 	labelsJSON := r.FormValue("labels")
 	if labelsJSON != "" {
-		if err := json.NewDecoder(strings.NewReader(labelsJSON)).Decode(&labels); err != nil {
+		if err := json.Unmarshal([]byte(labelsJSON), &labels); err != nil {
 			return nil, err
 		}
 		options.Labels = labels
 	}
 
+	var cacheFrom = []string{}
+	cacheFromJSON := r.FormValue("cachefrom")
+	if cacheFromJSON != "" {
+		if err := json.Unmarshal([]byte(cacheFromJSON), &cacheFrom); err != nil {
+			return nil, err
+		}
+		options.CacheFrom = cacheFrom
+	}
+
 	return options, nil
 }
 

api/types/client.go

@@ -151,6 +151,9 @@ type ImageBuildOptions struct {
 	// preserves the original image and creates a new one from the parent with all
 	// the changes applied to a single layer
 	Squash bool
+	// CacheFrom specifies images that are used for matching cache. Images
+	// specified here do not need to have a valid parent chain to match cache.
+	CacheFrom []string
 }
 
 // ImageBuildResponse holds information

builder/builder.go

@@ -153,10 +153,16 @@ type Image interface {
 	RunConfig() *container.Config
 }
 
-// ImageCache abstracts an image cache store.
+// ImageCacheBuilder represents a generator for stateful image cache.
+type ImageCacheBuilder interface {
+	// MakeImageCache creates a stateful image cache.
+	MakeImageCache(cacheFrom []string) ImageCache
+}
+
+// ImageCache abstracts an image cache.
 // (parent image, child runconfig) -> child image
 type ImageCache interface {
 	// GetCachedImageOnBuild returns a reference to a cached image whose parent equals `parent`
 	// and runconfig equals `cfg`. A cache miss is expected to return an empty ID and a nil error.
-	GetCachedImageOnBuild(parentID string, cfg *container.Config) (imageID string, err error)
+	GetCache(parentID string, cfg *container.Config) (imageID string, err error)
 }

builder/dockerfile/builder.go

@@ -75,6 +75,8 @@ type Builder struct {
 
 	// TODO: remove once docker.Commit can receive a tag
 	id string
+
+	imageCache builder.ImageCache
 }
 
 // BuildManager implements builder.Backend and is shared across all Builder objects.
@@ -136,6 +138,10 @@ func NewBuilder(clientCtx context.Context, config *types.ImageBuildOptions, back
 			LookingForDirectives: true,
 		},
 	}
+	if icb, ok := backend.(builder.ImageCacheBuilder); ok {
+		b.imageCache = icb.MakeImageCache(config.CacheFrom)
+	}
+
 	parser.SetEscapeToken(parser.DefaultEscapeToken, &b.directive) // Assume the default token for escape
 
 	if dockerfile != nil {

builder/dockerfile/internals.go

@@ -438,18 +438,16 @@ func (b *Builder) processImageFrom(img builder.Image) error {
 	return nil
 }
 
-// probeCache checks if `b.docker` implements builder.ImageCache and image-caching
-// is enabled (`b.UseCache`).
-// If so attempts to look up the current `b.image` and `b.runConfig` pair with `b.docker`.
+// probeCache checks if cache match can be found for current build instruction.
 // If an image is found, probeCache returns `(true, nil)`.
 // If no image is found, it returns `(false, nil)`.
 // If there is any error, it returns `(false, err)`.
 func (b *Builder) probeCache() (bool, error) {
-	c, ok := b.docker.(builder.ImageCache)
-	if !ok || b.options.NoCache || b.cacheBusted {
+	c := b.imageCache
+	if c == nil || b.options.NoCache || b.cacheBusted {
 		return false, nil
 	}
-	cache, err := c.GetCachedImageOnBuild(b.image, b.runConfig)
+	cache, err := c.GetCache(b.image, b.runConfig)
 	if err != nil {
 		return false, err
 	}

cli/command/image/build.go

@@ -55,6 +55,7 @@ type buildOptions struct {
 	rm             bool
 	forceRm        bool
 	pull           bool
+	cacheFrom      []string
 }
 
 // NewBuildCommand creates a new `docker build` command
@@ -98,6 +99,7 @@ func NewBuildCommand(dockerCli *command.DockerCli) *cobra.Command {
 	flags.BoolVar(&options.forceRm, "force-rm", false, "Always remove intermediate containers")
 	flags.BoolVarP(&options.quiet, "quiet", "q", false, "Suppress the build output and print image ID on success")
 	flags.BoolVar(&options.pull, "pull", false, "Always attempt to pull a newer version of the image")
+	flags.StringSliceVar(&options.cacheFrom, "cache-from", []string{}, "Images to consider as cache sources")
 
 	command.AddTrustedFlags(flags, true)
 
@@ -289,6 +291,7 @@ func runBuild(dockerCli *command.DockerCli, options buildOptions) error {
 		BuildArgs:      runconfigopts.ConvertKVStringsToMap(options.buildArgs.GetAll()),
 		AuthConfigs:    authConfig,
 		Labels:         runconfigopts.ConvertKVStringsToMap(options.labels),
+		CacheFrom:      options.cacheFrom,
 	}
 
 	response, err := dockerCli.Client().ImageBuild(ctx, body, buildOptions)

client/image_build.go

@@ -110,6 +110,13 @@ func imageBuildOptionsToQuery(options types.ImageBuildOptions) (url.Values, erro
 		return query, err
 	}
 	query.Set("labels", string(labelsJSON))
+
+	cacheFromJSON, err := json.Marshal(options.CacheFrom)
+	if err != nil {
+		return query, err
+	}
+	query.Set("cachefrom", string(cacheFromJSON))
+
 	return query, nil
 }
 

daemon/cache.go

@@ -0,0 +1,254 @@
+package daemon
+
+import (
+	"encoding/json"
+	"fmt"
+	"reflect"
+	"strings"
+
+	"github.com/Sirupsen/logrus"
+	containertypes "github.com/docker/docker/api/types/container"
+	"github.com/docker/docker/builder"
+	"github.com/docker/docker/dockerversion"
+	"github.com/docker/docker/image"
+	"github.com/docker/docker/layer"
+	"github.com/docker/docker/runconfig"
+	"github.com/pkg/errors"
+)
+
+// getLocalCachedImage returns the most recent created image that is a child
+// of the image with imgID, that had the same config when it was
+// created. nil is returned if a child cannot be found. An error is
+// returned if the parent image cannot be found.
+func (daemon *Daemon) getLocalCachedImage(imgID image.ID, config *containertypes.Config) (*image.Image, error) {
+	// Loop on the children of the given image and check the config
+	getMatch := func(siblings []image.ID) (*image.Image, error) {
+		var match *image.Image
+		for _, id := range siblings {
+			img, err := daemon.imageStore.Get(id)
+			if err != nil {
+				return nil, fmt.Errorf("unable to find image %q", id)
+			}
+
+			if runconfig.Compare(&img.ContainerConfig, config) {
+				// check for the most up to date match
+				if match == nil || match.Created.Before(img.Created) {
+					match = img
+				}
+			}
+		}
+		return match, nil
+	}
+
+	// In this case, this is `FROM scratch`, which isn't an actual image.
+	if imgID == "" {
+		images := daemon.imageStore.Map()
+		var siblings []image.ID
+		for id, img := range images {
+			if img.Parent == imgID {
+				siblings = append(siblings, id)
+			}
+		}
+		return getMatch(siblings)
+	}
+
+	// find match from child images
+	siblings := daemon.imageStore.Children(imgID)
+	return getMatch(siblings)
+}
+
+// MakeImageCache creates a stateful image cache.
+func (daemon *Daemon) MakeImageCache(sourceRefs []string) builder.ImageCache {
+	if len(sourceRefs) == 0 {
+		return &localImageCache{daemon}
+	}
+
+	cache := &imageCache{daemon: daemon, localImageCache: &localImageCache{daemon}}
+
+	for _, ref := range sourceRefs {
+		img, err := daemon.GetImage(ref)
+		if err != nil {
+			logrus.Warnf("Could not look up %s for cache resolution, skipping: %+v", ref, err)
+			continue
+		}
+		cache.sources = append(cache.sources, img)
+	}
+
+	return cache
+}
+
+// localImageCache is cache based on parent chain.
+type localImageCache struct {
+	daemon *Daemon
+}
+
+func (lic *localImageCache) GetCache(imgID string, config *containertypes.Config) (string, error) {
+	return getImageIDAndError(lic.daemon.getLocalCachedImage(image.ID(imgID), config))
+}
+
+// imageCache is cache based on history objects. Requires initial set of images.
+type imageCache struct {
+	sources         []*image.Image
+	daemon          *Daemon
+	localImageCache *localImageCache
+}
+
+func (ic *imageCache) restoreCachedImage(parent, target *image.Image, cfg *containertypes.Config) (image.ID, error) {
+	var history []image.History
+	rootFS := image.NewRootFS()
+	lenHistory := 0
+	if parent != nil {
+		history = parent.History
+		rootFS = parent.RootFS
+		lenHistory = len(parent.History)
+	}
+	history = append(history, target.History[lenHistory])
+	if layer := getLayerForHistoryIndex(target, lenHistory); layer != "" {
+		rootFS.Append(layer)
+	}
+
+	config, err := json.Marshal(&image.Image{
+		V1Image: image.V1Image{
+			DockerVersion: dockerversion.Version,
+			Config:        cfg,
+			Architecture:  target.Architecture,
+			OS:            target.OS,
+			Author:        target.Author,
+			Created:       history[len(history)-1].Created,
+		},
+		RootFS:     rootFS,
+		History:    history,
+		OSFeatures: target.OSFeatures,
+		OSVersion:  target.OSVersion,
+	})
+	if err != nil {
+		return "", errors.Wrap(err, "failed to marshal image config")
+	}
+
+	imgID, err := ic.daemon.imageStore.Create(config)
+	if err != nil {
+		return "", errors.Wrap(err, "failed to create cache image")
+	}
+
+	if parent != nil {
+		if err := ic.daemon.imageStore.SetParent(imgID, parent.ID()); err != nil {
+			return "", errors.Wrapf(err, "failed to set parent for %v to %v", target.ID(), parent.ID())
+		}
+	}
+	return imgID, nil
+}
+
+func (ic *imageCache) isParent(imgID, parentID image.ID) bool {
+	nextParent, err := ic.daemon.imageStore.GetParent(imgID)
+	if err != nil {
+		return false
+	}
+	if nextParent == parentID {
+		return true
+	}
+	return ic.isParent(nextParent, parentID)
+}
+
+func (ic *imageCache) GetCache(parentID string, cfg *containertypes.Config) (string, error) {
+	imgID, err := ic.localImageCache.GetCache(parentID, cfg)
+	if err != nil {
+		return "", err
+	}
+	if imgID != "" {
+		for _, s := range ic.sources {
+			if ic.isParent(s.ID(), image.ID(imgID)) {
+				return imgID, nil
+			}
+		}
+	}
+
+	var parent *image.Image
+	lenHistory := 0
+	if parentID != "" {
+		parent, err = ic.daemon.imageStore.Get(image.ID(parentID))
+		if err != nil {
+			return "", errors.Wrapf(err, "unable to find image %v", parentID)
+		}
+		lenHistory = len(parent.History)
+	}
+
+	for _, target := range ic.sources {
+		if !isValidParent(target, parent) || !isValidConfig(cfg, target.History[lenHistory]) {
+			continue
+		}
+
+		if len(target.History)-1 == lenHistory { // last
+			if parent != nil {
+				if err := ic.daemon.imageStore.SetParent(target.ID(), parent.ID()); err != nil {
+					return "", errors.Wrapf(err, "failed to set parent for %v to %v", target.ID(), parent.ID())
+				}
+			}
+			return target.ID().String(), nil
+		}
+
+		imgID, err := ic.restoreCachedImage(parent, target, cfg)
+		if err != nil {
+			return "", errors.Wrapf(err, "failed to restore cached image from %q to %v", parentID, target.ID())
+		}
+
+		ic.sources = []*image.Image{target} // avoid jumping to different target, tuned for safety atm
+		return imgID.String(), nil
+	}
+
+	return "", nil
+}
+
+func getImageIDAndError(img *image.Image, err error) (string, error) {
+	if img == nil || err != nil {
+		return "", err
+	}
+	return img.ID().String(), nil
+}
+
+func isValidParent(img, parent *image.Image) bool {
+	if len(img.History) == 0 {
+		return false
+	}
+	if parent == nil || len(parent.History) == 0 && len(parent.RootFS.DiffIDs) == 0 {
+		return true
+	}
+	if len(parent.History) >= len(img.History) {
+		return false
+	}
+	if len(parent.RootFS.DiffIDs) >= len(img.RootFS.DiffIDs) {
+		return false
+	}
+
+	for i, h := range parent.History {
+		if !reflect.DeepEqual(h, img.History[i]) {
+			return false
+		}
+	}
+	for i, d := range parent.RootFS.DiffIDs {
+		if d != img.RootFS.DiffIDs[i] {
+			return false
+		}
+	}
+	return true
+}
+
+func getLayerForHistoryIndex(image *image.Image, index int) layer.DiffID {
+	layerIndex := 0
+	for i, h := range image.History {
+		if i == index {
+			if h.EmptyLayer {
+				return ""
+			}
+			break
+		}
+		if !h.EmptyLayer {
+			layerIndex++
+		}
+	}
+	return image.RootFS.DiffIDs[layerIndex] // validate?
+}
+
+func isValidConfig(cfg *containertypes.Config, h image.History) bool {
+	// todo: make this format better than join that loses data
+	return strings.Join(cfg.Cmd, " ") == h.CreatedBy
+}

daemon/image.go

@@ -3,11 +3,9 @@ package daemon
 import (
 	"fmt"
 
-	containertypes "github.com/docker/docker/api/types/container"
 	"github.com/docker/docker/builder"
 	"github.com/docker/docker/image"
 	"github.com/docker/docker/reference"
-	"github.com/docker/docker/runconfig"
 )
 
 // ErrImageDoesNotExist is error returned when no image can be found for a reference.
@@ -71,54 +69,3 @@ func (daemon *Daemon) GetImageOnBuild(name string) (builder.Image, error) {
 	}
 	return img, nil
 }
-
-// GetCachedImage returns the most recent created image that is a child
-// of the image with imgID, that had the same config when it was
-// created. nil is returned if a child cannot be found. An error is
-// returned if the parent image cannot be found.
-func (daemon *Daemon) GetCachedImage(imgID image.ID, config *containertypes.Config) (*image.Image, error) {
-	// Loop on the children of the given image and check the config
-	getMatch := func(siblings []image.ID) (*image.Image, error) {
-		var match *image.Image
-		for _, id := range siblings {
-			img, err := daemon.imageStore.Get(id)
-			if err != nil {
-				return nil, fmt.Errorf("unable to find image %q", id)
-			}
-
-			if runconfig.Compare(&img.ContainerConfig, config) {
-				// check for the most up to date match
-				if match == nil || match.Created.Before(img.Created) {
-					match = img
-				}
-			}
-		}
-		return match, nil
-	}
-
-	// In this case, this is `FROM scratch`, which isn't an actual image.
-	if imgID == "" {
-		images := daemon.imageStore.Map()
-		var siblings []image.ID
-		for id, img := range images {
-			if img.Parent == imgID {
-				siblings = append(siblings, id)
-			}
-		}
-		return getMatch(siblings)
-	}
-
-	// find match from child images
-	siblings := daemon.imageStore.Children(imgID)
-	return getMatch(siblings)
-}
-
-// GetCachedImageOnBuild returns a reference to a cached image whose parent equals `parent`
-// and runconfig equals `cfg`. A cache miss is expected to return an empty ID and a nil error.
-func (daemon *Daemon) GetCachedImageOnBuild(imgID string, cfg *containertypes.Config) (string, error) {
-	cache, err := daemon.GetCachedImage(image.ID(imgID), cfg)
-	if cache == nil || err != nil {
-		return "", err
-	}
-	return cache.ID().String(), nil
-}

docs/reference/api/docker_remote_api.md

@@ -124,6 +124,7 @@ This section lists each version from latest to oldest.  Each listing includes a
 * `POST /containers/create` now validates IPAMConfig in NetworkingConfig, and returns error for invalid IPv4 and IPv6 addresses (`--ip` and `--ip6` in `docker create/run`).
 * `POST /containers/create` now takes a `Mounts` field in `HostConfig` which replaces `Binds` and `Volumes`. *note*: `Binds` and `Volumes` are still available but are exclusive with `Mounts`
 * `POST /build` now performs a preliminary validation of the `Dockerfile` before starting the build, and returns an error if the syntax is incorrect. Note that this change is _unversioned_ and applied to all API versions.
+* `POST /build` accepts `cachefrom` parameter to specify images used for build cache.
 
 ### v1.24 API changes
 

docs/reference/api/docker_remote_api_v1.25.md

@@ -1715,6 +1715,7 @@ or being killed.
         there must be a file with the corresponding path inside the tarball.
 -   **q** – Suppress verbose build output.
 -   **nocache** – Do not use the cache when building the image.
+-   **cachefrom** - JSON array of images used for build cache resolution.
 -   **pull** - Attempt to pull the image even if an older image exists locally.
 -   **rm** - Remove intermediate containers after a successful build (default behavior).
 -   **forcerm** - Always remove intermediate containers (includes `rm`).

docs/reference/builder.md

@@ -106,6 +106,13 @@ the `Using cache` message in the console output.
      ---> 7ea8aef582cc
     Successfully built 7ea8aef582cc
 
+Build cache is only used from images that have a local parent chain. This means
+that these images were created by previous builds or the whole chain of images
+was loaded with `docker load`. If you wish to use build cache of a specific
+image you can specify it with `--cache-from` option. Images specified with
+`--cache-from` do not need to have a parent chain and may be pulled from other
+registries.
+
 When you're done with your build, you're ready to look into [*Pushing a
 repository to its registry*](../tutorials/dockerrepos.md#contributing-to-docker-hub).
 

docs/reference/commandline/build.md

@@ -17,6 +17,7 @@ Build an image from a Dockerfile
 
 Options:
       --build-arg value         Set build-time variables (default [])
+      --cache-from value        Images to consider as cache sources (default [])
       --cgroup-parent string    Optional parent cgroup for the container
       --cpu-period int          Limit the CPU CFS (Completely Fair Scheduler) period
       --cpu-quota int           Limit the CPU CFS (Completely Fair Scheduler) quota